From webmaster at mn-linux.org Mon Feb 4 15:24:37 2008 From: webmaster at mn-linux.org (TCLUG Classifieds) Date: Mon, 4 Feb 2008 15:24:37 -0600 Subject: [tclug-list] New TCLUG Classified Ad Message-ID: <200802042124.m14LObM15962@crusader.real-time.com> New TCLUG Classified Ad Category: Computer Type of Ad: For Sale Subject: Panasas ActiveStor 3100 Panasas ACTIVESTOR 3100 High-Performance Parallel Storage for Linux Clusters Each unit is configured as follows: Panasas Storage Cluster with 10x500GB Storage Disk Blade Array Director Blade with 4GB RAM Dual Power Supplies Battery Backup Module 4xGIG E Connections Each unit has 5TB Raw storage and we have 10-pcs available @ $3995ea Sincerely, Casey M. DuBois N-VINT, Inc. 3240 Hanna Lake Industrial Park Dr. SE, Caledonia, MI 49316 616-656-5500 x20 Office 866-337-2686 Direct AOL IM: CaseyNVINT cdubois at n-vint.com ?To err is human, but to really foul things up you need a computer.? Seller Email address: cdubois at n-vint dot com http://www.mn-linux.org/cgi-bin/classifieds/index.cgi From samir.nassar+tclug at steamedpenguin.com Mon Feb 4 19:03:59 2008 From: samir.nassar+tclug at steamedpenguin.com (Samir M. Nassar) Date: Mon, 4 Feb 2008 19:03:59 -0600 Subject: [tclug-list] [OT] Meetup with Dries Buytert and the Drupal "Usability Team" Message-ID: <200802041904.00288.samir.nassar+tclug@steamedpenguin.com> When: Monday, Febrary 25, 2008 at around 18:00 Where: To be determined (around the University of Minnesota) http://groups.drupal.org/node/8648 Who this might interest: System Administrators, Web Developers, Content Managers, Public Relations professionals, journalists, and last but not least, linux users. -- Samir M. Nassar samir.nassar at steamedpenguin.com From erikerik at gmail.com Mon Feb 4 22:18:51 2008 From: erikerik at gmail.com (Erik Anderson) Date: Mon, 4 Feb 2008 22:18:51 -0600 Subject: [tclug-list] [OT] Meetup with Dries Buytert and the Drupal "Usability Team" In-Reply-To: <200802041904.00288.samir.nassar+tclug@steamedpenguin.com> References: <200802041904.00288.samir.nassar+tclug@steamedpenguin.com> Message-ID: On Feb 4, 2008 7:03 PM, Samir M. Nassar wrote: > When: Monday, Febrary 25, 2008 at around 18:00 > Where: To be determined (around the University of Minnesota) > > http://groups.drupal.org/node/8648 Very cool - thanks for the heads up. I'm going to try to be there! From dedrizen at yahoo.com Tue Feb 5 10:11:56 2008 From: dedrizen at yahoo.com (Daniel Burke) Date: Tue, 5 Feb 2008 08:11:56 -0800 (PST) Subject: [tclug-list] New to Linux Message-ID: <614465.45276.qm@web50301.mail.re2.yahoo.com> I am new to Linux, though not new to custom built PCs. My wife has given me the okay to spend some of our tax return to put together a new PC. I am hoping to learn some Linux as well as not compete over our one PC connected to the internet. My plan was to go with an ASUS motherboard, video card, 2GB RAM, DVD burner, 250GB hard drive (was thinking of doing a dual-boot with Windows XP on another partition), memory card reader, one of those 85% efficient power supply, and a 45W AMD dual-core processor. To connect to our wireless network I was considering a USB connected external device. The budget for this, plus monitor and case, is projected to be about $1,200.00 in tax refund. My initial list of questions/concerns are: I found that ASUS listed some driver download for Linux. In Windows I would know what to do with it. In Linux I do not. So what do I do? What do I look for to make sure the hardware I choose is compatible with Linux? To keep things simple for my wife, I was thinking of going with KDE. Is this a good choice? Our children are still very young but it won't be long before they will want to play on the computer. Any advice on setup that might make my life easier with the kids accessing the PC over the next decade? And thinking of children, is there any quick options to "lock" a Linux PC? In Windows XP, I just have to press the Window+L on the keyboard when our little two year old boy comes near to prevent him from using my account with my access to do anything (logging off and shutting down in Windows takes too long as little boys can be very quick). So far I am looking at: motherboard: ASUS M2N-SLI Deluxe processor: AMD Athlon X2 2.1GHz 45W Dual-Core I know that eats up a bit of the initial budget. I was hoping that a good backbone would give me future flexibility. Any suggestions or help with the setup of my first Linux machine would be greatly appreciated. --------------------------------- Looking for last minute shopping deals? Find them fast with Yahoo! Search. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/43f6eebf/attachment.htm From auditodd at comcast.net Tue Feb 5 10:33:56 2008 From: auditodd at comcast.net (auditodd at comcast.net) Date: Tue, 05 Feb 2008 16:33:56 +0000 Subject: [tclug-list] New to Linux Message-ID: <020520081633.19338.47A88FF40002E90900004B8A22007343640B0B019B070B9A0E@comcast.net> Daniel, It is really surprising how little you need to depend upon vendor Linux drivers for things to "just work". I bought a FOXCONN motherboard with onboard nVidia video and a 64bit AMD processor. It "just worked" with Kubuntu, no muss, no fuss. Video, sound, keyboard, mouse, LAN, etc. Wireless is another matter. Make sure your wireless choice is supported by Linux. KDE... I've set up Kubuntu v7.10 on an old HP as a spare for my Dad in case is Dell crashes, and he can barely tell the difference between his WindowsXP and the Kubuntu. -- ========== Todd Young -------------- Original message ---------------------- From: Daniel Burke > What do I look for to make sure the hardware I choose is compatible with > Linux? > > To keep things simple for my wife, I was thinking of going with KDE. Is this a > good choice? -------------- next part -------------- An embedded message was scrubbed... From: Daniel Burke Subject: [tclug-list] New to Linux Date: Tue, 5 Feb 2008 16:17:23 +0000 Size: 5791 Url: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/334dbc8d/attachment.eml From teeahr1 at gmail.com Tue Feb 5 11:07:33 2008 From: teeahr1 at gmail.com (p.daniels) Date: Tue, 05 Feb 2008 11:07:33 -0600 Subject: [tclug-list] put grub on the mbr after the fact? Message-ID: <47A897D5.5050206@gmail.com> I just installed the alpha of Kubuntu Hardy. Nice looking, feels quicker than Gutsy (but still one of the slowest KDE distros I've ever seen, WTF people?) Anyway, somehow grub did not get installed to the master boot record. Going in with the Super Grub live CD will boot it without issue. Is there a way to put grub onto the mbr after the fact? This is unfamiliar territory for me, so if the answer is RTFM, please be gentle and just point me to it. Pete can read good :) -pd- From florin at iucha.net Tue Feb 5 11:17:16 2008 From: florin at iucha.net (Florin Iucha) Date: Tue, 5 Feb 2008 11:17:16 -0600 Subject: [tclug-list] put grub on the mbr after the fact? In-Reply-To: <47A897D5.5050206@gmail.com> References: <47A897D5.5050206@gmail.com> Message-ID: <20080205171716.GY15915@iris.iucha.org> On Tue, Feb 05, 2008 at 11:07:33AM -0600, p.daniels wrote: > I just installed the alpha of Kubuntu Hardy. Nice looking, feels quicker > than Gutsy (but still one of the slowest KDE distros I've ever seen, WTF > people?) Anyway, somehow grub did not get installed to the master boot > record. Going in with the Super Grub live CD will boot it without issue. > Is there a way to put grub onto the mbr after the fact? This is > unfamiliar territory for me, so if the answer is RTFM, please be gentle > and just point me to it. Pete can read good :) Yes, it can. Search the archives, I have answered this in detail sometime last year. florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/1f439e0a/attachment.pgp From teeahr1 at gmail.com Tue Feb 5 11:13:04 2008 From: teeahr1 at gmail.com (p.daniels) Date: Tue, 05 Feb 2008 11:13:04 -0600 Subject: [tclug-list] New to Linux In-Reply-To: <614465.45276.qm@web50301.mail.re2.yahoo.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> Message-ID: <47A89920.20104@gmail.com> Daniel Burke wrote: > I found that ASUS listed some driver download for Linux. In Windows I > would know what to do with it. In Linux I do not. So what do I do? What Todd said. With very few exceptions (wireless being at the top of the list), you'll never really need to worry about drivers with any modern Linux distro. Check to make sure your wireless dongle is compatible (I've had good luck with Netgear) before you buy it. Here's a good link for that: http://www.fsf.org/resources/hw/net/wireless/cards.html Also on the subject of drivers, may I suggest Intel onboard video as opposed to buying a card from Nvidia or ATI. The Intel drivers are open source which the big names are not. Practically speaking, this means you'll never need to worry about your video driver. I bought Nvidia cards for years before getting a new tower with intel onboard, and I've never looked back. > To keep things simple for my wife, I was thinking of going with KDE. Is > this a good choice? Live CDs are your friend. Burn one of a Gnome distro, one for KDE, and one for XFCE, and take your pick. I deal with lots of noobs, and I do push Ubuntu (Gnome), even though I personally use KDE. > Our children are still very young but it won't be long before they will > want to play on the computer. Any advice on setup that might make my > life easier with the kids accessing the PC over the next decade? Ubuntu (or any distro that uses the sudo model) makes this very easy. There is no root account by default, and the "do this as root" password is the password of the original user (you). When you make new users, they don't have root access unless you give it to them. I know on Ubuntu when you make a new user, the menu items that require root access don't even appear in their menus. > And thinking of children, is there any quick options to "lock" a Linux > PC? In Windows XP, I just have to press the Window+L on the keyboard > when our little two year old boy comes near to prevent him from using my > account with my access to do anything (logging off and shutting down in > Windows takes too long as little boys can be very quick). Ubuntu (I know, Ubuntu Ubuntu Ubuntu) has a "lock screen" option in the shutdown menu. Other distros may do it differently. Anyway, welcome to TCLUG! From ryan.langseth at gmail.com Tue Feb 5 11:29:37 2008 From: ryan.langseth at gmail.com (Ryan Langseth) Date: Tue, 5 Feb 2008 11:29:37 -0600 Subject: [tclug-list] put grub on the mbr after the fact? In-Reply-To: <20080205171716.GY15915@iris.iucha.org> References: <47A897D5.5050206@gmail.com> <20080205171716.GY15915@iris.iucha.org> Message-ID: The command is grub-install [device_name] . I had to do that on my SuSE box last night for that system it was "grub-install /dev/sda" your device name will probably be different. Ryan On Feb 5, 2008 11:17 AM, Florin Iucha wrote: > On Tue, Feb 05, 2008 at 11:07:33AM -0600, p.daniels wrote: > > I just installed the alpha of Kubuntu Hardy. Nice looking, feels quicker > > than Gutsy (but still one of the slowest KDE distros I've ever seen, WTF > > people?) Anyway, somehow grub did not get installed to the master boot > > record. Going in with the Super Grub live CD will boot it without issue. > > Is there a way to put grub onto the mbr after the fact? This is > > unfamiliar territory for me, so if the answer is RTFM, please be gentle > > and just point me to it. Pete can read good :) > > Yes, it can. > > Search the archives, I have answered this in detail sometime last year. > > florin > > -- > Bruce Schneier expects the Spanish Inquisition. > http://geekz.co.uk/schneierfacts/fact/163 > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > From cschumann at twp-llc.com Tue Feb 5 12:12:25 2008 From: cschumann at twp-llc.com (Chris Schumann) Date: Tue, 5 Feb 2008 12:12:25 -0600 (CST) Subject: [tclug-list] New to Linux In-Reply-To: References: Message-ID: <23170.192.28.2.17.1202235145.squirrel@alpha.twp-llc.com> > Date: Tue, 5 Feb 2008 08:11:56 -0800 (PST) > From: Daniel Burke > I found that ASUS listed some driver download for Linux. In Windows I > would know what to do with it. In Linux I do not. So what do I do? Start by using the Google. Things like the motherboard model and linux will be a good start. You may not have to do anything. Since Asus now ships a Linux machine (EEE PC), they are aware of Linux and might try to make it handy to use on their products. > To keep things simple for my wife, I was thinking of going with KDE. > Is this a good choice? KDE is more configurable than Gnome, so it has more options, which can be a little more confusing. Paradox of choice and all that. There are also Linux distributions that purposely try to look like Windows to reduce the trauma for new users. I like Gnome. > Our children are still very young but it won't be long before they > will want to play on the computer. Any advice on setup that might make > my life easier with the kids accessing the PC over the next decade? Things change quickly. It's not a problem until it's a problem. Come back when you have actual issues. You might be an expert by then. > And thinking of children, is there any quick options to "lock" a Linux > PC? I know Gnome has a task bar thing that you can click to do that. I'm sure there are other options. Google is your friend if no one else chimes in. On video cards, Intel drivers are more open, but ATI is getting there. AMD bought them and are releasing documentation... slowly. Nvidia drivers and ATI's closed drivers are good and getting better too. Personally, I'm a fan of Fedora, but Ubuntu has a larger support community. If you want latest and greatest, Fedora and Ubuntu are good choices. If you want stable and secure, CentOS and Ubuntu LTS are better. From mbmiller at taxa.epi.umn.edu Tue Feb 5 12:16:43 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 12:16:43 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <47A89920.20104@gmail.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: On Tue, 5 Feb 2008, p.daniels wrote: > Ubuntu (or any distro that uses the sudo model) makes this very easy. > There is no root account by default, and the "do this as root" password > is the password of the original user (you). When you make new users, > they don't have root access unless you give it to them. I know on Ubuntu > when you make a new user, the menu items that require root access don't > even appear in their menus. The note above is mostly answering a question I was going to ask here. Isn't that system weakening security a little bit by essentially making the root password the same as one of the user passwords? If someone gets the user password, he also gets root permissions and can do what he pleases. Is there really no root account? On our Ubuntu system there is one: $ grep ^root /etc/passwd root:x:0:0:root:/root:/bin/bash Doesn't there have to be a root account if files are to be owned by root? What is the advantage of sudo over su? Does it log activity better? Mike From brian at ropers-huilman.net Tue Feb 5 12:35:09 2008 From: brian at ropers-huilman.net (Brian D. Ropers-Huilman) Date: Tue, 5 Feb 2008 12:35:09 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: On Feb 5, 2008 12:31 PM, Brian D. Ropers-Huilman wrote: > On a Macintosh or in the *buntu model, the first user created > typically has "full" sudo rights and can do anything on the machine. > This is _still_ a better security model than allowing root to login to > the box (locally or remotely) and having a root password set. I accidentally sent that before completing it. By limiting root access to sudo commands, you force the user into a mode of consciously making the decision to take administrative actions. Back in the day, it was not at all uncommon for an administrator (or user with such priveleges) to login as root and operate that way on the machine, all day long. This is a huge security exposure. Not having a root password, preventing remote and local root logins, and explicitly controlling access to root-level administrative commands is definitely a better way to fly. -- Brian D. Ropers-Huilman, Director Systems Administration and Technical Operations Minnesota Supercomputing Institute 599 Walter Library +1 612-626-5948 (V) 117 Pleasant Street S.E. +1 612-624-8861 (F) University of Minnesota Twin Cities Campus Minneapolis, MN 55455-0255 http://www.msi.umn.edu/ From brockn at gmail.com Tue Feb 5 12:36:46 2008 From: brockn at gmail.com (Brock Noland) Date: Tue, 5 Feb 2008 12:36:46 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: <741dcbb80802051036o400f27d6le09f981590cbc3df@mail.gmail.com> With sudo, you do everything that does not require root privileges as a normal user. Then only when you need root privileges do you execute a command as root. Except for the desktop distros, you set a separate root password. Thus users, even with root privileges via sudo, don't know the root password. sudo asks for the users password, not the root password. Unless setup as below, nothing is stopping you from running sudo su - root You can allow specific commands to be run as root. So user x (or group) can only run kill, service, apt-get, etc as root. Every sudo command is logged. On Feb 5, 2008 12:16 PM, Mike Miller wrote: > On Tue, 5 Feb 2008, p.daniels wrote: > > > Ubuntu (or any distro that uses the sudo model) makes this very easy. > > There is no root account by default, and the "do this as root" password > > is the password of the original user (you). When you make new users, > > they don't have root access unless you give it to them. I know on Ubuntu > > when you make a new user, the menu items that require root access don't > > even appear in their menus. > > The note above is mostly answering a question I was going to ask here. > Isn't that system weakening security a little bit by essentially making > the root password the same as one of the user passwords? If someone gets > the user password, he also gets root permissions and can do what he > pleases. > > Is there really no root account? On our Ubuntu system there is one: > > $ grep ^root /etc/passwd > > root:x:0:0:root:/root:/bin/bash > > Doesn't there have to be a root account if files are to be owned by root? > > What is the advantage of sudo over su? Does it log activity better? > > Mike > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > From erikerik at gmail.com Tue Feb 5 12:37:23 2008 From: erikerik at gmail.com (Erik Anderson) Date: Tue, 5 Feb 2008 12:37:23 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: On Feb 5, 2008 12:16 PM, Mike Miller wrote: > > Is there really no root account? On our Ubuntu system there is one: Yes, there is a root account. It just doesn't have a password by default. It's quite easy to set the root password if you want to - just run: $ sudo passwd root > What is the advantage of sudo over su? Does it log activity better? Yes, sudo logs all commands that are run through it. su doesn't. -erik From florin at iucha.net Tue Feb 5 12:50:31 2008 From: florin at iucha.net (Florin Iucha) Date: Tue, 5 Feb 2008 12:50:31 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: <20080205185031.GZ15915@iris.iucha.org> On Tue, Feb 05, 2008 at 12:37:23PM -0600, Erik Anderson wrote: > On Feb 5, 2008 12:16 PM, Mike Miller wrote: > > > > Is there really no root account? On our Ubuntu system there is one: > > Yes, there is a root account. It just doesn't have a password by > default. That would be terrible. In fact, it does have a password, but it is a random string. Cheers, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/5ae65852/attachment.pgp From ecrist at secure-computing.net Tue Feb 5 12:53:24 2008 From: ecrist at secure-computing.net (Eric F Crist) Date: Tue, 5 Feb 2008 12:53:24 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> On Feb 5, 2008, at 12:37 PM, Erik Anderson wrote: > On Feb 5, 2008 12:16 PM, Mike Miller > wrote: >> >> Is there really no root account? On our Ubuntu system there is one: > > Yes, there is a root account. It just doesn't have a password by > default. It's quite easy to set the root password if you want to - > just run: > > $ sudo passwd root > >> What is the advantage of sudo over su? Does it log activity better? > > Yes, sudo logs all commands that are run through it. su doesn't. This is slightly misguided. Even with sudo, you can sudo su and where the su to will be logged, anything done while su'd is not logged. Only commands invoked directly with sudo are logged. In this case, logging is no better than it is with su. ----- Eric F Crist Secure Computing Networks From clay at fandre.com Tue Feb 5 12:47:09 2008 From: clay at fandre.com (Clay Fandre) Date: Tue, 5 Feb 2008 12:47:09 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: Yes, there is a root account but by default the password isn't set. You can sudo to root and set it and use it like normal if you'd like. (but I would advise against that) sudo is said to be more secure for a number of reasons. First of all, you should run all admin commands prefixed with sudo (as opposed to running 'sudo su -' first) Also, it does additional logging which makes it easy to track down 'who dun-it'. It also lets you grant only certain commands to users rather than 'ALL' access. So if you need someone to be able to restart apache from time-to-time, you can give them rights to just restart apache. The problem with sudo is there are many ways to get around it, and it's not very easy to get specific. Luckily the next version of Ubuntu will incorporate PolicyKit which should allow for finer controls over things. Haven't used it yet, but it looks promising. https://wiki.ubuntu.com/HardyHeron/Alpha4 On Feb 5, 2008 12:16 PM, Mike Miller wrote: > On Tue, 5 Feb 2008, p.daniels wrote: > > > Ubuntu (or any distro that uses the sudo model) makes this very easy. > > There is no root account by default, and the "do this as root" password > > is the password of the original user (you). When you make new users, > > they don't have root access unless you give it to them. I know on Ubuntu > > when you make a new user, the menu items that require root access don't > > even appear in their menus. > > The note above is mostly answering a question I was going to ask here. > Isn't that system weakening security a little bit by essentially making > the root password the same as one of the user passwords? If someone gets > the user password, he also gets root permissions and can do what he > pleases. > > Is there really no root account? On our Ubuntu system there is one: > > $ grep ^root /etc/passwd > > root:x:0:0:root:/root:/bin/bash > > Doesn't there have to be a root account if files are to be owned by root? > > What is the advantage of sudo over su? Does it log activity better? > > Mike > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > From clay at fandre.com Tue Feb 5 13:12:04 2008 From: clay at fandre.com (Clay Fandre) Date: Tue, 5 Feb 2008 13:12:04 -0600 Subject: [tclug-list] sudo In-Reply-To: <20080205185031.GZ15915@iris.iucha.org> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> Message-ID: s/doesn't have a password/doesn't have a password that you know/ On Feb 5, 2008 12:50 PM, Florin Iucha wrote: > On Tue, Feb 05, 2008 at 12:37:23PM -0600, Erik Anderson wrote: > > On Feb 5, 2008 12:16 PM, Mike Miller wrote: > > > > > > Is there really no root account? On our Ubuntu system there is one: > > > > Yes, there is a root account. It just doesn't have a password by > > default. > > That would be terrible. In fact, it does have a password, but it is a > random string. > > Cheers, > florin > > -- > Bruce Schneier expects the Spanish Inquisition. > http://geekz.co.uk/schneierfacts/fact/163 > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > From brian at ropers-huilman.net Tue Feb 5 11:48:05 2008 From: brian at ropers-huilman.net (Brian D. Ropers-Huilman) Date: Tue, 5 Feb 2008 11:48:05 -0600 Subject: [tclug-list] New to Linux In-Reply-To: <614465.45276.qm@web50301.mail.re2.yahoo.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> Message-ID: On Feb 5, 2008 10:11 AM, Daniel Burke wrote: > I am new to Linux Welcome! > I am hoping to learn some Linux as well I will second the mention of the various LiveCDs already mentioned. They are an excellent way to both test out your hardware and also experiment with different flavors of Linux so you can find what you like. > The budget for this, ..., is projected to be about $1,200.00 I know you know PCs, but have you considered getting a Dell? If you do, that opens to door to receiving a system pre-installed where you'll know everything works out of the gate. You can still experiment and learn, but you'll start with a known-good system. > What do I look for to make sure the hardware I choose is compatible with > Linux? As has already been mentioned, hardware support in Linux has come a long way (I've used Linux on servers for over 13 years now and on my desktop {exclusively, I'd add) for 10) and things have dramatically changed. Most stuff "just works." If it doesn't Google Is Your Friend and the forums of the various distributions typically offer a quick resolution. > To keep things simple for my wife, I was thinking of going with KDE. Is this > a good choice? Over my long history with Linux, I've used fvwm, fluxbox, GNOME, E17, KDE, xfce, ... For what you're talking about, I happen to think that KDE will be the most "compatible" for you and your spouse. Since you're getting new hardware, there's really no reason to go minimalist (bye bye fvwm, fluxbox, xfce and E17). If you and your spouse have previously used and like Macintoshes, I'd go with GNOME. If you're long time Windows users, go with KDE. With that said, I would also voice support for Kubuntu, the KDE version of Ubuntu. It's a nice, fully functional KDE. > Our children are still very young but it won't be long before they will want > to play on the computer. Any advice on setup that might make my life easier > with the kids accessing the PC over the next decade? I also have two young children, seven and three, and both of them are able to use my home system. The seven year old will go login to his desktop (Switching Users if Mom or Dad have stayed logged in) and bring up Firefox to check his Google Mail. He's also quite adept at playing the plethora of games available (and installed by me) on the system. The three year old is still getting his mouse skills down, but there are enough options for his little mind to engage with as well. > And thinking of children, is there any quick options to "lock" a Linux PC? > In Windows XP, I just have to press the Window+L on the keyboard when our > little two year old boy comes near to prevent him from using my account with > my access to do anything (logging off and shutting down in Windows takes too > long as little boys can be very quick). I very much empathize with this! Yes, there certainly is a way. Exactly how you do it will depend on which desktop environment you choose (KDE, GNOME, etc.). I have my system mapped so does the instant locking. -- Brian D. Ropers-Huilman, , Director Systems Administration and Technical Operations Minnesota Supercomputing Institute 599 Walter Library +1 612-626-5948 (V) 117 Pleasant Street S.E. +1 612-624-8861 (F) University of Minnesota Twin Cities Campus Minneapolis, MN 55455-0255 http://www.msi.umn.edu/ From brian at ropers-huilman.net Tue Feb 5 12:31:09 2008 From: brian at ropers-huilman.net (Brian D. Ropers-Huilman) Date: Tue, 5 Feb 2008 12:31:09 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> Message-ID: On Feb 5, 2008 12:16 PM, Mike Miller wrote: > On Tue, 5 Feb 2008, p.daniels wrote: > > Ubuntu (or any distro that uses the sudo model) makes this very easy. > > There is no root account by default, and the "do this as root" password > > is the password of the original user (you). When you make new users, > > they don't have root access unless you give it to them. I know on Ubuntu > > when you make a new user, the menu items that require root access don't > > even appear in their menus. > > The note above is mostly answering a question I was going to ask here. > Isn't that system weakening security a little bit by essentially making > the root password the same as one of the user passwords? If someone gets > the user password, he also gets root permissions and can do what he > pleases. > > Is there really no root account? On our Ubuntu system there is one: > > $ grep ^root /etc/passwd > > root:x:0:0:root:/root:/bin/bash > > Doesn't there have to be a root account if files are to be owned by root? > > What is the advantage of sudo over su? Does it log activity better? > > Mike There's some misinformation above. Start by looking at a man page for sudo [ http://www.gratisoft.us/sudo/man/sudo.html ] and then maybe hit Wikipedia [ http://en.wikipedia.org/wiki/Sudo ]. Yes, there is a root account. sudo is a better way to facilitate using su, providing a granular, limiting access that is also auditable (logging). Because of this, sudo is typically setup to allow limited administrative operations. As such, a compromised account will still be limited to what the systems administrator allowed that account to do in the first place, which is typically not much. On a Macintosh or in the *buntu model, the first user created typically has "full" sudo rights and can do anything on the machine. This is _still_ a better security model than allowing root to login to the box (locally or remotely) and having a root password set. -- Brian D. Ropers-Huilman, Director Systems Administration and Technical Operations Minnesota Supercomputing Institute 599 Walter Library +1 612-626-5948 (V) 117 Pleasant Street S.E. +1 612-624-8861 (F) University of Minnesota Twin Cities Campus Minneapolis, MN 55455-0255 http://www.msi.umn.edu/ From mbmiller at taxa.epi.umn.edu Tue Feb 5 13:46:15 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 13:46:15 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> Message-ID: On Tue, 5 Feb 2008, Clay Fandre wrote: > s/doesn't have a password/doesn't have a password that you know/ I can believe that because (A) it makes sense [as Florin suggested -- a random string would be optimal] and (B) several people have said it. One respondent suggested that this command will allow the user to change the root password: $ sudo passwd root But usually the first step in running the passwd command is to enter the existing password, which you don't know (doesn't it work that way for root passwords). This may not be an important issue, but it is still possible to change the root password by editing the /etc/shadow file -- you just have to know the encrypted form of your password. This can be obtained as follows: perl -le 'print crypt("password", "salt");' ...where 'password' is the unencrypted password and 'salt' is the salt string (only the first two characters are used) for the crypt command. The first two characters of the output are the salt. If you have root permissions on a UNIX/Linux machine, you can check that this works by reading /etc/shadow, entering your password and the first two characters of your encrypted password as your salt. This is your salt: sudo egrep '^username:' /etc/shadow | gawk -F':' '{print $2}' | cut -c -2 ...where 'username' is your user name. I tried it and it worked perfectly. (It is not a great idea to run the perl command from the command line because it will leave your password in your command history, so it would be safer to write the perl command into a one-line script, execute it, and delete it.) Mike From mbmiller at taxa.epi.umn.edu Tue Feb 5 13:51:42 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 13:51:42 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> Message-ID: Thanks for all of the useful information about sudo. It works pretty much like I hoped. I'll have to spend a little time studying it and then configuring it so that various users can do various tasks. The example of restarting apache was excellent and it got me thinking about how nice it would be to have a group of users with permission to do that kind of thing. You can really improve service uptime that way, but you have to be able to trust the users not to do bad things. Mike From ecrist at secure-computing.net Tue Feb 5 13:55:49 2008 From: ecrist at secure-computing.net (Eric F Crist) Date: Tue, 5 Feb 2008 13:55:49 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> Message-ID: <5AFDA5A8-043D-47EB-B775-0CA5268E57B3@secure-computing.net> On Feb 5, 2008, at 1:46 PM, Mike Miller wrote: > On Tue, 5 Feb 2008, Clay Fandre wrote: > >> s/doesn't have a password/doesn't have a password that you know/ > > I can believe that because (A) it makes sense [as Florin suggested > -- a > random string would be optimal] and (B) several people have said it. > > One respondent suggested that this command will allow the user to > change > the root password: > > $ sudo passwd root > > But usually the first step in running the passwd command is to enter > the > existing password, which you don't know (doesn't it work that way > for root > passwords). Mike, you're wrong here. On every *nix system I've been on, the root user doesn't have to enter the old password in order to change the password for a given user. The command above would work, while it would only ask the user issuing the command for their current password, and the new password (twice) for the given user. Alternatively, you change the second delimited field in /etc/ shadow.passwd to be blank, thus clearing out the password for root. ----- Eric F Crist Secure Computing Networks From mbmiller at taxa.epi.umn.edu Tue Feb 5 14:07:42 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 14:07:42 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <5AFDA5A8-043D-47EB-B775-0CA5268E57B3@secure-computing.net> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> <5AFDA5A8-043D-47EB-B775-0CA5268E57B3@secure-computing.net> Message-ID: On Tue, 5 Feb 2008, Eric F Crist wrote: >> But usually the first step in running the passwd command is to enter >> the existing password, which you don't know (doesn't it work that way >> for root passwords). > > Mike, you're wrong here. Well, I forgot the question mark before the close paren. > On every *nix system I've been on, the root user doesn't have to enter > the old password in order to change the password for a given user. The > command above would work, while it would only ask the user issuing the > command for their current password, and the new password (twice) for the > given user. Yep, you are right. That command was: sudo passwd root Thanks for the tip! Mike From josh at joshwelch.com Tue Feb 5 14:04:29 2008 From: josh at joshwelch.com (Josh Welch) Date: Tue, 05 Feb 2008 20:04:29 +0000 Subject: [tclug-list] sudo In-Reply-To: <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> Message-ID: <20080205200429.m58akd4jcowgc0w4@joshwelch.com> Quoting Eric F Crist : >> >> Yes, sudo logs all commands that are run through it. su doesn't. > > > This is slightly misguided. Even with sudo, you can sudo su > and where the su to will be logged, anything done while su'd is > not logged. Only commands invoked directly with sudo are logged. In > this case, logging is no better than it is with su. > Note that the proper approach here would be to simply disallow doing a sudo to su if you're on a multi-user system where such things matter. One of the nice things about sudo is that you can specify with a fair degree of granularity what users are allowed to issue what commands as the superuser. Josh From crumley at belka.space.umn.edu Tue Feb 5 14:39:48 2008 From: crumley at belka.space.umn.edu (Jim Crumley) Date: Tue, 5 Feb 2008 14:39:48 -0600 Subject: [tclug-list] sudo In-Reply-To: <20080205200429.m58akd4jcowgc0w4@joshwelch.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> Message-ID: <20080205143948.A1204@pchelka.space.umn.edu> On Tue, Feb 05, 2008 at 08:04:29PM +0000, Josh Welch wrote: > Note that the proper approach here would be to simply disallow doing a > sudo to su if you're on a multi-user system where such things matter. > One of the nice things about sudo is that you can specify with a fair > degree of granularity what users are allowed to issue what commands as > the superuser. The problem with the blacklist route of dealing with sudo, is that there are often holes. Many programs allow you to run shell commands (vi, emacs, etc.), so you really need to restrict their usage as well, if you are going to go this route. -- Jim Crumley |Twin Cities Linux Users Group Mailing List (TCLUG) Ruthless Debian Zealot |http://www.mn-linux.org/ Never laugh at live dragons | From ecrist at secure-computing.net Tue Feb 5 14:43:32 2008 From: ecrist at secure-computing.net (Eric F Crist) Date: Tue, 5 Feb 2008 14:43:32 -0600 Subject: [tclug-list] sudo In-Reply-To: <20080205200429.m58akd4jcowgc0w4@joshwelch.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> Message-ID: <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> On Feb 5, 2008, at 2:04 PM, Josh Welch wrote: > Quoting Eric F Crist : > >>> >>> Yes, sudo logs all commands that are run through it. su doesn't. >> >> >> This is slightly misguided. Even with sudo, you can sudo su >> and where the su to will be logged, anything done while su'd >> is >> not logged. Only commands invoked directly with sudo are logged. In >> this case, logging is no better than it is with su. >> > > Note that the proper approach here would be to simply disallow doing a > sudo to su if you're on a multi-user system where such things matter. > One of the nice things about sudo is that you can specify with a fair > degree of granularity what users are allowed to issue what commands as > the superuser. Hardly a work-around as I could execute sudo . It really boils down to a couple of options: 1) You trust your users, give them sudo access. 2) You don't trust your users, don't give them sudo access. 3) You don't trust your users, give them a limited set of commands. * With this, I would recommend a 'take it all away' and give them what they need approach. HTH ----- Eric F Crist Secure Computing Networks From josh at joshwelch.com Tue Feb 5 15:01:40 2008 From: josh at joshwelch.com (Josh Welch) Date: Tue, 05 Feb 2008 21:01:40 +0000 Subject: [tclug-list] sudo In-Reply-To: <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> Message-ID: <20080205210140.8cokmg7ascc8ggk8@joshwelch.com> Quoting Eric F Crist : > On Feb 5, 2008, at 2:04 PM, Josh Welch wrote: > >> Quoting Eric F Crist : >> >>>> >>>> Yes, sudo logs all commands that are run through it. su doesn't. >>> >>> >>> This is slightly misguided. Even with sudo, you can sudo su >>> and where the su to will be logged, anything done while su'd is >>> not logged. Only commands invoked directly with sudo are logged. In >>> this case, logging is no better than it is with su. >>> >> >> Note that the proper approach here would be to simply disallow doing a >> sudo to su if you're on a multi-user system where such things matter. >> One of the nice things about sudo is that you can specify with a fair >> degree of granularity what users are allowed to issue what commands as >> the superuser. > > > Hardly a work-around as I could execute sudo . Ummm, what makes you think I gave you the access to `sudo bash` if I didn't give the access to `sudo su`? ;) > It really boils down to a couple of options: > > 1) You trust your users, give them sudo access. > 2) You don't trust your users, don't give them sudo access. > 3) You don't trust your users, give them a limited set of commands. > * With this, I would recommend a 'take it all away' and give them what > they need approach. This is the way I've always given sudo access, users get only the commands they need. Of course I've never met a user I trust :) Josh From josh at joshwelch.com Tue Feb 5 15:06:51 2008 From: josh at joshwelch.com (Josh Welch) Date: Tue, 05 Feb 2008 21:06:51 +0000 Subject: [tclug-list] sudo In-Reply-To: <20080205143948.A1204@pchelka.space.umn.edu> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> <20080205143948.A1204@pchelka.space.umn.edu> Message-ID: <20080205210651.qnroglk2swocokoc@joshwelch.com> Quoting Jim Crumley : > On Tue, Feb 05, 2008 at 08:04:29PM +0000, Josh Welch wrote: >> Note that the proper approach here would be to simply disallow doing a >> sudo to su if you're on a multi-user system where such things matter. >> One of the nice things about sudo is that you can specify with a fair >> degree of granularity what users are allowed to issue what commands as >> the superuser. > > The problem with the blacklist route of dealing with sudo, is > that there are often holes. Many programs allow you to run shell > commands (vi, emacs, etc.), so you really need to restrict their > usage as well, if you are going to go this route. > I misspoke. As I noted to someone else, possibly in private mail, he proper way to give sudo access is to give only specific access, which would in effect disallow `sudo su` as well as everything else not explicitly allowed. Can't get anything by you people. ;) Josh W From ecrist at secure-computing.net Tue Feb 5 15:51:33 2008 From: ecrist at secure-computing.net (Eric F Crist) Date: Tue, 5 Feb 2008 15:51:33 -0600 Subject: [tclug-list] sudo In-Reply-To: <20080205210140.8cokmg7ascc8ggk8@joshwelch.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> <20080205210140.8cokmg7ascc8ggk8@joshwelch.com> Message-ID: <3C63CDD5-0364-4525-8392-D062168EA1F3@secure-computing.net> On Feb 5, 2008, at 3:01 PM, Josh Welch wrote: > Quoting Eric F Crist : > >> On Feb 5, 2008, at 2:04 PM, Josh Welch wrote: >>> >>> Note that the proper approach here would be to simply disallow >>> doing a >>> sudo to su if you're on a multi-user system where such things >>> matter. >>> One of the nice things about sudo is that you can specify with a >>> fair >>> degree of granularity what users are allowed to issue what >>> commands as >>> the superuser. >> >> >> Hardly a work-around as I could execute sudo . > > Ummm, what makes you think I gave you the access to `sudo bash` if I > didn't give the access to `sudo su`? ;) You said you would disallow doing a sudo to su. You said nothing about disallowing other commands. My point is that there are other ways to obtain a root shell without going the su route. As someone else mentioned, vim, emacs, poorly written shell scripts dumped into $PATH, etc. The more secure, or safer, method may be to white-list rather than black-list. At least, that's been my experience. ----- Eric F Crist Secure Computing Networks From mbmiller at taxa.epi.umn.edu Tue Feb 5 16:08:15 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 16:08:15 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <3C63CDD5-0364-4525-8392-D062168EA1F3@secure-computing.net> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> <20080205210140.8cokmg7ascc8ggk8@joshwelch.com> <3C63CDD5-0364-4525-8392-D062168EA1F3@secure-computing.net> Message-ID: On Tue, 5 Feb 2008, Eric F Crist wrote: > You said you would disallow doing a sudo to su. You said nothing about > disallowing other commands. My point is that there are other ways to > obtain a root shell without going the su route. As someone else > mentioned, vim, emacs, poorly written shell scripts dumped into $PATH, > etc. The more secure, or safer, method may be to white-list rather than > black-list. At least, that's been my experience. I think you have to use an disallow-all, allow-specifics kind of approach. What about this? sudo cp -p /usr/bash /usr/bash2 sudo bash2 -l If bash had been disallowed, but bash2 hadn't been disallowed, then you're screwed. I think you have to allow a very limited set of commands, very cautiously. It's probably best to create a special user, not called root, in a special group who has certain special permissions -- can't that be done? Mike From dave at sherohman.org Tue Feb 5 16:31:21 2008 From: dave at sherohman.org (Dave Sherohman) Date: Tue, 5 Feb 2008 16:31:21 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> Message-ID: <20080205223121.GD3431@sherohman.org> On Tue, Feb 05, 2008 at 01:46:15PM -0600, Mike Miller wrote: > On Tue, 5 Feb 2008, Clay Fandre wrote: > > s/doesn't have a password/doesn't have a password that you know/ > > I can believe that because (A) it makes sense [as Florin suggested -- a > random string would be optimal] and (B) several people have said it. Well, actually, no... A random string would *not* be optimal, as there would still be the (admittedly very small) chance that a brute-force attack could guess it. Having no password (which is distinct from having an empty password) is indeed the optimal solution if you wish to disable login on an account, root or otherwise. On some systems, this can be accomplished by emptying the user's password hash value in /etc/shadow (or /etc/passwd, but why would you not be using shadow if you care at all about security?), but other systems will treat that as an empty password, so bad solution. The proper way to ensure that an account has no valid password is to insert characters into the user's hashed password which ensure that the hashing algorithm can never produce a matching hash. Altering the length of the hash can accomplish this for fixed-length hashes (such as MD5) or a character which isn't a part of the set used to encode the hash's output can work for others (e.g., inserting a G into a hexadecimal hash). If you take a look at your /etc/shadow, you'll most likely see several accounts with a password hash of simply "*" or "!", which fits both methods - no respectable cryptographic hash will produce a single-character result and few will use the characters ! or * in their output. But, then, I'm sure Clay and Florin already know this and it's just the ambiguity of the phrase "doesn't have a password" causing confusion... (Does it mean you can access the account without entering a password, that the password is an empty string (nothing), or that there is the absence of any valid password (known or unknown)?) -- News aggregation meets world domination. Can you see the fnews? http://seethefnews.com/ From johntrammell at gmail.com Tue Feb 5 16:38:09 2008 From: johntrammell at gmail.com (John J. Trammell) Date: Tue, 5 Feb 2008 16:38:09 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <7566207B-0441-4872-A974-35683383F8B8@secure-computing.net> <20080205200429.m58akd4jcowgc0w4@joshwelch.com> <7B8F527A-FD55-4A71-B209-4744707EA4AE@secure-computing.net> <20080205210140.8cokmg7ascc8ggk8@joshwelch.com> <3C63CDD5-0364-4525-8392-D062168EA1F3@secure-computing.net> Message-ID: <68dbb6fe0802051438o20f5e93bs3a48b1d14aafe164@mail.gmail.com> Obligatory XKCD reference: http://xkcd.com/149/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/5bea651e/attachment.htm From mbmiller at taxa.epi.umn.edu Tue Feb 5 17:11:47 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Tue, 5 Feb 2008 17:11:47 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <20080205223121.GD3431@sherohman.org> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> <20080205223121.GD3431@sherohman.org> Message-ID: Thanks, Dave. Very interesting. How about: A random string is the hardest password to guess. In that sense a random string is optimal. But use of * or ! in the shadow file does not provide a password so that is optimal when you don't need a password. I had seen those asterisks and double-bangs !! in the shadow file for years and never knew what that was about. Before that they used to be in /etc/password. Remember that? There wasn't always a shadow file and the encrypted passwords used to be readable by all! Mike From pcrequest at gmail.com Tue Feb 5 18:06:56 2008 From: pcrequest at gmail.com (Aaron Lewis) Date: Tue, 5 Feb 2008 18:06:56 -0600 Subject: [tclug-list] Linux+ certification Message-ID: <8cb4302d0802051606l7d4e4f4dr1122898d1b24506c@mail.gmail.com> I'm wondering if there's anyone interested in taking a Linux+ certification class. I've registered in the past with www.minneapolis.edu however they've never had enough persons register and the course was cancelled. If I can find 2 or 3 more people for evening classes sometime in the last half of 2008 that will be enough. Or perhaps someone knows where a class is offered already... Or if you know of a good self study book. Thanks. From tonyyarusso at gmail.com Tue Feb 5 22:55:16 2008 From: tonyyarusso at gmail.com (Tony Yarusso) Date: Tue, 5 Feb 2008 22:55:16 -0600 Subject: [tclug-list] Linux+ certification In-Reply-To: <8cb4302d0802051606l7d4e4f4dr1122898d1b24506c@mail.gmail.com> References: <8cb4302d0802051606l7d4e4f4dr1122898d1b24506c@mail.gmail.com> Message-ID: <254fef0f0802052055j51f58317t71b9a5ae470d06d6@mail.gmail.com> On Feb 5, 2008 6:06 PM, Aaron Lewis wrote: > I'm wondering if there's anyone interested in taking a Linux+ > certification class. I've registered in the past with > www.minneapolis.edu however they've never had enough persons register > and the course was cancelled. If I can find 2 or 3 more people for > evening classes sometime in the last half of 2008 that will be enough. > Or perhaps someone knows where a class is offered already... Or if > you know of a good self study book. Thanks. > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > Saint Paul College (http://www.saintpaul.edu/) offers a number of classes geared toward certifications, include Linux ones. -- Tony Yarusso http://tonyyarusso.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080205/0b31b370/attachment.htm From tonyyarusso at gmail.com Wed Feb 6 00:26:58 2008 From: tonyyarusso at gmail.com (Tony Yarusso) Date: Wed, 6 Feb 2008 00:26:58 -0600 Subject: [tclug-list] Open standards for government documents Message-ID: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> As noted on my blog, I attended my caucus tonight, and decided to present a resolution for mandating the use of ISO-approved open standards for all new government documents and all being newly converted to electronic form, mentioning ODF as the current best option but leaving room to consider others as appropriate in the future. My precinct easily passed the resolution, although for many of them it was the first they had heard of the issue. I was wondering if anyone else had similar resolutions brought up in their precinct, and if so, what was the result? The only concern raised against mine was wondering what the potential cost would be, although I think we have a solid argument there in that it would cost essentially nothing to implement open formats in a forward-only manner, and the real cost only comes in with retroactively converting existing documents (and the prioritization of my resolution was worded accordingly), and regardless would be far less than the cost of trying to recover data after a vendor went bankrupt, changed its terms, or any other similar drastic blocking event to current documents. Given that open document standards have been proposed twice already in the Minnesota Legislature (but been ignored as a low priority), and enjoyed increasing success in being passed into law in other states and countries, I think we have an opportunity to make a significant push for this in Minnesota in the coming year. Additionally, the benefits of open formats provide a great starting point for selling the benefits of open source software in government usage as well (state agencies, legislators, ***public schools!***, etc.), which is another thing that I would very much like to see us get involved in in the coming months and years. So, I wanted to take the opportunity of this local caucus night to see where other people around the state were at with respect to standardized, free, open, non-proprietary formats in government office documents, and start getting people talking about how we might go about making a significant publicity push on the issue with the common population as well as gathering support from local and state representatives. Let the ideas flow! -- Tony Yarusso http://tonyyarusso.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/d4ef319a/attachment.htm From pcrequest at gmail.com Wed Feb 6 00:43:40 2008 From: pcrequest at gmail.com (Aaron Lewis) Date: Wed, 6 Feb 2008 00:43:40 -0600 Subject: [tclug-list] Linux+ certification In-Reply-To: <000001c86867$e9548610$040a0a0a@W4NM4N02> References: <8cb4302d0802051606l7d4e4f4dr1122898d1b24506c@mail.gmail.com> <000001c86867$e9548610$040a0a0a@W4NM4N02> Message-ID: <8cb4302d0802052243u208a80benec0eb4c1a2b133b9@mail.gmail.com> I'm glad to see some interest. General info: http://www.minneapolis.edu/continuinged/certificatesandprograms/informationtechnology/comptia.cfm Use the contact link to express your interest. This is key! I've been told they need at least a few registered to get a class going. Cost: Linux+ is not on the schedule, but have a look at the other courses to get a general idea: https://webproc.mnscu.edu/registration/search/results.html;jsessionid=ED6E4A9310F38F053D606FD8ACF5E991?campusid=305&searchcampusid=305&yrtr=20085&category=IT&subcategory=CERT&openonly=false See if your employer can help with the cost, or if you can get a lifetime learner tax deduction or credit http://www.irs.gov/publications/p970/ or perhaps it can qualify as a business expense. Unemployment benefits will sometimes help you retrain in a new field too. Location: My IT classes have been at either Minneapolis or Egan MN. Ask the school if they could consider a remote option if you're not in the vicinity. On Feb 5, 2008 8:28 PM, Andrew von Nagy wrote: > How much does the course cost? I may be interested. From brian at ropers-huilman.net Wed Feb 6 06:52:58 2008 From: brian at ropers-huilman.net (Brian D. Ropers-Huilman) Date: Wed, 6 Feb 2008 06:52:58 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> Message-ID: On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: > As noted on my blog, I attended my caucus tonight, and decided to present a > resolution for mandating the use of ISO-approved open standards for all new > government documents and all being newly converted to electronic form, > > I was wondering if anyone else had similar resolutions brought up in their > precinct, and if so, what was the result? The only concern raised against > mine was wondering what the potential cost would be, although I think we > have a solid argument there in that it would cost essentially nothing to > implement open formats in a forward-only manner, and the real cost only > comes in with retroactively converting existing documents I am a strong proponent of open source and have even been accused of evangelicalizing at times. I pushed my extended family to move to OpenOffice years ago, as I did. Having said that, however, we need to be careful when we talk about costs. There certainly will be a cost involved in such a conversion, not in the cost of the format or of the software to produce documents in that format, but in terms of training. Like it or not, there are differences in the applications that support open formats and those that don't and those differences will have to be trained before users are fully accepting of the new applications and their new formats. I just needed to point that one out. We did not stay for our actual caucus last night as our kids would have melted. I'm not aware of any discussion in that regard either, but I applaud you for bringing it up and am excited that your resolution passed. -- Brian D. Ropers-Huilman From troythetechguy at gmail.com Wed Feb 6 07:46:51 2008 From: troythetechguy at gmail.com (Troy) Date: Wed, 6 Feb 2008 07:46:51 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> Message-ID: <34de7f3d0802060546p6dfb598fwf125281092955712@mail.gmail.com> Great idea Tony. I wish I would have thought of offering the open standard as a resolution as my caucus too. As Brian points out, there is the cost of training involved with any conversion. However, I think this touches on the fundamental issue of how we teach people to use software. This topic was mentioned in another list that I am a member, and the consensus was we teach kids to use MS Word, not word processing software. It would be nice if schools at least touched on alternatives to MS products. Troy On Feb 6, 2008 6:52 AM, Brian D. Ropers-Huilman wrote: > On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: > > As noted on my blog, I attended my caucus tonight, and decided to > present a > > resolution for mandating the use of ISO-approved open standards for all > new > > government documents and all being newly converted to electronic form, > > > > I was wondering if anyone else had similar resolutions brought up in > their > > precinct, and if so, what was the result? The only concern raised > against > > mine was wondering what the potential cost would be, although I think we > > have a solid argument there in that it would cost essentially nothing to > > implement open formats in a forward-only manner, and the real cost only > > comes in with retroactively converting existing documents > > I am a strong proponent of open source and have even been accused of > evangelicalizing at times. I pushed my extended family to move to > OpenOffice years ago, as I did. > > Having said that, however, we need to be careful when we talk about > costs. There certainly will be a cost involved in such a conversion, > not in the cost of the format or of the software to produce documents > in that format, but in terms of training. Like it or not, there are > differences in the applications that support open formats and those > that don't and those differences will have to be trained before users > are fully accepting of the new applications and their new formats. > > I just needed to point that one out. > > We did not stay for our actual caucus last night as our kids would > have melted. I'm not aware of any discussion in that regard either, > but I applaud you for bringing it up and am excited that your > resolution passed. > > -- > Brian D. Ropers-Huilman > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > -- Website of the week: http://www.ubuntu.com/ The Free Alternative to M$ Office: http://www.openoffice.org My Blog: http://troythetechguy.blogspot.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/de445f54/attachment.htm From mbmiller at taxa.epi.umn.edu Wed Feb 6 08:46:15 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 6 Feb 2008 08:46:15 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <7b7c42a30802060542h39bdb6ebifd6793b5e29edafd@mail.gmail.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <47A89920.20104@gmail.com> <20080205185031.GZ15915@iris.iucha.org> <20080205223121.GD3431@sherohman.org> <7b7c42a30802060542h39bdb6ebifd6793b5e29edafd@mail.gmail.com> Message-ID: On Wed, 6 Feb 2008, Andy Schmid wrote: > On Feb 5, 2008 5:11 PM, Mike Miller wrote: > >> Thanks, Dave. Very interesting. How about: A random string is the >> hardest password to guess. > > > I disagree. There is the chance (albeit very slim to none) that a > random string can produce a password such as '1234', which can be easily > cracked. I thought about that too, but the thing is, if the wouldbe cracker knows that it is a random string (and he would know if that was the design of the system), there will be no benefit to his guessing first things like "1234," but if he knows that you have disallowed things like "1234", then you have helped him by cutting back on the number of things he must guess. So when using random strings you would *not* want to have rules like "must include both upper case lower case letters, digits and non-alphanumeric characters," because that rule would help a brute-force attacker. Mike From andyschmid at gmail.com Wed Feb 6 08:58:20 2008 From: andyschmid at gmail.com (Andy Schmid) Date: Wed, 6 Feb 2008 08:58:20 -0600 Subject: [tclug-list] sudo In-Reply-To: References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <20080205185031.GZ15915@iris.iucha.org> <20080205223121.GD3431@sherohman.org> <7b7c42a30802060542h39bdb6ebifd6793b5e29edafd@mail.gmail.com> Message-ID: <7b7c42a30802060658k19c7d285ya357f7ffa4b5dd4b@mail.gmail.com> On Feb 6, 2008 8:46 AM, Mike Miller wrote: > On Wed, 6 Feb 2008, Andy Schmid wrote: > > > On Feb 5, 2008 5:11 PM, Mike Miller wrote: > > > >> Thanks, Dave. Very interesting. How about: A random string is the > >> hardest password to guess. > > > > > > I disagree. There is the chance (albeit very slim to none) that a > > random string can produce a password such as '1234', which can be easily > > cracked. > > I thought about that too, but the thing is, if the wouldbe cracker knows > that it is a random string (and he would know if that was the design of > the system), there will be no benefit to his guessing first things like > "1234," but if he knows that you have disallowed things like "1234", then > you have helped him by cutting back on the number of things he must guess. > > So when using random strings you would *not* want to have rules like "must > include both upper case lower case letters, digits and non-alphanumeric > characters," because that rule would help a brute-force attacker. > > Mike > This is a good point. But most brute force attacks are done using common passwords across many hosts (typically from worms). If you have constraints put in place that are wide enough, the number of password permutations is still astronomical, with the chance of weak passwords being produced eliminated. Though, its a good idea all around to disable login access for the root account, as well as any other accounts you do not want logging in. Andy -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/9b7d48b7/attachment.htm From mbmiller at taxa.epi.umn.edu Wed Feb 6 10:32:50 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 6 Feb 2008 10:32:50 -0600 (CST) Subject: [tclug-list] sudo In-Reply-To: <7b7c42a30802060658k19c7d285ya357f7ffa4b5dd4b@mail.gmail.com> References: <614465.45276.qm@web50301.mail.re2.yahoo.com> <20080205185031.GZ15915@iris.iucha.org> <20080205223121.GD3431@sherohman.org> <7b7c42a30802060542h39bdb6ebifd6793b5e29edafd@mail.gmail.com> <7b7c42a30802060658k19c7d285ya357f7ffa4b5dd4b@mail.gmail.com> Message-ID: On Wed, 6 Feb 2008, Andy Schmid wrote: > Though, its a good idea all around to disable login access for the root > account, as well as any other accounts you do not want logging in. Yep. I understand that concept. I like to become root sometimes, but I like the idea of sudo logging, so maybe I'd try to do things that way. I'm still mostly on Solaris (sparc) right now but it is time for a change! Mike From strayf at freeshell.org Wed Feb 6 10:35:42 2008 From: strayf at freeshell.org (Steve Cayford) Date: Wed, 06 Feb 2008 10:35:42 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> Message-ID: <47A9E1DE.2020302@freeshell.org> Tony Yarusso wrote: > [...] > I was wondering if anyone else had similar resolutions brought up in > their precinct, and if so, what was the result? The only concern raised > against mine was wondering what the potential cost would be, although I > think we have a solid argument there in that it would cost essentially > nothing to implement open formats in a forward-only manner, and the real > cost only comes in with retroactively converting existing documents (and > the prioritization of my resolution was worded accordingly), and > regardless would be far less than the cost of trying to recover data > after a vendor went bankrupt, changed its terms, or any other similar > drastic blocking event to current documents. > [...] Our precinct did not have time to go through all the resolutions and they had to be passed along to the senate district conference en masse. So I don't know what all the resolutions contained. I should have included a resolution like that myself, but didn't think of it. As far as the cost... won't the older documents have to be converted to something else at some point anyway? It seems like converting to an open format could actually be cheaper than converting to Microsoft's next document format. -Steve From troy.johnson at health.state.mn.us Wed Feb 6 10:55:22 2008 From: troy.johnson at health.state.mn.us (Troy.A Johnson) Date: Wed, 06 Feb 2008 10:55:22 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> Message-ID: <47A99218.9048.009E.0@health.state.mn.us> >>> On 2/6/2008 at 6:52 AM, in message , "Brian D. Ropers-Huilman" wrote: > On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: >> As noted on my blog, I attended my caucus tonight, and decided to present a >> resolution for mandating the use of ISO-approved open standards for all new >> government documents and all being newly converted to electronic form, >> I was wondering if anyone else had similar resolutions brought up in their >> precinct, and if so, what was the result? The only concern raised against >> mine was wondering what the potential cost would be, although I think we >> have a solid argument there in that it would cost essentially nothing to >> implement open formats in a forward-only manner, and the real cost only >> comes in with retroactively converting existing documents ... > Having said that, however, we need to be careful when we talk about > costs. There certainly will be a cost involved in such a conversion, > not in the cost of the format or of the software to produce documents > in that format, but in terms of training. Like it or not, there are > differences in the applications that support open formats and those > that don't and those differences will have to be trained before users > are fully accepting of the new applications and their new formats. Keep in mind, however, that though some might not need training for new versions of Microsoft Office, many do. And that training costs are not limited to initial product roll out. There are many factors to be considered: - Initial and ongoing end user training - Upgrades and patching - Buying, managing, and tracking licenses - Researching, interpreting, and compensating for changing licensing terms - Negotiating pricing at time of purchase All of these take time and should be weighed on the cost side, and I think there may be more. There are some of the same costs for OpenOffice or similar products, but the purchasing and licensing issues go out the window. It gets even more expensive if you want to convert documents, for whatever reason. Unannounced, undercover document conversion projects occur with almost every new version of Microsoft Office. They show up in higher numbers of Microsoft Office related trouble tickets, if nowhere else. From brian at ropers-huilman.net Wed Feb 6 12:27:08 2008 From: brian at ropers-huilman.net (Brian D. Ropers-Huilman) Date: Wed, 6 Feb 2008 12:27:08 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: <34de7f3d0802060546p6dfb598fwf125281092955712@mail.gmail.com> References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> <34de7f3d0802060546p6dfb598fwf125281092955712@mail.gmail.com> Message-ID: On Feb 6, 2008 7:46 AM, Troy wrote: > As Brian points out, there is the cost of training involved with any > conversion. However, I think this touches on the fundamental issue of how > we teach people to use software. This topic was mentioned in another list > that I am a member, and the consensus was we teach kids to use MS Word, not > word processing software. It would be nice if schools at least touched on > alternatives to MS products. Very much agreed. I've had issues with this for many years. Quite frankly, I am of the belief that a "computer literate" person should be able to walk up to any modern desktop system (Linux, OS X, or Windows [pick your flavor]) and be productive. We need to teach philosophies, not applications. Whether it's Word Perfect, OpenOffice Writer, Abiword, Microsoft Word, or whatever, people should know what's capable. There will always be some specifics to learn, but we need to teach concepts and approaches, capabilities and philosophies, rather than specific applications. -- Brian D. Ropers-Huilman, Director Systems Administration and Technical Operations Minnesota Supercomputing Institute 599 Walter Library +1 612-626-5948 (V) 117 Pleasant Street S.E. +1 612-624-8861 (F) University of Minnesota Twin Cities Campus Minneapolis, MN 55455-0255 http://www.msi.umn.edu/ From mbmiller at taxa.epi.umn.edu Wed Feb 6 14:43:47 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 6 Feb 2008 14:43:47 -0600 (CST) Subject: [tclug-list] Open standards for government documents In-Reply-To: References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> <34de7f3d0802060546p6dfb598fwf125281092955712@mail.gmail.com> Message-ID: On Wed, 6 Feb 2008, Brian D. Ropers-Huilman wrote: > Very much agreed. I've had issues with this for many years. Quite > frankly, I am of the belief that a "computer literate" person should be > able to walk up to any modern desktop system (Linux, OS X, or Windows > [pick your flavor]) and be productive. We need to teach philosophies, > not applications. Whether it's Word Perfect, OpenOffice Writer, Abiword, > Microsoft Word, or whatever, people should know what's capable. There > will always be some specifics to learn, but we need to teach concepts > and approaches, capabilities and philosophies, rather than specific > applications. Good ideas, but when we first begin to teach about software, I think we should teach students how to use something, and that something should be an open-source program. We don't want to be a Microsoft training ground. Regarding the idea that more training is required for Word than for OO-Write: I'm sure that is true *now* but as Open Office becomes more widely used in coming years (because it is taught to every child in school?), the opposite will eventually become true and Word will be the program that requires extra training. I'm ccing this to the Ubuntu list because everyone else was doing that, but I'm not on the Ubuntu list -- do you have any info about that list? Maybe I'll want to sign up. Mike From Dean.Benjamin at mm.com Wed Feb 6 21:07:48 2008 From: Dean.Benjamin at mm.com (Dean.Benjamin at mm.com) Date: Wed, 06 Feb 2008 21:07:48 -0600 Subject: [tclug-list] Open standards for government documents In-Reply-To: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.co m> References: <254fef0f0802052226p6a3eaac6u8dbe001e941f53ac@mail.gmail.com> Message-ID: <6.1.2.0.2.20080206200833.02538598@pop.mm.com> An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/f8bddb5c/attachment.htm From bunjee at charter.net Thu Feb 7 06:43:21 2008 From: bunjee at charter.net (Danny) Date: Thu, 07 Feb 2008 06:43:21 -0600 Subject: [tclug-list] ATI Radeon video card in Ubuntu 7.10 In-Reply-To: References: Message-ID: <1202388201.10652.2.camel@patty> Can anyone out there tell me how to configure a Radeon 1950Pro video card to work with special effects? I would appreciate it. Danny On Wed, 2008-02-06 at 10:57 -0600, tclug-list-request at mn-linux.org wrote: > Send tclug-list mailing list submissions to > tclug-list at mn-linux.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > or, via email, send a message with subject or body 'help' to > tclug-list-request at mn-linux.org > > You can reach the person managing the list at > tclug-list-owner at mn-linux.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of tclug-list digest..." > > > Today's Topics: > > 1. Re: Open standards for government documents > (Brian D. Ropers-Huilman) > 2. Re: Open standards for government documents (Troy) > 3. Re: sudo (Mike Miller) > 4. Re: sudo (Andy Schmid) > 5. Re: sudo (Mike Miller) > 6. Re: Open standards for government documents (Steve Cayford) > 7. Re: Open standards for government documents (Troy.A Johnson) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 6 Feb 2008 06:52:58 -0600 > From: "Brian D. Ropers-Huilman" > Subject: Re: [tclug-list] Open standards for government documents > To: "Tony Yarusso" > Cc: ubuntu-us-mn at lists.ubuntu.com, tclug-list at mn-linux.org > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: > > As noted on my blog, I attended my caucus tonight, and decided to present a > > resolution for mandating the use of ISO-approved open standards for all new > > government documents and all being newly converted to electronic form, > > > > I was wondering if anyone else had similar resolutions brought up in their > > precinct, and if so, what was the result? The only concern raised against > > mine was wondering what the potential cost would be, although I think we > > have a solid argument there in that it would cost essentially nothing to > > implement open formats in a forward-only manner, and the real cost only > > comes in with retroactively converting existing documents > > I am a strong proponent of open source and have even been accused of > evangelicalizing at times. I pushed my extended family to move to > OpenOffice years ago, as I did. > > Having said that, however, we need to be careful when we talk about > costs. There certainly will be a cost involved in such a conversion, > not in the cost of the format or of the software to produce documents > in that format, but in terms of training. Like it or not, there are > differences in the applications that support open formats and those > that don't and those differences will have to be trained before users > are fully accepting of the new applications and their new formats. > > I just needed to point that one out. > > We did not stay for our actual caucus last night as our kids would > have melted. I'm not aware of any discussion in that regard either, > but I applaud you for bringing it up and am excited that your > resolution passed. > > -- > Brian D. Ropers-Huilman > > > > ------------------------------ > > Message: 2 > Date: Wed, 6 Feb 2008 07:46:51 -0600 > From: Troy > Subject: Re: [tclug-list] Open standards for government documents > To: "Brian D. Ropers-Huilman" > Cc: ubuntu-us-mn at lists.ubuntu.com, tclug-list at mn-linux.org > Message-ID: > <34de7f3d0802060546p6dfb598fwf125281092955712 at mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Great idea Tony. I wish I would have thought of offering the open standard > as a resolution as my caucus too. > > As Brian points out, there is the cost of training involved with any > conversion. However, I think this touches on the fundamental issue of how > we teach people to use software. This topic was mentioned in another list > that I am a member, and the consensus was we teach kids to use MS Word, not > word processing software. It would be nice if schools at least touched on > alternatives to MS products. > > Troy > > On Feb 6, 2008 6:52 AM, Brian D. Ropers-Huilman > wrote: > > > On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: > > > As noted on my blog, I attended my caucus tonight, and decided to > > present a > > > resolution for mandating the use of ISO-approved open standards for all > > new > > > government documents and all being newly converted to electronic form, > > > > > > I was wondering if anyone else had similar resolutions brought up in > > their > > > precinct, and if so, what was the result? The only concern raised > > against > > > mine was wondering what the potential cost would be, although I think we > > > have a solid argument there in that it would cost essentially nothing to > > > implement open formats in a forward-only manner, and the real cost only > > > comes in with retroactively converting existing documents > > > > I am a strong proponent of open source and have even been accused of > > evangelicalizing at times. I pushed my extended family to move to > > OpenOffice years ago, as I did. > > > > Having said that, however, we need to be careful when we talk about > > costs. There certainly will be a cost involved in such a conversion, > > not in the cost of the format or of the software to produce documents > > in that format, but in terms of training. Like it or not, there are > > differences in the applications that support open formats and those > > that don't and those differences will have to be trained before users > > are fully accepting of the new applications and their new formats. > > > > I just needed to point that one out. > > > > We did not stay for our actual caucus last night as our kids would > > have melted. I'm not aware of any discussion in that regard either, > > but I applaud you for bringing it up and am excited that your > > resolution passed. > > > > -- > > Brian D. Ropers-Huilman > > > > _______________________________________________ > > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > > tclug-list at mn-linux.org > > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > > > > > -- > Website of the week: > http://www.ubuntu.com/ > > The Free Alternative to M$ Office: > http://www.openoffice.org > > My Blog: > http://troythetechguy.blogspot.com > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/de445f54/attachment-0001.htm > > ------------------------------ > > Message: 3 > Date: Wed, 6 Feb 2008 08:46:15 -0600 (CST) > From: Mike Miller > Subject: Re: [tclug-list] sudo > To: Andy Schmid > Cc: tclug-list at mn-linux.org > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > On Wed, 6 Feb 2008, Andy Schmid wrote: > > > On Feb 5, 2008 5:11 PM, Mike Miller wrote: > > > >> Thanks, Dave. Very interesting. How about: A random string is the > >> hardest password to guess. > > > > > > I disagree. There is the chance (albeit very slim to none) that a > > random string can produce a password such as '1234', which can be easily > > cracked. > > I thought about that too, but the thing is, if the wouldbe cracker knows > that it is a random string (and he would know if that was the design of > the system), there will be no benefit to his guessing first things like > "1234," but if he knows that you have disallowed things like "1234", then > you have helped him by cutting back on the number of things he must guess. > > So when using random strings you would *not* want to have rules like "must > include both upper case lower case letters, digits and non-alphanumeric > characters," because that rule would help a brute-force attacker. > > Mike > > > > ------------------------------ > > Message: 4 > Date: Wed, 6 Feb 2008 08:58:20 -0600 > From: "Andy Schmid" > Subject: Re: [tclug-list] sudo > To: "Mike Miller" > Cc: tclug-list at mn-linux.org > Message-ID: > <7b7c42a30802060658k19c7d285ya357f7ffa4b5dd4b at mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > On Feb 6, 2008 8:46 AM, Mike Miller wrote: > > > On Wed, 6 Feb 2008, Andy Schmid wrote: > > > > > On Feb 5, 2008 5:11 PM, Mike Miller wrote: > > > > > >> Thanks, Dave. Very interesting. How about: A random string is the > > >> hardest password to guess. > > > > > > > > > I disagree. There is the chance (albeit very slim to none) that a > > > random string can produce a password such as '1234', which can be easily > > > cracked. > > > > I thought about that too, but the thing is, if the wouldbe cracker knows > > that it is a random string (and he would know if that was the design of > > the system), there will be no benefit to his guessing first things like > > "1234," but if he knows that you have disallowed things like "1234", then > > you have helped him by cutting back on the number of things he must guess. > > > > So when using random strings you would *not* want to have rules like "must > > include both upper case lower case letters, digits and non-alphanumeric > > characters," because that rule would help a brute-force attacker. > > > > Mike > > > > This is a good point. But most brute force attacks are done using common > passwords across many hosts (typically from worms). If you have constraints > put in place that are wide enough, the number of password permutations is > still astronomical, with the chance of weak passwords being produced > eliminated. Though, its a good idea all around to disable login access for > the root account, as well as any other accounts you do not want logging in. > > Andy > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080206/9b7d48b7/attachment-0001.htm > > ------------------------------ > > Message: 5 > Date: Wed, 6 Feb 2008 10:32:50 -0600 (CST) > From: Mike Miller > Subject: Re: [tclug-list] sudo > To: TCLUG List > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > On Wed, 6 Feb 2008, Andy Schmid wrote: > > > Though, its a good idea all around to disable login access for the root > > account, as well as any other accounts you do not want logging in. > > Yep. I understand that concept. I like to become root sometimes, but I > like the idea of sudo logging, so maybe I'd try to do things that way. > > I'm still mostly on Solaris (sparc) right now but it is time for a change! > > Mike > > > > ------------------------------ > > Message: 6 > Date: Wed, 06 Feb 2008 10:35:42 -0600 > From: Steve Cayford > Subject: Re: [tclug-list] Open standards for government documents > To: TCLUG > Message-ID: <47A9E1DE.2020302 at freeshell.org> > Content-Type: text/plain; charset=UTF-8; format=flowed > > Tony Yarusso wrote: > > [...] > > I was wondering if anyone else had similar resolutions brought up in > > their precinct, and if so, what was the result? The only concern raised > > against mine was wondering what the potential cost would be, although I > > think we have a solid argument there in that it would cost essentially > > nothing to implement open formats in a forward-only manner, and the real > > cost only comes in with retroactively converting existing documents (and > > the prioritization of my resolution was worded accordingly), and > > regardless would be far less than the cost of trying to recover data > > after a vendor went bankrupt, changed its terms, or any other similar > > drastic blocking event to current documents. > > [...] > > Our precinct did not have time to go through all the resolutions and they > had to be passed along to the senate district conference en masse. So I > don't know what all the resolutions contained. I should have included a > resolution like that myself, but didn't think of it. > > As far as the cost... won't the older documents have to be converted to > something else at some point anyway? It seems like converting to an open > format could actually be cheaper than converting to Microsoft's next > document format. > > -Steve > > > > ------------------------------ > > Message: 7 > Date: Wed, 06 Feb 2008 10:55:22 -0600 > From: "Troy.A Johnson" > Subject: Re: [tclug-list] Open standards for government documents > To: "Tony Yarusso" , "Brian D. Ropers-Huilman" > > Cc: ubuntu-us-mn at lists.ubuntu.com, tclug-list at mn-linux.org > Message-ID: <47A99218.9048.009E.0 at health.state.mn.us> > Content-Type: text/plain; charset=US-ASCII > > >>> On 2/6/2008 at 6:52 AM, in message > , "Brian D. > Ropers-Huilman" wrote: > > On Feb 6, 2008 12:26 AM, Tony Yarusso wrote: > >> As noted on my blog, I attended my caucus tonight, and decided to present a > >> resolution for mandating the use of ISO-approved open standards for all new > >> government documents and all being newly converted to electronic form, > >> I was wondering if anyone else had similar resolutions brought up in their > >> precinct, and if so, what was the result? The only concern raised against > >> mine was wondering what the potential cost would be, although I think we > >> have a solid argument there in that it would cost essentially nothing to > >> implement open formats in a forward-only manner, and the real cost only > >> comes in with retroactively converting existing documents > ... > > Having said that, however, we need to be careful when we talk about > > costs. There certainly will be a cost involved in such a conversion, > > not in the cost of the format or of the software to produce documents > > in that format, but in terms of training. Like it or not, there are > > differences in the applications that support open formats and those > > that don't and those differences will have to be trained before users > > are fully accepting of the new applications and their new formats. > > Keep in mind, however, that though some might not need training > for new versions of Microsoft Office, many do. And that training costs > are not limited to initial product roll out. There are many factors to be > considered: > > - Initial and ongoing end user training > - Upgrades and patching > - Buying, managing, and tracking licenses > - Researching, interpreting, and compensating for changing licensing terms > - Negotiating pricing at time of purchase > > All of these take time and should be weighed on the cost side, > and I think there may be more. There are some of the same costs > for OpenOffice or similar products, but the purchasing and licensing > issues go out the window. > > It gets even more expensive if you want to convert documents, > for whatever reason. Unannounced, undercover document > conversion projects occur with almost every new version of > Microsoft Office. They show up in higher numbers of Microsoft > Office related trouble tickets, if nowhere else. > > > > > > > > > ------------------------------ > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > End of tclug-list Digest, Vol 38, Issue 6 > ***************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080207/b81b76c9/attachment-0001.htm From johntrammell at gmail.com Thu Feb 7 08:58:42 2008 From: johntrammell at gmail.com (John J. Trammell) Date: Thu, 7 Feb 2008 08:58:42 -0600 Subject: [tclug-list] ATI Radeon video card in Ubuntu 7.10 In-Reply-To: <1202388201.10652.2.camel@patty> References: <1202388201.10652.2.camel@patty> Message-ID: <68dbb6fe0802070658w2ccc5c64ufd7e5ee91bf6d5b3@mail.gmail.com> On Feb 7, 2008 6:43 AM, Danny wrote: > Can anyone out there tell me how to configure a Radeon 1950Pro video card > to work with special effects? > I would appreciate it. > First you must perform 4 hours of community service for not trimming your post. I mean, the *lights dimmed* when I got that email. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080207/0bc53fc5/attachment.htm From erikerik at gmail.com Thu Feb 7 10:19:53 2008 From: erikerik at gmail.com (Erik Anderson) Date: Thu, 7 Feb 2008 10:19:53 -0600 Subject: [tclug-list] ATI Radeon video card in Ubuntu 7.10 In-Reply-To: <68dbb6fe0802070658w2ccc5c64ufd7e5ee91bf6d5b3@mail.gmail.com> References: <1202388201.10652.2.camel@patty> <68dbb6fe0802070658w2ccc5c64ufd7e5ee91bf6d5b3@mail.gmail.com> Message-ID: On Feb 7, 2008 8:58 AM, John J. Trammell wrote: > > First you must perform 4 hours of community service for not trimming your > post. I mean, the lights dimmed when I got that email. So *that's* what that was! I was wondering... -- Erik Anderson http://andersonfam.org From bdunnette at gmail.com Thu Feb 7 12:25:39 2008 From: bdunnette at gmail.com (Brian Dunnette) Date: Thu, 7 Feb 2008 12:25:39 -0600 Subject: [tclug-list] ATI Radeon video card in Ubuntu 7.10 In-Reply-To: <1202388201.10652.2.camel@patty> References: <1202388201.10652.2.camel@patty> Message-ID: > > Can anyone out there tell me how to configure a Radeon 1950Pro video card > to work with special effects? > I would appreciate it. > Danny > > > Looks like the "Envy" app might be your best bet for finding & installing the appropriate drivers: http://albertomilone.com/nvidia_scripts1.html Again, both Google and the delete key (as John mentioned) are your friends -- let us know if you're still unable to get acceleration for the Radeon! -Brian D. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080207/d24977f5/attachment.htm From c.r.troyer at usfamily.net Thu Feb 7 21:49:00 2008 From: c.r.troyer at usfamily.net (Cyprian Troyer) Date: Thu, 07 Feb 2008 21:49:00 -0600 (CST) Subject: [tclug-list] Slackware 12.0 and wireless networking Message-ID: <47ABD112.6010102@usfamily.net> Anyone have any success yet with a Slackware 12.0 installation talking to a wireless card? I'm trying to get a Trendnet TEW-423PI with a Realtek 8185 chipset working. I downloaded the Linux drivers from Realtek's website, and they seem to be doing what they should. I have tried both the manual method described in Realtek's readme file, and the Slackware method - loading everything through the rc. scripts (took some editing to eliminate all the error messages). Either way I can use iwlist to scan the card, and the results look good - the ap adresses are correct, the ESSID comes back correct, the protocol is identified correctly... But the Quality and Signal level are always reported as zero. I know that the signal strength is better than that, because when I boot the Windoze side of the machine, it reports a signal strength of 85% and I can connect to the router without problem. (It's a Belkin 54 g WIFI router). Any guidance is appreciated. --- http://USFamily.Net/dialup.html - $8.25/mo! -- http://www.usfamily.net/dsl.html - $19.99/mo! --- From tclug at b-o-b.homelinux.com Thu Feb 7 22:34:50 2008 From: tclug at b-o-b.homelinux.com (Robert De Mars) Date: Thu, 07 Feb 2008 22:34:50 -0600 Subject: [tclug-list] Slackware 12.0 and wireless networking In-Reply-To: <47ABD112.6010102@usfamily.net> References: <47ABD112.6010102@usfamily.net> Message-ID: <47ABDBEA.4060802@b-o-b.homelinux.com> Cyprian Troyer wrote: > Anyone have any success yet with a Slackware 12.0 installation talking > to a wireless card? Yes > I'm trying to get a Trendnet TEW-423PI with a > Realtek 8185 chipset working. I downloaded the Linux drivers from > Realtek's website, and they seem to be doing what they should. I have > tried both the manual method described in Realtek's readme file, and the > Slackware method - loading everything through the rc. scripts (took some > editing to eliminate all the error messages). Either way I can use > iwlist to scan the card, and the results look good - the ap adresses > are correct, the ESSID comes back correct, the protocol is identified > correctly... But the Quality and Signal level are always reported as zero. > I know that the signal strength is better than that, because when I > boot the Windoze side of the machine, it reports a signal strength of > 85% and I can connect to the router without problem. (It's a Belkin 54 g > WIFI router). > Any guidance is appreciated. I had a similar problem. I ended up using ndiswrapper with the windows drivers that came with the card. I took notes on how I was able to get it working, but they are at home (where I am not at the moment). I ended up using ndiswrapper with the windows driver that came with my card. The setup was pretty easy as I remember. I then used iwconfig to config the interface. I then added a start up script to rc.local & a shutdown script to rc.local_shutdown. If this email dose not help lead you in the right direction, let me know. I will send you my notes when I get home. Good Luck! Bob De Mars > > > --- http://USFamily.Net/dialup.html - $8.25/mo! -- http://www.usfamily.net/dsl.html - $19.99/mo! --- > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list From chewie at wookimus.net Wed Feb 13 17:15:29 2008 From: chewie at wookimus.net (Chad Walstrom) Date: Wed, 13 Feb 2008 17:15:29 -0600 Subject: [tclug-list] Perl question: Signals and system calls Message-ID: <3641.1202944529@skuld.wookimus.net> O.K. I'm tired of writing BASH scripts. I've decided to jump wholesale into Perl as any good Systems Engineer should, right? I'm running into a bit of a problem: signal handling. Let's say I want to run system application SYSAPP_A once for each element in LIST_A, and if that were to fail, run SYSAPP_B for that element. If I don't trap for signals, and I try to interrupt the Perl script with CTRL-C (SIGINT), it kills either SYSAPP_A or SYSAPP_B and then continues on looping over LIST_A elements. What I want is for the Perl script to die unconditionally. So, I try this: $SIG{INT} = sub { die "Um, I'm outta here!\n"; }; This kind of works, but if I hit CTRL-C during SYSAPP_A's run, it fails and SYSAPP_B tries to execute. SYSAPP_A and SYSAPP_B are each called with the "system" built-in. What do I need to do to make sure the script dies unconditionally when any child also receives an interrupt? Chad From troy.johnson at health.state.mn.us Wed Feb 13 17:56:19 2008 From: troy.johnson at health.state.mn.us (Troy.A Johnson) Date: Wed, 13 Feb 2008 17:56:19 -0600 Subject: [tclug-list] Perl question: Signals and system calls In-Reply-To: <3641.1202944529@skuld.wookimus.net> References: <3641.1202944529@skuld.wookimus.net> Message-ID: <47B32F41.9048.009E.0@health.state.mn.us> >>> On 2/13/2008 at 5:15 PM, in message <3641.1202944529 at skuld.wookimus.net>, Chad Walstrom wrote: > O.K. I'm tired of writing BASH scripts. I've decided to jump wholesale > into Perl as any good Systems Engineer should, right? I'm running into > a bit of a problem: signal handling. > Let's say I want to run system application SYSAPP_A once for each > element in LIST_A, and if that were to fail, run SYSAPP_B for that > element. If I don't trap for signals, and I try to interrupt the Perl > script with CTRL-C (SIGINT), it kills either SYSAPP_A or SYSAPP_B and > then continues on looping over LIST_A elements. > What I want is for the Perl script to die unconditionally. So, I try this: > $SIG{INT} = sub { die "Um, I'm outta here!\n"; }; > This kind of works, but if I hit CTRL-C during SYSAPP_A's run, it fails > and SYSAPP_B tries to execute. SYSAPP_A and SYSAPP_B are each called > with the "system" built-in. > What do I need to do to make sure the script dies unconditionally when > any child also receives an interrupt? Chad, Check the return value of 'system' and 'die' if it is not zero, if that works. If that won't work for you, you could check the output of commands by using backticks ($value = `SYSAPP_A`;) instead of using 'system'. Maybe you can use $SIG{CHLD} for something? Use 'perldoc perlvar', 'perldoc perlipc', 'perldoc -f system', 'perldoc -f die', and especially 'perldoc perldoc' to your advantage. Best of luck, Troy From florin at iucha.net Wed Feb 13 18:02:06 2008 From: florin at iucha.net (Florin Iucha) Date: Wed, 13 Feb 2008 18:02:06 -0600 Subject: [tclug-list] Perl question: Signals and system calls In-Reply-To: <3641.1202944529@skuld.wookimus.net> References: <3641.1202944529@skuld.wookimus.net> Message-ID: <20080214000205.GA3247@iris.iucha.org> On Wed, Feb 13, 2008 at 05:15:29PM -0600, Chad Walstrom wrote: > O.K. I'm tired of writing BASH scripts. I've decided to jump wholesale > into Perl as any good Systems Engineer should, right? I'm running into > a bit of a problem: signal handling. > > Let's say I want to run system application SYSAPP_A once for each > element in LIST_A, and if that were to fail, run SYSAPP_B for that > element. If I don't trap for signals, and I try to interrupt the Perl > script with CTRL-C (SIGINT), it kills either SYSAPP_A or SYSAPP_B and > then continues on looping over LIST_A elements. > > What I want is for the Perl script to die unconditionally. So, I try this: > > $SIG{INT} = sub { die "Um, I'm outta here!\n"; }; > > This kind of works, but if I hit CTRL-C during SYSAPP_A's run, it fails > and SYSAPP_B tries to execute. SYSAPP_A and SYSAPP_B are each called > with the "system" built-in. > > What do I need to do to make sure the script dies unconditionally when > any child also receives an interrupt? I'm a bit rusty on the signals and I don't have the Stevens[1] handy but I think you want trap SIGCHLD in the main process and handle the expiration of the child in an appropriate manner. Here's what 10 secs of Googling brought up: http://www.unix.org.ua/orelly/perl/cookbook/ch16_20.htm Cheers, florin 1: http://www.kohala.com/start/apue.html -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080213/9749238c/attachment.pgp From chewie at wookimus.net Thu Feb 14 10:00:56 2008 From: chewie at wookimus.net (Chad Walstrom) Date: Thu, 14 Feb 2008 10:00:56 -0600 Subject: [tclug-list] Perl question: Signals and system calls In-Reply-To: <20080214000205.GA3247@iris.iucha.org> References: <3641.1202944529@skuld.wookimus.net> <20080214000205.GA3247@iris.iucha.org> Message-ID: <7012.1203004856@skuld.wookimus.net> Thanks for the link, Florin. I notice that it's a Ukranian site where Coprights aren't really honored. ;-) I did have a copy lying around that I hadn't yet cracked open, but just found this gem[1]: The system function (but not backticks) ignores SIGINT and SIGQUIT while child processes are running. That way those signals will kill only the child process. If you want your main program to die as well, check the return value of system, or the value of the $? variable. if (($signo = system(@arglist)) &= 127) { die "program killed by signal $signo\n"; } To get the effect of a system that ignores SIGINT, install your own signal handler and then manually fork and exec: if ($pid = fork) { # parent catches INT and berates user local $SIG{INT} = sub { print "Tsk tsk, no process interruptus\n" }; waitpid($pid, 0); } else { die "cannot fork: $!" unless defined $pid; # child ignores INT and does its thing $SIG{INT} = "IGNORE"; exec("summarize", "/etc/logfiles") or die "Can't exec: $!\n"; } 1. Christiansen, T. and Torkington, N., ``Perl Cookbook'', 1999. O'Reilly & Associates, Inc. http://safari.oreilly.com/1565922433 From admin at lctn.org Thu Feb 14 13:51:29 2008 From: admin at lctn.org (Raymond Norton) Date: Thu, 14 Feb 2008 13:51:29 -0600 Subject: [tclug-list] keyspan usb serial converter Message-ID: <47B49BC1.4030205@lctn.org> Does anyone know if it is still required to rebuild the kernel in Ubuntu 1.7.10, in order to utilize a Keyspan usa-19qw adapter? I can see it using lsusb: Bus 001 Device 005: ID 06cd:0118 Keyspan USA-19QW PDA Bus 001 Device 001: ID 0000:0000 Bus 002 Device 001: ID 0000:0000 From erikerik at gmail.com Thu Feb 14 14:05:07 2008 From: erikerik at gmail.com (Erik Anderson) Date: Thu, 14 Feb 2008 14:05:07 -0600 Subject: [tclug-list] keyspan usb serial converter In-Reply-To: <47B49BC1.4030205@lctn.org> References: <47B49BC1.4030205@lctn.org> Message-ID: On Thu, Feb 14, 2008 at 1:51 PM, Raymond Norton wrote: > Does anyone know if it is still required to rebuild the kernel in Ubuntu > 1.7.10, in order to utilize a Keyspan usa-19qw adapter? I can see it > using lsusb: What does dmesg show after plugging in the device? I'd guess that Ubuntu already has the needed module(s) for this. From admin at lctn.org Thu Feb 14 14:22:38 2008 From: admin at lctn.org (admin at lctn.org) Date: Thu, 14 Feb 2008 14:22:38 -0600 (CST) Subject: [tclug-list] keyspan usb serial converter In-Reply-To: References: <47B49BC1.4030205@lctn.org> Message-ID: <49688.10.10.1.1.1203020558.squirrel@lctn.org> > What does dmesg show after plugging in the device? I'd guess that > Ubuntu already has the needed module(s) for this. > It shows the following: [10271.604000] usb 1-1: new full speed USB device using uhci_hcd and address 6 [10271.752000] usb 1-1: configuration #1 chosen from 1 choice It does not light up though, so I checked the keyspan site, and they had documentation that indicated Ubuntu/Debian did not incorporate the driver in the kernel. I wasn't sure if it was up to date info. From tonyyarusso at gmail.com Thu Feb 14 19:59:19 2008 From: tonyyarusso at gmail.com (Tony Yarusso) Date: Thu, 14 Feb 2008 19:59:19 -0600 Subject: [tclug-list] keyspan usb serial converter In-Reply-To: <49688.10.10.1.1.1203020558.squirrel@lctn.org> References: <47B49BC1.4030205@lctn.org> <49688.10.10.1.1.1203020558.squirrel@lctn.org> Message-ID: <254fef0f0802141759g1059deb5rbdd7635e54c0d626@mail.gmail.com> On Thu, Feb 14, 2008 at 2:22 PM, wrote: > > > What does dmesg show after plugging in the device? I'd guess that > > Ubuntu already has the needed module(s) for this. > > > > > It shows the following: > > [10271.604000] usb 1-1: new full speed USB device using uhci_hcd and > address 6 > [10271.752000] usb 1-1: configuration #1 chosen from 1 choice > > > It does not light up though, so I checked the keyspan site, and they had > documentation that indicated Ubuntu/Debian did not incorporate the driver > in the kernel. I wasn't sure if it was up to date info. > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > I have a Keyspan USA-19HS, and its drivers are included. They were not in 7.04, but I talked to some of the kernel guys and I think after reviewing some licensing stuff they were able to put them in the restricted modules bit, and it is now working in 7.10. -- Tony Yarusso http://tonyyarusso.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080214/25b69b7a/attachment.htm From josh at radkeland.org Fri Feb 15 14:11:27 2008 From: josh at radkeland.org (Joshua Radke) Date: Fri, 15 Feb 2008 14:11:27 -0600 Subject: [tclug-list] USB Bus Resets Message-ID: <47B5F1EF.3050407@radkeland.org> Ok, this one has really gotten me, and google hasn't bailed me out yet, so I was hoping someone on the list could. Some relevant system Info are an Athlon XP 1800+, nForce2 chipset motherboard, RT2561(?) wireless card, 40 GB internal (IDE) disk, and 500GB External Western Digital MyBook USB drive, running MythBuntu 7.10. The kernel is 2.6.22-14-generic #1 SMP Tue Dec 18 08:02:57 UTC 2007 i686 GNU/Linux. Note that I (until recently) had the same setup/problems with a Fedora 7/MythTV setup (on the same computer). Relevant portion of /var/log/messages: Feb 15 12:23:21 venus kernel: [23306.544344] usb 3-6: USB disconnect, address 3 Feb 15 12:23:25 venus kernel: [23310.022118] lost page write due to I/O error on dm-0 Feb 15 12:23:33 venus kernel: [23318.120350] usb 3-6: new high speed USB device using ehci_hcd and address 4 Feb 15 12:23:33 venus kernel: [23318.254619] usb 3-6: configuration #1 chosen from 1 choice Feb 15 12:23:33 venus kernel: [23318.255811] scsi1 : SCSI emulation for USB Mass Storage devices Feb 15 12:23:36 venus kernel: [23321.068823] printk: 475 messages suppressed. Feb 15 12:23:36 venus kernel: [23321.068831] lost page write due to I/O error on dm-0 Feb 15 12:23:36 venus kernel: [23321.068842] lost page write due to I/O error on dm-0 Feb 15 12:23:38 venus kernel: [23323.251000] scsi 1:0:0:0: Direct-Access WD 5000AAK External 1.06 PQ: 0 ANSI: 0 Feb 15 12:23:38 venus kernel: [23323.253219] sd 1:0:0:0: [sdb] 976773168 512-byte hardware sectors (500108 MB) Feb 15 12:23:38 venus kernel: [23323.253842] sd 1:0:0:0: [sdb] Write Protect is off Feb 15 12:23:38 venus kernel: [23323.256121] sd 1:0:0:0: [sdb] 976773168 512-byte hardware sectors (500108 MB) Feb 15 12:23:38 venus kernel: [23323.256836] sd 1:0:0:0: [sdb] Write Protect is off Feb 15 12:23:38 venus kernel: [23323.256849] sdb: sdb1 Feb 15 12:23:38 venus kernel: [23323.258946] sd 1:0:0:0: [sdb] Attached SCSI disk Feb 15 12:23:38 venus kernel: [23323.259003] sd 1:0:0:0: Attached scsi generic sg0 type 0 ... and a few minutes later the wife (at home) patiently reboots the computer, and wonders why I got her hooked on this silly machine. This issue is that the USB bus resets, and when the external disk is on an LVM2 volume, the dm entry is not removed. When the USB disk comes back online, it gets another /dev/sd? entry, and cannot be accessed until we've done some cleanup, as in (as root, of course): /etc/init.d/myth-backend stop umount /storage dmsetup delete vgscan lvscan vgchange -ay mount /storage /etc/init.d/myth-backend start Phew! Now ... I've considered adding this as a script to a /etc/udev/rules.d/10-local.rules file (to match the serial number of the disk, for example), but this is not a solution, only a way to let my wife only notice the issues minimally (though it would have the side effect of chopping up recordings into pieces with the backend restarts).. Note that the previous Fedora setup had the storage partition on a primary partition (not LVM), so the USB drive would (upon reconnect) end up with the same letter, and reading/writing from myth-backend would simply resume after a pause. Some other things I've tried: boot the kernel with acpi=off (no noticable difference) Add the following to /etc/udev/rules.d/10-local.rules SYSFS{serial}=="574341533830343230313137",RUN+="/bin/sh -c '/bin/echo 64 > /sys/block/%k/device/max_sectors'" Some of these were suggested in (https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/61235) Picking through sdparm -al /dev/sda shows '0' for IDLE and STANDBY entries. Note that the problem appeared to be much more pronounced (resetting every 10's of minutes instead of one to three times per day) when I had the PCI bus underclocked in the BIOS ... I have since gotten a better cooling solution, and no longer need to do that for stability. Finally, downgrading the usb to 1.1 (modprobe -r ehci_hcd) appeared to solve the USB resets, but there was not quite enough bandwidth left to both play and record to the disk. (Playback would skip every couple of seconds ... the kids didn't care, but it'd drive my wife and I nuts!). Any hints? Thanks, Josh From jpschewe at mtu.net Sun Feb 17 11:25:53 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Sun, 17 Feb 2008 11:25:53 -0600 Subject: [tclug-list] Wireless problem Message-ID: <47B86E21.9050905@mtu.net> I've got an Intel 3645 A/B/G card in my laptop and it works fine in Windows, however not so good in Linux. I can connect to a network and send packets, but I don't receive any packets. If I watch the output of ifconfig I can see the RX packets dropped count increasing. I have a sniffer on my wireless access point (OpenBSD) and don't see the packets getting through either. Does anyone have any ideas on this? I'm running openSUSE 10.3 and using the stock drivers that ship with it. -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From jack at jacku.com Sun Feb 17 22:39:46 2008 From: jack at jacku.com (Jack Ungerleider) Date: Sun, 17 Feb 2008 22:39:46 -0600 Subject: [tclug-list] Wireless problem In-Reply-To: <47B86E21.9050905@mtu.net> References: <47B86E21.9050905@mtu.net> Message-ID: <200802172239.46831.jack@jacku.com> On Sunday 17 February 2008 11:25 am, Jon Schewe wrote: > I've got an Intel 3645 A/B/G card in my laptop and it works fine in > Windows, however not so good in Linux. I can connect to a network and > send packets, but I don't receive any packets. If I watch the output of > ifconfig I can see the RX packets dropped count increasing. I have a > sniffer on my wireless access point (OpenBSD) and don't see the packets > getting through either. Does anyone have any ideas on this? > > I'm running openSUSE 10.3 and using the stock drivers that ship with it. There's a firmware shim package that might not get load by default. If its not there then the wireless won't connect correctly. -- Jack Ungerleider jack at jacku.com http://www.jacku.com From goeko at Goecke-Dolan.com Mon Feb 18 01:06:45 2008 From: goeko at Goecke-Dolan.com (Brian Dolan-Goecke) Date: Mon, 18 Feb 2008 01:06:45 -0600 Subject: [tclug-list] Audacity on Linux - Topic @PenguinsUnbound Linux Meeting Feb 23, 2008 Message-ID: <47B92E85.3060103@Goecke-Dolan.com> This months PenguinsUnbound.net meeting will be Saturday February 23, 2008 at TIES, 1667 Snelling Ave. N., St. Paul, MN 55108 from 10:00 to 12:00. (See the web site http://www.penguinsunbound.com/Location_for_Meetings for directions and more info.) We will be looking at and working with Audacity. My brief outline for the meeting, Getting Started Installing Audacity Starting Audacity Opening a file New projects Working with file Cutting file up Converting formats Creating a Ring Tone Please come and share your knowledge of Audacity! http://www.penguinsunbound.com/Future_Meetings/20080223_-_Audacity Thanks, hope to see you there. ==>brian. From daniel.armbrust.list at gmail.com Tue Feb 19 09:52:53 2008 From: daniel.armbrust.list at gmail.com (Dan Armbrust) Date: Tue, 19 Feb 2008 09:52:53 -0600 Subject: [tclug-list] How to find the width of a console? Message-ID: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> I have a program that is writing info out to the console - and I need to know where to wrap the lines. How does one go about finding the width of a console on linux? Or (shudder) on windows? I've discovered that on linux, at the command prompt, I can do "echo $COLUMNS", and it gives me the number that I want, but this variable doesn't seem to be exported. It doesn't appear in my env - and consequently, I haven't figured out a way to get at it from a program that I launch on the console. And I haven't yet found anything on windows that would tell me the current console width. Does anyone know offhand? I'm having a lot of difficulty crafting a useful google query for this one. Thanks, Dan From florin at iucha.net Tue Feb 19 11:03:52 2008 From: florin at iucha.net (Florin Iucha) Date: Tue, 19 Feb 2008 11:03:52 -0600 Subject: [tclug-list] How to find the width of a console? In-Reply-To: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> References: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> Message-ID: <20080219170352.GO3247@iris.iucha.org> On Tue, Feb 19, 2008 at 09:52:53AM -0600, Dan Armbrust wrote: > I have a program that is writing info out to the console - and I need > to know where to wrap the lines. > > How does one go about finding the width of a console on linux? Or > (shudder) on windows? //--- cut here --- cut here --- cut here --- cut here --- cut here --- /* * Compile me like this: * "cc -lncurses screen_size.c" */ #include int main(void) { int bx, by, mx, my; initscr(); getbegyx(stdscr, by, bx); getmaxyx(stdscr, my, mx); endwin(); printf("%d rows and %d columns\n", mx - bx, my - by); return 0; } //--- cut here --- cut here --- cut here --- cut here --- cut here --- There is curses for windows, too. Cheers, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080219/025b533b/attachment.pgp From jima at beer.tclug.org Tue Feb 19 11:18:05 2008 From: jima at beer.tclug.org (Jima) Date: Tue, 19 Feb 2008 11:18:05 -0600 (CST) Subject: [tclug-list] How to find the width of a console? In-Reply-To: <20080219170352.GO3247@iris.iucha.org> References: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> <20080219170352.GO3247@iris.iucha.org> Message-ID: On Tue, 19 Feb 2008, Florin Iucha wrote: > There is curses for windows, too. There are lots of curses for Windows. Some might even call Windows a curse, itself. Jima From klieber at gentoo.org Tue Feb 19 10:36:59 2008 From: klieber at gentoo.org (Kurt Lieber) Date: Tue, 19 Feb 2008 10:36:59 -0600 Subject: [tclug-list] How to find the width of a console? In-Reply-To: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> References: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> Message-ID: <82d43d110802190836g59560520xbb7b7432a16c8caa@mail.gmail.com> On 2/19/08, Dan Armbrust wrote: > > How does one go about finding the width of a console on linux? Or > (shudder) on windows? Here's a python snippet that might help for windows, plus some code in the comments that might apply to linux: http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/440694 hth --kurt -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080219/46a556d2/attachment.htm From daniel.armbrust.list at gmail.com Tue Feb 19 13:14:20 2008 From: daniel.armbrust.list at gmail.com (Dan Armbrust) Date: Tue, 19 Feb 2008 13:14:20 -0600 Subject: [tclug-list] How to find the width of a console? In-Reply-To: <20080219170352.GO3247@iris.iucha.org> References: <82f04dc40802190752k17ab4bc0m3e04a9e72d91f035@mail.gmail.com> <20080219170352.GO3247@iris.iucha.org> Message-ID: <82f04dc40802191114s2ab384b3ic55487031ce88ce3@mail.gmail.com> Thanks. I'll try it out. Dan On Feb 19, 2008 11:03 AM, Florin Iucha wrote: > On Tue, Feb 19, 2008 at 09:52:53AM -0600, Dan Armbrust wrote: > > I have a program that is writing info out to the console - and I need > > to know where to wrap the lines. > > > > How does one go about finding the width of a console on linux? Or > > (shudder) on windows? > > //--- cut here --- cut here --- cut here --- cut here --- cut here --- > > /* > * Compile me like this: > * "cc -lncurses screen_size.c" > */ > > #include > > int main(void) > { > int bx, by, mx, my; > > initscr(); > > getbegyx(stdscr, by, bx); > getmaxyx(stdscr, my, mx); > > endwin(); > > printf("%d rows and %d columns\n", mx - bx, my - by); > > return 0; > } > > //--- cut here --- cut here --- cut here --- cut here --- cut here --- > > There is curses for windows, too. > > Cheers, > florin > > -- > Bruce Schneier expects the Spanish Inquisition. > http://geekz.co.uk/schneierfacts/fact/163 > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > From jkey at tomobiki.dyndns.org Tue Feb 19 20:23:35 2008 From: jkey at tomobiki.dyndns.org (jkey) Date: Tue, 19 Feb 2008 20:23:35 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <47B5F1EF.3050407@radkeland.org> References: <47B5F1EF.3050407@radkeland.org> Message-ID: <47BB8F27.2000109@tomobiki.dyndns.org> Joshua Radke wrote: > Ok, this one has really gotten me, and google hasn't bailed me out yet, > so I was hoping someone on the list could. > > Some relevant system Info are an Athlon XP 1800+, nForce2 chipset > motherboard, RT2561(?) wireless card, 40 GB internal (IDE) disk, and > 500GB External Western Digital MyBook USB drive, running MythBuntu > 7.10. The kernel is 2.6.22-14-generic #1 SMP Tue Dec 18 08:02:57 UTC > 2007 i686 GNU/Linux. > > Note that I (until recently) had the same setup/problems with a Fedora > 7/MythTV setup (on the same computer). > > Relevant portion of /var/log/messages: > Feb 15 12:23:21 venus kernel: [23306.544344] usb 3-6: USB disconnect, > address 3 > Feb 15 12:23:25 venus kernel: [23310.022118] lost page write due to I/O > error on dm-0 > > Feb 15 12:23:33 venus kernel: [23318.120350] usb 3-6: new high speed USB > device using ehci_hcd and address 4 > Feb 15 12:23:33 venus kernel: [23318.254619] usb 3-6: configuration #1 > chosen from 1 choice > Feb 15 12:23:33 venus kernel: [23318.255811] scsi1 : SCSI emulation for > USB Mass Storage devices > Feb 15 12:23:36 venus kernel: [23321.068823] printk: 475 messages > suppressed. > Feb 15 12:23:36 venus kernel: [23321.068831] lost page write due to I/O > error on dm-0 > Feb 15 12:23:36 venus kernel: [23321.068842] lost page write due to I/O > error on dm-0 > Feb 15 12:23:38 venus kernel: [23323.251000] scsi 1:0:0:0: > Direct-Access WD 5000AAK External 1.06 PQ: 0 ANSI: 0 > Feb 15 12:23:38 venus kernel: [23323.253219] sd 1:0:0:0: [sdb] 976773168 > 512-byte hardware sectors (500108 MB) > Feb 15 12:23:38 venus kernel: [23323.253842] sd 1:0:0:0: [sdb] Write > Protect is off > Feb 15 12:23:38 venus kernel: [23323.256121] sd 1:0:0:0: [sdb] 976773168 > 512-byte hardware sectors (500108 MB) > Feb 15 12:23:38 venus kernel: [23323.256836] sd 1:0:0:0: [sdb] Write > Protect is off > Feb 15 12:23:38 venus kernel: [23323.256849] sdb: sdb1 > Feb 15 12:23:38 venus kernel: [23323.258946] sd 1:0:0:0: [sdb] Attached > SCSI disk > Feb 15 12:23:38 venus kernel: [23323.259003] sd 1:0:0:0: Attached scsi > generic sg0 type 0 > > ... and a few minutes later the wife (at home) patiently reboots the > computer, and wonders why I got her hooked on this silly machine. > > This issue is that the USB bus resets, and when the external disk is on > an LVM2 volume, the dm entry is not removed. When the USB disk comes > back online, it gets another /dev/sd? entry, and cannot be accessed > until we've done some cleanup, as in (as root, of course): > > /etc/init.d/myth-backend stop > umount /storage > dmsetup delete > vgscan > lvscan > vgchange -ay > mount /storage > /etc/init.d/myth-backend start > > Phew! Now ... I've considered adding this as a script to a > /etc/udev/rules.d/10-local.rules file (to match the serial number of the > disk, for example), but this is not a solution, only a way to let my > wife only notice the issues minimally (though it would have the side > effect of chopping up recordings into pieces with the backend restarts).. > > Note that the previous Fedora setup had the storage partition on a > primary partition (not LVM), so the USB drive would (upon reconnect) end > up with the same letter, and reading/writing from myth-backend would > simply resume after a pause. > > Some other things I've tried: > boot the kernel with > acpi=off (no noticable difference) > Add the following to /etc/udev/rules.d/10-local.rules > SYSFS{serial}=="574341533830343230313137",RUN+="/bin/sh -c '/bin/echo 64 > > /sys/block/%k/device/max_sectors'" > > Some of these were suggested in > (https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/61235) > > Picking through > sdparm -al /dev/sda > shows '0' for IDLE and STANDBY entries. > > Note that the problem appeared to be much more pronounced (resetting > every 10's of minutes instead of one to three times per day) when I had > the PCI bus underclocked in the BIOS ... I have since gotten a better > cooling solution, and no longer need to do that for stability. > > Finally, downgrading the usb to 1.1 (modprobe -r ehci_hcd) appeared to > solve the USB resets, but there was not quite enough bandwidth left to > both play and record to the disk. (Playback would skip every couple of > seconds ... the kids didn't care, but it'd drive my wife and I nuts!). > > Any hints? > > Thanks, > > Josh > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list Have you checked dmesg to see if you are getting any acpi errors? Have you tried a plugin USB board to see if that helps? You could also try rebuilding the kernel with usb debug enable and see if that pin points the problem. Joseph From tclug at beitsahour.net Tue Feb 19 20:59:42 2008 From: tclug at beitsahour.net (Munir Nassar) Date: Tue, 19 Feb 2008 20:59:42 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <47B5F1EF.3050407@radkeland.org> References: <47B5F1EF.3050407@radkeland.org> Message-ID: <47BB979E.2000609@beitsahour.net> Joshua Radke wrote: > Ok, this one has really gotten me, and google hasn't bailed me out yet, > so I was hoping someone on the list could. > > Some relevant system Info are an Athlon XP 1800+, nForce2 chipset > motherboard, RT2561(?) wireless card, 40 GB internal (IDE) disk, and > 500GB External Western Digital MyBook USB drive, running MythBuntu > 7.10. The kernel is 2.6.22-14-generic #1 SMP Tue Dec 18 08:02:57 UTC > 2007 i686 GNU/Linux. > > Note that the problem appeared to be much more pronounced (resetting > every 10's of minutes instead of one to three times per day) when I had > the PCI bus underclocked in the BIOS ... I have since gotten a better > cooling solution, and no longer need to do that for stability. it sounds to me like your hardware is suspect to begin with, if you had to underclock the system to get stability out of it there is a good chance something got damaged while it was overheating. it may be that it is still overheating. Try using a PCI USB2 card to see if it gets any better performance. while you are at it get a combo usb2/firewire card. Firewire uses less CPU and is imho more reliable. > Finally, downgrading the usb to 1.1 (modprobe -r ehci_hcd) appeared to > solve the USB resets, but there was not quite enough bandwidth left to > both play and record to the disk. (Playback would skip every couple of > seconds ... the kids didn't care, but it'd drive my wife and I nuts!). Full Speed usb1 runs at 12mbps, High Speed USB (High Speed) runs at 480Mbps, thats 2.5% of the theretical max meaning the system has to do a lot less.(mind you when using USB1 or USB2 the CPU still gets exercised a lot more than if you use Firewire.) It is also possible that the drive is getting kicked out because it is bad. at usb2 you are reaching the failed sectors a lot faster than with usb1. bad sectors tend to propagate so the reason it is getting kicked out more often is simply because more of it is bad. > Any hints? I would say that since you had heating issues with the system in the past either the issues still exist or your hardware was damaged somewhat by the overheating. try something like lucifer or cpuburn and bonnie++ to stress test the system. Test the drive in another system, use dd_rescue to dump the disk onto another or simply send it to /dev/null. dd_rescue will report if it came across any sectors it could not read (without locking your system up like dd does) dd_rescue /dev/sda /dev/null i do not know if the WDDIAG tools can test USB drives but it is worth a try as well. From Larry.Pint at ntuminc.com Wed Feb 20 09:41:00 2008 From: Larry.Pint at ntuminc.com (Larry R. Pint) Date: Wed, 20 Feb 2008 09:41:00 -0600 Subject: [tclug-list] Help needed Message-ID: The company I work with had their old Linux server die last Wednesday evening. Our systems consultants came in Thursday morning and we spent most of Thursday trying to get it back up with no luck. Later Thursday we moved things over to a new server. Things never got completely set up right and we still have some problems. Despite our calls for help, the system consultant has not been back in since Thursday. He was supposed to be here by 9:00 AM this morning, but as of 9:40 he is not here. We need a new Linux consulting firm. We are running Red hat Linux Enterprise version 4 on a dell server. Our problems revolve around permissions issues, Samba issues, printing issues and terminal emulation issues. We're not really doing anything fancy. Help is needed immediately. Please respond to me directly at larry.pint at ntuminc.com with offers of help and or recommendations of a good consulting firm that really knows Linux. This is a paid consulting position. Larry R. Pint National Truck Underwriting Managers, Inc. larry.pint at ntuminc.com 952-229-3451 952-893-1882 (FAX) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080220/01c99d8e/attachment.htm From josh at radkeland.org Wed Feb 20 09:49:35 2008 From: josh at radkeland.org (Joshua Radke) Date: Wed, 20 Feb 2008 09:49:35 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <47BB979E.2000609@beitsahour.net> References: <47B5F1EF.3050407@radkeland.org> <47BB979E.2000609@beitsahour.net> Message-ID: <47BC4C0F.8010907@radkeland.org> Munir Nassar wrote: > Joshua Radke wrote: >> Note that the problem appeared to be much more pronounced (resetting >> every 10's of minutes instead of one to three times per day) when I >> had the PCI bus underclocked in the BIOS ... I have since gotten a >> better cooling solution, and no longer need to do that for stability. > it sounds to me like your hardware is suspect to begin with, if you > had to underclock the system to get stability out of it there is a > good chance something got damaged while it was overheating. it may be > that it is still overheating. > The underclocking was to manage temperature, since I had a bad heatsink/fan, poorly mounted (by me). First, I'd like to thank everyone for their contributions. I pounded on the system last weekend, and made a few more discoveries. The punch line is that I've retired the drive for live use, and will be using it for backups instead. I have replaced the original disk (Western Digital MyBook, 500gb - having problems), with another off brand 300 gb disk I had, and everything's 'better'. I got a chanced to stress them a bit while moving recordings between the two, and here's what I discovered: 1) I can read from the troubled drive while writing to the OK one full speed, and in the hours of copying, I never got a single reset. 2) I can (almost) guarantee a reset (or 2 to 8 resets per hour) by writing to the WD Mybook while also reading from it. 3) The WD caused no resets until it had about 50 GB (used) on it. Other things tried: e2fsck --force (ok) badblocks (ok) New USB cable (and 2 others I had laying around) Move Drive to all three different Mobo Available USB hubs. Conclusion (best guesses; I encourage smarter folks to refine/discard these on the list!) 1) Based on an Ubuntu support page I can't seem to find anywhere; In kernels newer than ~2.6.9, some combinations of USB external storage and USB chipsets 'just do this'. 2) I can't recommend Western Digital Mybooks for external storage (at least for streaming). 3) I can recommend ID 0c0b:b352 Dura Micro, Inc. (Acomdata) (from lsusb -v). This one was sold as a Cavalry ?Pre-Formatted? CAUE37320 320GB. (buyer beware, these seem to have a fairly high DOA rate) 4) "Always back up your data" applies equally to media servers *laugh*. 5) On all external USB drives, do some serious concurrent Read/Write stress testing upon receipt so you can exchange if it's got issues. Thanks again, everyone! Josh From Larry.Pint at ntuminc.com Wed Feb 20 10:55:02 2008 From: Larry.Pint at ntuminc.com (Larry R. Pint) Date: Wed, 20 Feb 2008 10:55:02 -0600 Subject: [tclug-list] Help needed In-Reply-To: <47BC5A3C.5030801@Goecke-Dolan.com> Message-ID: Thanks to all the responded. Our regular consultant is on his way here. Hopefully our immediate crisis will be resolved. Longer term I will be recommending that we use a different firm for our Linux support. Larry > -----Original Message----- > From: Brian Dolan-Goecke [mailto:goeko at Goecke-Dolan.com] > Sent: Wednesday, February 20, 2008 10:50 AM > To: Larry R. Pint > Subject: Re: [tclug-list] Help needed > > > Larry, > > I am really sorry to hear that. I would like to help you, but I would > not be able to come during the day. I did forward your message to a > friend that might have time durning the day to help you. > > If you still need help, I can come help this evening (after 6:00) you > can give me a call on my cell phone 612-759-0967. Or you can send me an > email. > > Best of luck. > > ==>brian. > > > Larry R. Pint wrote: > > > > The company I work with had their old Linux server die last Wednesday > > evening. Our systems consultants came in Thursday morning and we spent > > most of Thursday trying to get it back up with no luck. Later Thursday > > we moved things over to a new server. Things never got completely set > > up right and we still have some problems. Despite our calls for help, > > the system consultant has not been back in since Thursday. He was > > supposed to be here by 9:00 AM this morning, but as of 9:40 he is not > > here. > > > > We need a new Linux consulting firm. We are running Red hat Linux > > Enterprise version 4 on a dell server. Our problems revolve around > > permissions issues, Samba issues, printing issues and terminal > > emulation issues. We're not really doing anything fancy. > > > > Help is needed immediately. Please respond to me directly at > > larry.pint at ntuminc.com with offers of > > help and or recommendations of a good consulting firm that really > > knows Linux. This is a paid consulting position. > > > > Larry R. Pint > > > > National Truck Underwriting Managers, Inc. > > > > larry.pint at ntuminc.com > > > > 952-229-3451 > > > > 952-893-1882 (FAX) > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > > tclug-list at mn-linux.org > > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > From jeff.rasmussen at gmail.com Wed Feb 20 11:30:57 2008 From: jeff.rasmussen at gmail.com (Jeff Rasmussen) Date: Wed, 20 Feb 2008 11:30:57 -0600 Subject: [tclug-list] ATI Radeon video card in Ubuntu 7.10 In-Reply-To: <1202388201.10652.2.camel@patty> References: <1202388201.10652.2.camel@patty> Message-ID: <9d6c82530802200930l7b244370pe4ae3883bfb3220a@mail.gmail.com> You will need to download the newest ATI driver from http://ati.amd.com When you find and download the driver for your model, you will need to change the file to an executable 'chmod +x ati*.run' Then instead of running the program directly, you can build Ubuntu packages so that upgrades will be easier. 'ati*.run --buildpkg Ubuntu/gutsy' If you don't use Gutsy, then you can list all packages by running 'ati*.run --listpkg' When you finish with building the packages install them with 'sudo dpkg -i *.deb' I believe everything will work at that point. Just restart your computer or just X windows. Jeff Rasmussen -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080220/254c4fd5/attachment.htm From goeko at Goecke-Dolan.com Thu Feb 21 12:49:38 2008 From: goeko at Goecke-Dolan.com (Brian Dolan-Goecke) Date: Thu, 21 Feb 2008 12:49:38 -0600 Subject: [tclug-list] *Updated* Audacity on Linux - Topic @PenguinsUnbound Linux Meeting Feb 23, 2008 Message-ID: <47BDC7C2.6050802@Goecke-Dolan.com> This months PenguinsUnbound.net meeting will be Saturday February 23, 2008 at TIES, 1667 Snelling Ave. N., St. Paul, MN 55108 from 10:00 to 12:00. (See the web site http://www.penguinsunbound.com/Location_for_Meetings for directions and more info.) We will be looking at and working with Audacity. My brief outline for the meeting, Getting Started Installing Audacity Starting Audacity Opening a file New projects Working with file Cutting file up Converting formats Creating a Ring Tone Please come and share your knowledge of Audacity! http://www.penguinsunbound.com/Future_Meetings/20080223_-_Audacity Thanks, hope to see you there. *UPDATE* Bill, is kinda enough to be coming this Saturday and will be * _GIVING_ _AWAY_ a OLPC * So stop by and talk a little about Linux, and maybe go home with a new PC! ==>brian. From admin at lctn.org Fri Feb 22 12:39:17 2008 From: admin at lctn.org (Raymond Norton) Date: Fri, 22 Feb 2008 12:39:17 -0600 Subject: [tclug-list] postfix question Message-ID: <47BF16D5.9090704@lctn.org> Anyone have experience with recipient_bcc_maps in postfix? I got "always_bcc" to work easily, but must be misunderstanding "recipient_bcc_maps" I added "recipient_bcc_maps = hash:/etc/postfix/recipient_bcc" to main.cf, and created "recipient_bcc", and ran postmap against it. The content of recipient_bcc is "user at domain.com acrchive at domain.com". It's not working,. so obviously I am misunderstanding how it works. Ultimately, I need to archive mail for 20 different domains, each needing their own archive folder. From admin at lctn.org Fri Feb 22 13:36:51 2008 From: admin at lctn.org (Raymond Norton) Date: Fri, 22 Feb 2008 13:36:51 -0600 Subject: [tclug-list] postfix question In-Reply-To: <20080222190435.GA1651@fire.msi.umn.edu> References: <47BF16D5.9090704@lctn.org> <20080222190435.GA1651@fire.msi.umn.edu> Message-ID: <47BF2453.3040608@lctn.org> Kevin Prigge wrote: > Dumb question, did you restart postfix after making the change? I > think that's only read once, it's not dynamic. I restarted mailscanner, did a postfix reload and am tailing the logs with "tail -f /var/log/mail.log -n10000 | grep "archive"" Tail raps away when main.cf has "always_bcc" configured, but shows nothing when recipient_bcc is configured. From jpschewe at mtu.net Fri Feb 22 13:58:45 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Fri, 22 Feb 2008 13:58:45 -0600 Subject: [tclug-list] Problem with laser printer Message-ID: <47BF2975.4090305@mtu.net> I've got an HP Laserjet 4 printer with a network card in it. When my wife prints from Windows 2000, everything looks nice and dark. When I print from Linux, everything looks washed out. Anyone have any ideas? -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From nate at refried.org Fri Feb 22 14:17:32 2008 From: nate at refried.org (Nate Straz) Date: Fri, 22 Feb 2008 15:17:32 -0500 Subject: [tclug-list] Problem with laser printer In-Reply-To: <47BF2975.4090305@mtu.net> References: <47BF2975.4090305@mtu.net> Message-ID: <20080222201732.GA12281@refried.org> On Fri, Feb 22, 2008 at 01:58:45PM -0600, Jon Schewe wrote: > I've got an HP Laserjet 4 printer with a network card in it. When my > wife prints from Windows 2000, everything looks nice and dark. When I > print from Linux, everything looks washed out. Anyone have any ideas? Is Linux configured to use EconoMode? Nate From cncole at earthlink.net Fri Feb 22 14:40:38 2008 From: cncole at earthlink.net (Chuck Cole) Date: Fri, 22 Feb 2008 14:40:38 -0600 Subject: [tclug-list] FW: *Updated* Audacity on Linux - Topic @PenguinsUnbound Linux Meeting Feb 23, 2008 - ride needed Message-ID: Can I bum a ride from Apple Valley? I'm very near Hwy 42 and Pilot Knob. My van died and I won;t have it replaced that soon. Chuck > -----Original Message----- > From: tclug-list-bounces at mn-linux.org > [mailto:tclug-list-bounces at mn-linux.org]On Behalf Of Brian Dolan-Goecke > Sent: Thursday, February 21, 2008 12:50 PM > To: tclug-list; tclug-announce > Subject: [tclug-list] *Updated* Audacity on Linux - Topic > @PenguinsUnbound Linux Meeting Feb 23, 2008 > > > This months PenguinsUnbound.net meeting will be Saturday February 23, > 2008 at TIES, > 1667 Snelling Ave. N., St. Paul, MN 55108 > from 10:00 to 12:00. > (See the web site http://www.penguinsunbound.com/Location_for_Meetings > for directions and more info.) > > We will be looking at and working with Audacity. > > My brief outline for the meeting, > > Getting Started > Installing Audacity > Starting Audacity > Opening a file > New projects > Working with file > Cutting file up > Converting formats > > Creating a Ring Tone > > Please come and share your knowledge of Audacity! > > > http://www.penguinsunbound.com/Future_Meetings/20080223_-_Audacity > > Thanks, hope to see you there. > > *UPDATE* > > Bill, is kinda enough to be coming this Saturday and will be > * _GIVING_ _AWAY_ a OLPC * > > So stop by and talk a little about Linux, and maybe go home with a new PC! > > ==>brian. > From jpschewe at mtu.net Sun Feb 24 18:49:47 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Sun, 24 Feb 2008 18:49:47 -0600 Subject: [tclug-list] Problem with laser printer In-Reply-To: <20080222201732.GA12281@refried.org> References: <47BF2975.4090305@mtu.net> <20080222201732.GA12281@refried.org> Message-ID: <47C210AB.6030307@mtu.net> Nate Straz wrote: > On Fri, Feb 22, 2008 at 01:58:45PM -0600, Jon Schewe wrote: > >> I've got an HP Laserjet 4 printer with a network card in it. When my >> wife prints from Windows 2000, everything looks nice and dark. When I >> print from Linux, everything looks washed out. Anyone have any ideas? >> > > Is Linux configured to use EconoMode? > No, it's configured for "High Quality". Resolution is set at 300 dpi, Grayscale, Balck Cartr. I just tried changing this to be controlled by "Printout Mode" and that didn't seem to help. It's using the "HP LaserJet 4 Foomatic/hpijs" ppd file. -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From jpschewe at mtu.net Sun Feb 24 18:55:14 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Sun, 24 Feb 2008 18:55:14 -0600 Subject: [tclug-list] Problem with laser printer In-Reply-To: <47C210AB.6030307@mtu.net> References: <47BF2975.4090305@mtu.net> <20080222201732.GA12281@refried.org> <47C210AB.6030307@mtu.net> Message-ID: <47C211F2.6000505@mtu.net> Jon Schewe wrote: > Nate Straz wrote: > >> On Fri, Feb 22, 2008 at 01:58:45PM -0600, Jon Schewe wrote: >> >> >>> I've got an HP Laserjet 4 printer with a network card in it. When my >>> wife prints from Windows 2000, everything looks nice and dark. When I >>> print from Linux, everything looks washed out. Anyone have any ideas? >>> >>> >> Is Linux configured to use EconoMode? >> >> > No, it's configured for "High Quality". Resolution is set at 300 dpi, > Grayscale, Balck Cartr. > I just tried changing this to be controlled by "Printout Mode" and that > didn't seem to help. > > It's using the "HP LaserJet 4 Foomatic/hpijs" ppd file. > > I just switched to the ljet4 PPD file instead of hpijs and it's much better now. -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From john.meier at gmail.com Mon Feb 25 12:05:24 2008 From: john.meier at gmail.com (John Meier) Date: Mon, 25 Feb 2008 12:05:24 -0600 Subject: [tclug-list] OT- Local shop for 3ware SATA RAID Message-ID: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> In a bit of a sticky wicket here - need to score a 3ware 8006-2LP SATA card or similar locally. Anyone know of any shops in the downtown areas ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080225/dec5788f/attachment.htm From srcfoo at gmail.com Mon Feb 25 12:16:54 2008 From: srcfoo at gmail.com (Eric Peterson) Date: Mon, 25 Feb 2008 12:16:54 -0600 Subject: [tclug-list] OT- Local shop for 3ware SATA RAID In-Reply-To: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> References: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> Message-ID: <579c6fd30802251016o6992f23dy3fd269e0869db61d@mail.gmail.com> I believe Micro Center carriers that card. On 2/25/08, John Meier wrote: > In a bit of a sticky wicket here - need to score a 3ware 8006-2LP SATA card > or similar locally. Anyone know of any shops in the downtown areas ? > > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > From john.meier at gmail.com Mon Feb 25 12:39:53 2008 From: john.meier at gmail.com (John Meier) Date: Mon, 25 Feb 2008 12:39:53 -0600 Subject: [tclug-list] OT- Local shop for 3ware SATA RAID In-Reply-To: <579c6fd30802251016o6992f23dy3fd269e0869db61d@mail.gmail.com> References: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> <579c6fd30802251016o6992f23dy3fd269e0869db61d@mail.gmail.com> Message-ID: <65293fcc0802251039i341e0acbg20da2fc9f775c9e5@mail.gmail.com> On Mon, Feb 25, 2008 at 12:16 PM, Eric Peterson wrote: > I believe Micro Center carriers that card. > Thanks - didn't see it on the web site, I'll call over there! > > On 2/25/08, John Meier wrote: > > In a bit of a sticky wicket here - need to score a 3ware 8006-2LP SATA > card > > or similar locally. Anyone know of any shops in the downtown areas ? > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080225/8bb7399f/attachment.htm From admin at lctn.org Mon Feb 25 12:58:09 2008 From: admin at lctn.org (Raymond Norton) Date: Mon, 25 Feb 2008 12:58:09 -0600 Subject: [tclug-list] solution for archiving mail gateway Message-ID: <47C30FC1.3040903@lctn.org> What's open source solutions are there for archiving email on a mail gateway . We need to archive per domain (20). We are using postfix with mailscanner, and it seems always_bcc, and recipient_bcc are not going to work, or may provide incomplete information. From andyzib at gmail.com Mon Feb 25 14:38:02 2008 From: andyzib at gmail.com (Andrew Zbikowski) Date: Mon, 25 Feb 2008 14:38:02 -0600 Subject: [tclug-list] solution for archiving mail gateway In-Reply-To: <47C30FC1.3040903@lctn.org> References: <47C30FC1.3040903@lctn.org> Message-ID: For incoming mail I once had a system wide procmail rule that would pipe a copy of every incoming mail to something like bzcat >> /mnt/Archive/username-Incoming-Archive.bz2. When you bunzip2ed the archive you got a mbox mail file. If a user needed to go through their archive I just created a decompressed copy of the file and put it under ~/mail, and they subscribed to it as they would any other IMAP folder. I never got around for something similar for outgoing mail. -- Andrew S. Zbikowski | http://andy.zibnet.us IT Outhouse Blog Thing | http://www.itouthouse.com From admin at lctn.org Mon Feb 25 14:43:15 2008 From: admin at lctn.org (Raymond Norton) Date: Mon, 25 Feb 2008 14:43:15 -0600 Subject: [tclug-list] solution for archiving mail gateway In-Reply-To: References: <47C30FC1.3040903@lctn.org> Message-ID: <47C32863.9040505@lctn.org> > When you bunzip2ed the archive you got a mbox mail file. If a user > needed to go through their archive I just created a decompressed copy > of the file and put it under ~/mail, and they subscribed to it as they > would any other IMAP folder. > Always_bcc is similar, but it does not show who the message was sent to, unless you open it, which wont work for us. From jima at beer.tclug.org Mon Feb 25 15:01:27 2008 From: jima at beer.tclug.org (Jima) Date: Mon, 25 Feb 2008 15:01:27 -0600 (CST) Subject: [tclug-list] solution for archiving mail gateway In-Reply-To: References: <47C30FC1.3040903@lctn.org> Message-ID: On Mon, 25 Feb 2008, Andrew Zbikowski wrote: > For incoming mail I once had a system wide procmail rule that would > pipe a copy of every incoming mail to something like bzcat >> > /mnt/Archive/username-Incoming-Archive.bz2. > > When you bunzip2ed the archive you got a mbox mail file. If a user > needed to go through their archive I just created a decompressed copy > of the file and put it under ~/mail, and they subscribed to it as they > would any other IMAP folder. > > I never got around for something similar for outgoing mail. They were doing the same at my work when I started here. They really did want outgoing, though, so I ended up finding a solution years ago (for sendmail, anyway): http://www.freakout.de/logall.c There's a similar solution for exim: unseen save /var/log/fullmail I'm guessing one might be able to hack either solution to work per-domain. Not sure what hacks are possible in postfix; I've never tried. Jima From nassarmu at beitsahour.net Mon Feb 25 18:50:33 2008 From: nassarmu at beitsahour.net (Munir Nassar) Date: Mon, 25 Feb 2008 18:50:33 -0600 Subject: [tclug-list] OT- Local shop for 3ware SATA RAID In-Reply-To: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> References: <65293fcc0802251005o4f211c99r449d49900c1cce81@mail.gmail.com> Message-ID: <47C36259.7050805@beitsahour.net> John Meier wrote: > In a bit of a sticky wicket here - need to score a 3ware 8006-2LP SATA > card or similar locally. Anyone know of any shops in the downtown areas ? After years of using 3ware cards, i have yet to find a local source. At some point General NanoSystems had a couple of the models in stock but nothing that could be relied on. We get all of ours from NewEgg. Nice cards, we use them for almost all of our servers my only complaint is that they lack a proc interface to poke at them and 3dmd or tw_cli must be used. From josh at joshwelch.com Tue Feb 26 13:18:28 2008 From: josh at joshwelch.com (Josh Welch) Date: Tue, 26 Feb 2008 19:18:28 +0000 Subject: [tclug-list] Removing Xen remnant Message-ID: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> I've got a box running Xen that had a domain go flaky on it, I ended up doing an `xm destroy` on that domain to kill it. Now I always see that domain when I do an `xm list`. It has no ID and no state, but it shows up in the display like so: Name ID Mem VCPUs State Time(s) 10_virtmach1 512 1 0.0 Domain-0 0 404 1 r----- 112527.4 Anyone know where xen gets this data from? Thanks, Josh From tclug at natecarlson.com Tue Feb 26 14:26:35 2008 From: tclug at natecarlson.com (Nate Carlson) Date: Tue, 26 Feb 2008 14:26:35 -0600 (CST) Subject: [tclug-list] Removing Xen remnant In-Reply-To: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> References: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> Message-ID: On Tue, 26 Feb 2008, Josh Welch wrote: > I've got a box running Xen that had a domain go flaky on it, I ended up > doing an `xm destroy` on that domain to kill it. Now I always see that > domain when I do an `xm list`. It has no ID and no state, but it shows > up in the display like so: > > Name ID Mem VCPUs State > Time(s) > 10_virtmach1 512 1 0.0 > Domain-0 0 404 1 r----- 112527.4 > > Anyone know where xen gets this data from? What version of Xen? It's most likely in XenSource; I haven't see it that corrupted before, though. ;) There were a lot of bugs with this type of thing in 3.x < 3.2. ------------------------------------------------------------------------ | nate carlson | natecars at natecarlson.com | http://www.natecarlson.com | | depriving some poor village of its idiot since 1981 | ------------------------------------------------------------------------ From marc at e-skinner.net Tue Feb 26 14:05:25 2008 From: marc at e-skinner.net (Marc Skinner) Date: Tue, 26 Feb 2008 14:05:25 -0600 Subject: [tclug-list] Removing Xen remnant In-Reply-To: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> References: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> Message-ID: <47C47105.3050204@e-skinner.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 which distro are you running? in RedHat based systems: /etc/xen will have all your xen config files, make sure you delete that, and then if you are using File based images make sure you remove that as well. xm destroy - just stops the virt machine, it doesn't "destroy" it from the system. Josh Welch wrote: > I've got a box running Xen that had a domain go flaky on it, I ended > up doing an `xm destroy` on that domain to kill it. Now I always see > that domain when I do an `xm list`. It has no ID and no state, but it > shows up in the display like so: > > Name ID Mem VCPUs State > Time(s) > 10_virtmach1 512 1 0.0 > Domain-0 0 404 1 r----- 112527.4 > > Anyone know where xen gets this data from? > > Thanks, > Josh > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) iD8DBQFHxHEFvE9HrEfeE4cRAonkAKDeptlPQMA53k2h4ndti9doLojDvQCfb2OV e6XJirvsIwQopHKDUyvyz8Y= =RnhF -----END PGP SIGNATURE----- From josh at joshwelch.com Tue Feb 26 15:44:29 2008 From: josh at joshwelch.com (Josh Welch) Date: Tue, 26 Feb 2008 21:44:29 +0000 Subject: [tclug-list] Removing Xen remnant In-Reply-To: References: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> Message-ID: <20080226214429.jc1oyfzrnoooosos@joshwelch.com> Quoting Nate Carlson : > On Tue, 26 Feb 2008, Josh Welch wrote: >> I've got a box running Xen that had a domain go flaky on it, I >> ended up doing an `xm destroy` on that domain to kill it. Now I >> always see that domain when I do an `xm list`. It has no ID and no >> state, but it shows up in the display like so: >> >> Name ID Mem VCPUs State >> Time(s) >> 10_virtmach1 512 1 >> 0.0 >> Domain-0 0 404 1 >> r----- 112527.4 >> >> Anyone know where xen gets this data from? > > What version of Xen? It's most likely in XenSource; I haven't see it > that corrupted before, though. ;) Did you mean XenStore? > There were a lot of bugs with this type of thing in 3.x < 3.2. Looks like I'm running 3.1.1. Perhaps I need to look at upgrading. Not sure if Oracle Virtual Server has newer packages for Xen yet. I tracked down the issue to a config file hanging around. It was /var/lib/send/domains//config.sxp. Deleting that and restarting xend made the phantom domain go away. Thanks, Josh From tclug at natecarlson.com Tue Feb 26 16:22:50 2008 From: tclug at natecarlson.com (Nate Carlson) Date: Tue, 26 Feb 2008 16:22:50 -0600 (CST) Subject: [tclug-list] Removing Xen remnant In-Reply-To: <20080226214429.jc1oyfzrnoooosos@joshwelch.com> References: <20080226191828.aulntvtt4os4ow8k@joshwelch.com> <20080226214429.jc1oyfzrnoooosos@joshwelch.com> Message-ID: On Tue, 26 Feb 2008, Josh Welch wrote: >> What version of Xen? It's most likely in XenSource; I haven't see it >> that corrupted before, though. ;) > > Did you mean XenStore? Indeed. Sigh. ;) > Looks like I'm running 3.1.1. Perhaps I need to look at upgrading. Not > sure if Oracle Virtual Server has newer packages for Xen yet. > > I tracked down the issue to a config file hanging around. It was > /var/lib/send/domains//config.sxp. Deleting that and restarting > xend made the phantom domain go away. Sweet! ------------------------------------------------------------------------ | nate carlson | natecars at natecarlson.com | http://www.natecarlson.com | | depriving some poor village of its idiot since 1981 | ------------------------------------------------------------------------ From daniel.armbrust.list at gmail.com Wed Feb 27 16:13:31 2008 From: daniel.armbrust.list at gmail.com (Dan Armbrust) Date: Wed, 27 Feb 2008 16:13:31 -0600 Subject: [tclug-list] shell script - find shell script parent directory? Message-ID: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> I'm trying to find a variable or a command that will tell me where the currently executing script lives. So, If I'm running /foo/bar.sh, I'm looking for a something that will give me "/foo/", while I'm inside the script bar.sh. On windows, this is "%~dp0". On linux, `pwd` isn't right all the time, because it tells the location the script was executed from, not where it lives. $0 is not normalized, it can have all sorts of interesting paths in it. The best I have come up with so far is: PROG=`which $0` PROGDIR=`dirname $PROG` echo $PROGDIR But I'm not yet convinced that this will always work. Is there an easier way to do this? Thanks, Dan From mbmiller at taxa.epi.umn.edu Wed Feb 27 16:45:23 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 16:45:23 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> Message-ID: On Wed, 27 Feb 2008, Dan Armbrust wrote: > I'm trying to find a variable or a command that will tell me where the > currently executing script lives. > > So, If I'm running /foo/bar.sh, I'm looking for a something that will > give me "/foo/", while I'm inside the script bar.sh. > > On windows, this is "%~dp0". > > On linux, `pwd` isn't right all the time, because it tells the > location the script was executed from, not where it lives. > > $0 is not normalized, it can have all sorts of interesting paths in it. > > The best I have come up with so far is: > > PROG=`which $0` > PROGDIR=`dirname $PROG` > echo $PROGDIR > > But I'm not yet convinced that this will always work. You'll have a problem with spaces. This would fix that problem by using quotation marks: PROG=`which "$0"` PROGDIR=`dirname "$PROG"` echo $PROGDIR But if that's all you want to do, you can shorten it to this single line: echo $(dirname "$(which "$0")") I *think* that will always work in bash, but then I am not an expert on bash scripts. Did you see this?: http://www.linuxquestions.org/questions/programming-9/discovering-script-directory-path-531088/ That might help. This guy is totally wrong, so ignore him!: http://unixjunkie.blogspot.com/2006/08/old-but-useful-shell-tricks.html Mike From florin at iucha.net Wed Feb 27 17:13:30 2008 From: florin at iucha.net (Florin Iucha) Date: Wed, 27 Feb 2008 17:13:30 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> Message-ID: <20080227231329.GJ3247@iris.iucha.org> On Wed, Feb 27, 2008 at 04:13:31PM -0600, Dan Armbrust wrote: > I'm trying to find a variable or a command that will tell me where the > currently executing script lives. > [snip] > > The best I have come up with so far is: > > PROG=`which $0` > PROGDIR=`dirname $PROG` > echo $PROGDIR > > But I'm not yet convinced that this will always work. Well, what happens if $SCRIPT is not in $PATH? 'which' won't find it.... Sorry, I don't have a good solution handy. florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080227/56d2d74d/attachment.pgp From tclug at beitsahour.net Wed Feb 27 17:23:19 2008 From: tclug at beitsahour.net (Munir Nassar) Date: Wed, 27 Feb 2008 17:23:19 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> Message-ID: <47C5F0E7.8000508@beitsahour.net> Dan Armbrust wrote: > I'm trying to find a variable or a command that will tell me where the > currently executing script lives. > > So, If I'm running /foo/bar.sh, I'm looking for a something that will > give me "/foo/", while I'm inside the script bar.sh. > $ dirname /foo/bar.sh /foo $ From jpschewe at mtu.net Wed Feb 27 17:59:55 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Wed, 27 Feb 2008 17:59:55 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> Message-ID: <47C5F97B.7050702@mtu.net> Dan Armbrust wrote: > I'm trying to find a variable or a command that will tell me where the > currently executing script lives. > > So, If I'm running /foo/bar.sh, I'm looking for a something that will > give me "/foo/", while I'm inside the script bar.sh. > > On windows, this is "%~dp0". > > On Linux you have two options: 1) If you can guarantee that dirname exists then this works: mypath=`dirname $0` cd ${mypath} mypath=`pwd` 2) If you're not sure that dirname exists, this works well: mypath="`echo $0 | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`" cd ${mypath} mypath=`pwd` -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:20:05 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:20:05 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <20080227231329.GJ3247@iris.iucha.org> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <20080227231329.GJ3247@iris.iucha.org> Message-ID: On Wed, 27 Feb 2008, Florin Iucha wrote: > Well, what happens if $SCRIPT is not in $PATH? 'which' won't find > it.... Yes it will. $ mkdir foo $ touch foo/script $ chmod 755 foo/script $ which foo/script ./foo/script Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:21:48 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:21:48 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C5F0E7.8000508@beitsahour.net> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F0E7.8000508@beitsahour.net> Message-ID: On Wed, 27 Feb 2008, Munir Nassar wrote: > Dan Armbrust wrote: >> I'm trying to find a variable or a command that will tell me where the >> currently executing script lives. >> >> So, If I'm running /foo/bar.sh, I'm looking for a something that will >> give me "/foo/", while I'm inside the script bar.sh. >> > > $ dirname /foo/bar.sh > /foo > $ $ cd foo $ dirname bar.sh . Oops. He wants something that works when the full path is not given on the command line. See my earlier message. Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:26:45 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:26:45 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C5F97B.7050702@mtu.net> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: On Wed, 27 Feb 2008, Jon Schewe wrote: > On Linux you have two options: > 1) If you can guarantee that dirname exists then this works: > mypath=`dirname $0` That won't work when there are spaces in $0 -- see my earlier message. > cd ${mypath} That won't work when there are spaces in $mypath. > 2) If you're not sure that dirname exists, this works well: > mypath="`echo $0 | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`" That doesn't find the path to the file, it just takes the path that is given, but if the path isn't given, it fails. See my earlier message and the original posting. > cd ${mypath} Again, that won't work when there are spaces in $mypath. Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:39:17 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:39:17 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: I think some people aren't getting what the problem is here. The script has been executed from the command line and we want the script to report the directory in which it is located. There are four ways that the script could have been executed: (1) absolute path example: /usr/local/bin/script.bash (2) relative path example: ../../../local/bin/script.bash example: ./local/bin/script.bash example: local/bin/script.bash (3) in search path ($PATH) example: script.bash (4) as above but using globbing example: local/bin/scri* In 1-3, the variable $0 will be the command as entered. In "4" the filename is expanded and $0 becomes that filename, so you really only have to deal with the first three. You have to remember that filenames or directory names can contain spaces. This is a really important little "gottcha" because you can do a whole lot of testing and think everything is fine because you never use spaces in filenames. Some people like spaces (are they crazy? ;-), and your program will not seem so great to them! Luckily, use of "which," with appropriate quoting around shell variables to deal with the spaces, seems to work for every case. Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:42:34 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:42:34 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <20080227231329.GJ3247@iris.iucha.org> Message-ID: On Wed, 27 Feb 2008, Mike Miller wrote: > On Wed, 27 Feb 2008, Florin Iucha wrote: > >> Well, what happens if $SCRIPT is not in $PATH? 'which' won't find it.... > > Yes it will. > > $ mkdir foo > $ touch foo/script > $ chmod 755 foo/script > $ which foo/script > ./foo/script Oops. What was I thinking? Yes, it doesn't find the full path, which is what you want. You were right, Florin, as usual. Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:50:27 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:50:27 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <20080227231329.GJ3247@iris.iucha.org> Message-ID: On Wed, 27 Feb 2008, Mike Miller wrote: > On Wed, 27 Feb 2008, Mike Miller wrote: > >> On Wed, 27 Feb 2008, Florin Iucha wrote: >> >>> Well, what happens if $SCRIPT is not in $PATH? 'which' won't find >>> it.... >> >> Yes it will. >> >> $ mkdir foo >> $ touch foo/script >> $ chmod 755 foo/script >> $ which foo/script >> ./foo/script > > Oops. What was I thinking? Yes, it doesn't find the full path, which > is what you want. You were right, Florin, as usual. Here's the problem: It did work for me when I used it in a script, then it didn't work from the command line. The reason is that I have an alias that added the "--show-dot" option: $ which which alias which='alias | /usr/bin/which --tty-only --read-alias --show-dot --show-tilde' /usr/bin/which But within the script, it was using /usr/bin/which instead of the alias. You can turn off the alias with a backslash: $ \which which /usr/bin/which So I will stick by my earlier idea that "which" is the way to go, but you have to be sure that you are getting the full path from "which." Will use of "\which" guarantee that full paths are given? Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 18:53:35 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 18:53:35 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: On Wed, 27 Feb 2008, Mike Miller wrote: > There are four ways that the script could have been executed: > > (1) absolute path > example: /usr/local/bin/script.bash > > (2) relative path > example: ../../../local/bin/script.bash > example: ./local/bin/script.bash > example: local/bin/script.bash > > (3) in search path ($PATH) > example: script.bash > > (4) as above but using globbing > example: local/bin/scri* It is also possible to have used ${HOME} or ~ in the file specification, or even some other variable like ${SCRIPT} or ${SOME_DIR}. I think those end up working like "4" and the variable or ~ is expanded to a string that then becomes $0. Mike From brockn at gmail.com Wed Feb 27 18:59:33 2008 From: brockn at gmail.com (Brock Noland) Date: Wed, 27 Feb 2008 18:59:33 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <20080227231329.GJ3247@iris.iucha.org> Message-ID: <741dcbb80802271659o2a9ad8f8jc8e19699934915ea@mail.gmail.com> > (1) absolute path > example: /usr/local/bin/script.bash [[ ${0:0:1} == "/" ]] > (2) relative path > example: ../../../local/bin/script.bash > example: ./local/bin/script.bash > example: local/bin/script.bash $PWD/`dirname` > (3) in search path ($PATH) > example: script.bash which $0 > (4) as above but using globbing > example: local/bin/scri* $PWD/`dirname` If you are correct that these are the only possibilities. Check for 1 and 3 first, then go with 3/4. Brock From jpschewe at mtu.net Wed Feb 27 19:08:49 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Wed, 27 Feb 2008 19:08:49 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: <47C609A1.1060206@mtu.net> Mike Miller wrote: > On Wed, 27 Feb 2008, Jon Schewe wrote: > > >> On Linux you have two options: >> 1) If you can guarantee that dirname exists then this works: >> mypath=`dirname $0` >> > > That won't work when there are spaces in $0 -- see my earlier message. > > So so put quotes around $0. -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From jpschewe at mtu.net Wed Feb 27 19:11:28 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Wed, 27 Feb 2008 19:11:28 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: <47C60A40.70100@mtu.net> Mike Miller wrote: > You have to remember that filenames or directory names can contain spaces. > This is a really important little "gottcha" because you can do a whole lot > of testing and think everything is fine because you never use spaces in > filenames. Some people like spaces (are they crazy? ;-), and your program > will not seem so great to them! > > Don't put spaces in filenames in the first place... Also are you handling backslashes in filenames and control characters? -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From kjh at flyballdogs.com Wed Feb 27 19:26:33 2008 From: kjh at flyballdogs.com (Kathryn Hogg) Date: Wed, 27 Feb 2008 19:26:33 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> Message-ID: <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> If you want to print just the directory that the script is located in do (cd $(dirname $0); pwd) if you want the full path to the script then this will suffice echo $(cd $(dirname $0); pwd)/$(basename $0) -- Kathryn http://womensfooty.com From mbmiller at taxa.epi.umn.edu Wed Feb 27 19:45:15 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 19:45:15 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <741dcbb80802271659o2a9ad8f8jc8e19699934915ea@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <20080227231329.GJ3247@iris.iucha.org> <741dcbb80802271659o2a9ad8f8jc8e19699934915ea@mail.gmail.com> Message-ID: On Wed, 27 Feb 2008, Brock Noland wrote: >> (1) absolute path >> example: /usr/local/bin/script.bash > > [[ ${0:0:1} == "/" ]] I don't know what that does or how it is used. I put it in a script and nothing happened. >> (2) relative path >> example: ../../../local/bin/script.bash >> example: ./local/bin/script.bash >> example: local/bin/script.bash > > $PWD/`dirname` > >> (3) in search path ($PATH) >> example: script.bash > > which $0 > >> (4) as above but using globbing >> example: local/bin/scri* > > $PWD/`dirname` > > > If you are correct that these are the only possibilities. Check for 1 > and 3 first, then go with 3/4. Are you saying that it is better to have a series of checks followed by a series of different commands, one for each case, instead of doing this... echo $(dirname "$(\which "$0")") ...which seems to always work for any case? Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 19:46:10 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 19:46:10 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C609A1.1060206@mtu.net> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <47C609A1.1060206@mtu.net> Message-ID: On Wed, 27 Feb 2008, Jon Schewe wrote: > So so put quotes around $0. I think you are definitely making progress. ;-) Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 20:02:38 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 20:02:38 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C60A40.70100@mtu.net> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <47C60A40.70100@mtu.net> Message-ID: On Wed, 27 Feb 2008, Jon Schewe wrote: > Mike Miller wrote: >> You have to remember that filenames or directory names can contain spaces. >> This is a really important little "gottcha" because you can do a whole lot >> of testing and think everything is fine because you never use spaces in >> filenames. Some people like spaces (are they crazy? ;-), and your program >> will not seem so great to them! >> > Don't put spaces in filenames in the first place... I know what you mean, but I don't want to give someone a script where the first line of documentation says "make sure that none of your filenames contains a space." > Also are you handling backslashes in filenames and control characters? Backslashes -- yes. Control characters -- mostly, I guess. I just tried it with a carriage return in the path and it worked, but it might not look right on your output device, whatever that is. Then I put a newline in the path. That didn't quite work out. The output used a space where the newline should have been. So ... I'll bet you could see this coming ... "Don't put newlines in filenames in the first place..." Mike From mbmiller at taxa.epi.umn.edu Wed Feb 27 20:11:28 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 20:11:28 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> Message-ID: On Wed, 27 Feb 2008, Kathryn Hogg wrote: > If you want to print just the directory that the script is located in do > > (cd $(dirname $0); pwd) That assumes that the script is in "dirname $0" but what if the script is in /usr/local/bin and /usr/local/bin is in the path and the user typed this: script.bash Then "dirname $0" will fail, and that is why the original poster suggested use of "which". Also, you need quotes around the $0. > if you want the full path to the script then this will suffice > echo $(cd $(dirname $0); pwd)/$(basename $0) That has the same problem as above. Mike From strayf at freeshell.org Wed Feb 27 20:42:05 2008 From: strayf at freeshell.org (Steve Cayford) Date: Wed, 27 Feb 2008 20:42:05 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> Message-ID: <47C61F7D.2090609@freeshell.org> Mike Miller wrote: > On Wed, 27 Feb 2008, Kathryn Hogg wrote: > >> If you want to print just the directory that the script is located in do >> >> (cd $(dirname $0); pwd) > > That assumes that the script is in "dirname $0" but what if the script is > in /usr/local/bin and /usr/local/bin is in the path and the user typed > this: > > script.bash > > Then "dirname $0" will fail, and that is why the original poster suggested > use of "which". Also, you need quotes around the $0. > Hm. Not on my system. When I run a script in /usr/local/bin by typing test.sh, `dirname "$0"` returns /usr/local/bin. The path is getting expanded before the script runs. That seems to work when running sh, bash, or tcsh at the command line or in the script itself. -Steve From hick0088 at tc.umn.edu Wed Feb 27 22:10:07 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Wed, 27 Feb 2008 22:10:07 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> Message-ID: <1204171807.15788.17.camel@localhost> On Wed, 2008-02-27 at 19:26 -0600, Kathryn Hogg wrote: > If you want to print just the directory that the script is located in do > > (cd $(dirname $0); pwd) This looks like a winner to me. To handle spaces and at least some other nastiness, this seems to work (cd "$(dirname "$0")"; pwd) I've had it work with directories including newlines and backslashes. I was a little surprised, but apparently bash separately parses double-quotes which are within $( .. ) blobs. I might change the semicolon to '&&'. If there's somehow an error, you'd most likely get a zero-length return string instead of the original workind dir, and $? would be set non-zero. -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080227/3584077f/attachment-0001.pgp From mbmiller at taxa.epi.umn.edu Wed Feb 27 22:15:36 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Wed, 27 Feb 2008 22:15:36 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C61F7D.2090609@freeshell.org> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <47C61F7D.2090609@freeshell.org> Message-ID: On Wed, 27 Feb 2008, Steve Cayford wrote: > Mike Miller wrote: >> On Wed, 27 Feb 2008, Kathryn Hogg wrote: >> >>> If you want to print just the directory that the script is located in >>> do >>> >>> (cd $(dirname $0); pwd) >> >> That assumes that the script is in "dirname $0" but what if the script >> is in /usr/local/bin and /usr/local/bin is in the path and the user >> typed this: >> >> script.bash >> >> Then "dirname $0" will fail, and that is why the original poster >> suggested use of "which". Also, you need quotes around the $0. > > Hm. Not on my system. When I run a script in /usr/local/bin by typing > test.sh, `dirname "$0"` returns /usr/local/bin. The path is getting > expanded before the script runs. > > That seems to work when running sh, bash, or tcsh at the command line or > in the script itself. My apologies to Kathryn Hogg. It seems that you are correct because if I run "echo $0" within the script it returns the full path from $PATH even when nothing but the filename was typed on the command line. Another guy named Brock Noland wrote to tell me that "which" in Solaris does not return the full path. He is quite right about that. I guess it is a GNU thing. He also thought Kathryn's method would work. So then, does Kathryn's method always work? (with quotes around $0): (cd $(dirname "$0"); pwd) I think maybe it does. It is an elegant solution. Mike From strayf at freeshell.org Wed Feb 27 23:37:40 2008 From: strayf at freeshell.org (Steve Cayford) Date: Wed, 27 Feb 2008 23:37:40 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <47C61F7D.2090609@freeshell.org> Message-ID: <47C648A4.3030408@freeshell.org> Mike Miller wrote: > [...] > So then, does Kathryn's method always work? (with quotes around $0): > > (cd $(dirname "$0"); pwd) > Note that if $0 is a relative path and you change directories in your script you'll no longer get the right result. However you should be able to do this early in the script and save the path for later if necessary. -Steve From kjh at flyballdogs.com Wed Feb 27 23:46:40 2008 From: kjh at flyballdogs.com (Kathryn Hogg) Date: Wed, 27 Feb 2008 23:46:40 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> Message-ID: <36891.192.168.0.7.1204177600.squirrel@www.flyballdogs.com> Mike Miller wrote: > On Wed, 27 Feb 2008, Kathryn Hogg wrote: > >> If you want to print just the directory that the script is located in do >> >> (cd $(dirname $0); pwd) > > That assumes that the script is in "dirname $0" but what if the script is > in /usr/local/bin and /usr/local/bin is in the path and the user typed > this: > > script.bash If I put the script in a directory in my search path and then run the script without a pathname, $0 is set to the full path name to the executable. So dirname $0 works. I'm using bash 3.2.33 on fedora 8 and I'm pretty sure that ksh has worked like this on various unix implementations (hp-ux, solaris, aix, tru64) for many years. -- Kathryn http://womensfooty.com From kjh at flyballdogs.com Wed Feb 27 23:51:24 2008 From: kjh at flyballdogs.com (Kathryn Hogg) Date: Wed, 27 Feb 2008 23:51:24 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <47C61F7D.2090609@freeshell.org> Message-ID: <13982.192.168.0.7.1204177884.squirrel@www.flyballdogs.com> Mike Miller wrote: > My apologies to Kathryn Hogg. It seems that you are correct because if I > run "echo $0" within the script it returns the full path from $PATH even > when nothing but the filename was typed on the command line. To be honest, I was pleasantly surprised the first time I found out that $0 would expand to the full pathname many years ago. > So then, does Kathryn's method always work? (with quotes around $0): > > (cd $(dirname "$0"); pwd) > > I think maybe it does. It is an elegant solution. That definitely won't work. It will fix the dirname but then cd would see something like cd /yes/i am/stupid. so we need to quote the arguments to both dirname and cd. -- Kathryn http://womensfooty.com From tclug at freakzilla.com Thu Feb 28 00:46:46 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 00:46:46 -0600 (CST) Subject: [tclug-list] WPA2 encryption Message-ID: I'll admit, last time I tried to do wireless on Linux WEP was barely even around. I recently put Linux on my old Powerbook. It's pretty cool, every piece of hardware is detected and works, including the wireless card. But I use WPA2 and I have NO idea how to get that in there. The WiFi setup tool that comes with Yellowdog Linux asks for a "key", but apparently that's not the actual WPA passphrase. I've done some googling and it looks like I need to generate the key somehow... but I'm not too sure how. No links to key generation utilities actually worked. Anyone? -Yaron -- From mbmiller at taxa.epi.umn.edu Thu Feb 28 01:13:09 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Thu, 28 Feb 2008 01:13:09 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <13982.192.168.0.7.1204177884.squirrel@www.flyballdogs.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <47C61F7D.2090609@freeshell.org> <13982.192.168.0.7.1204177884.squirrel@www.flyballdogs.com> Message-ID: On Wed, 27 Feb 2008, Kathryn Hogg wrote: > Mike Miller wrote: >> So then, does Kathryn's method always work? (with quotes around $0): >> >> (cd $(dirname "$0"); pwd) >> >> I think maybe it does. It is an elegant solution. > > That definitely won't work. It will fix the dirname but then cd would see > something like > cd /yes/i am/stupid. so we need to quote the arguments to both dirname > and cd. Yep, I forgot that too. Quotes are needed... cd "you/are smarter/than/me" Mike From mbmiller at taxa.epi.umn.edu Thu Feb 28 01:18:01 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Thu, 28 Feb 2008 01:18:01 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204171807.15788.17.camel@localhost> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> Message-ID: On Wed, 27 Feb 2008, Mike Hicks wrote: > On Wed, 2008-02-27 at 19:26 -0600, Kathryn Hogg wrote: >> If you want to print just the directory that the script is located in do >> >> (cd $(dirname $0); pwd) > > This looks like a winner to me. To handle spaces and at least some > other nastiness, this seems to work > > (cd "$(dirname "$0")"; pwd) > > I've had it work with directories including newlines and backslashes. I > was a little surprised, but apparently bash separately parses > double-quotes which are within $( .. ) blobs. I guess you can do plenty of nesting of $() terms also. I was pleasantly surprised to see that it handled newlines because the one I wrote earlier did not handle them correctly. > I might change the semicolon to '&&'. If there's somehow an error, > you'd most likely get a zero-length return string instead of the > original workind dir, and $? would be set non-zero. That's another new one for me -- I didn't know that "&&" could be used instead of the semicolon. Steve Crayford was definitely right about this: Note that if $0 is a relative path and you change directories in your script you'll no longer get the right result. However you should be able to do this early in the script and save the path for later if necessary. Good call, Steve. Mike From cncole at earthlink.net Thu Feb 28 05:16:54 2008 From: cncole at earthlink.net (Chuck Cole) Date: Thu, 28 Feb 2008 05:16:54 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: Message-ID: > -----Original Message----- > From: tclug-list-bounces at mn-linux.org > [mailto:tclug-list-bounces at mn-linux.org]On Behalf Of Yaron > > > I recently put Linux on my old Powerbook. It's pretty cool, every piece of > hardware is detected and works, including the wireless card. But I use > WPA2 and I have NO idea how to get that in there. The WiFi setup tool that > comes with Yellowdog Linux asks for a "key", but apparently that's not the > actual WPA passphrase. > > I've done some googling and it looks like I need to generate the key > somehow... but I'm not too sure how. No links to key generation utilities > actually worked. > > Anyone? > > > -Yaron > Linksys routers can do this. I copy the resulting key to a text file that I keep in a flash drive so I can merely paste in the long key when I need it. Chuck From jpschewe at mtu.net Thu Feb 28 07:22:19 2008 From: jpschewe at mtu.net (Jon Schewe) Date: Thu, 28 Feb 2008 07:22:19 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> Message-ID: <47C6B58B.1010200@mtu.net> > That's another new one for me -- I didn't know that "&&" could be used > instead of the semicolon. > > But be careful, && and ; mean different things. A semicolon between two commands executes both commands. Using && executes the first command and if it returns success, then the second command is executed, if the first command fails, the second command doesn't execute. -- Jon Schewe | http://mtu.net/~jpschewe If you see an attachment named signature.asc, this is my digital signature. See http://www.gnupg.org for more information. For I am convinced that neither death nor life, neither angels nor demons, neither the present nor the future, nor any powers, neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord. - Romans 8:38-39 From hick0088 at tc.umn.edu Thu Feb 28 08:18:13 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 08:18:13 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <47C6B58B.1010200@mtu.net> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> Message-ID: <1204208293.15788.22.camel@localhost> On Thu, 2008-02-28 at 07:22 -0600, Jon Schewe wrote: > > That's another new one for me -- I didn't know that "&&" could be used > > instead of the semicolon. > > > > > But be careful, && and ; mean different things. A semicolon between two > commands executes both commands. Using && executes the first command > and if it returns success, then the second command is executed, if the > first command fails, the second command doesn't execute. Yep, and conversely, '||' can be used to execute a command if the first fails. You can do one-liner if-else commands this way (though you have to be pretty careful about it): [mike at 3po][~]$ true && echo true || echo false true [mike at 3po][~]$ false && echo true || echo false false -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/d1ca1055/attachment-0001.pgp From tlunde at gmail.com Thu Feb 28 08:20:38 2008 From: tlunde at gmail.com (Thomas Lunde) Date: Thu, 28 Feb 2008 08:20:38 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <47BC4C0F.8010907@radkeland.org> References: <47B5F1EF.3050407@radkeland.org> <47BB979E.2000609@beitsahour.net> <47BC4C0F.8010907@radkeland.org> Message-ID: <4e291ac30802280620s6b3603b1mf69011376e68681b@mail.gmail.com> (Sorry for resurrecting an old thread...) On 2/20/08, Joshua Radke wrote: > The punch line > is that I've retired the drive for live use, and will be using it for > backups instead. I've been bitten HARD by this in the past and strongly urge you not to do this. When you need your backup, you'll really need it. Do you want to find out at that point that it was a bad drive and you didn't _really_ have a backup? My $0.02 thomas From hick0088 at tc.umn.edu Thu Feb 28 08:48:45 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 08:48:45 -0600 Subject: [tclug-list] SSH's "Password:" prompt Message-ID: <1204210125.15788.32.camel@localhost> This is a more general Unix question. We've got a bunch of somewhat older FreeBSD boxes in a lab at work (most are running 5.4), and when I SSH to them with a username, as in ssh mhicks at hostname the prompt I get back just says "Password: ". However, if I connect to a Solaris box in our lab, the prompt comes back as "mhicks at hostname's password: ". The prompt also shows up in that manner if I have an entry in my ~/.ssh/config file like Host hostname User mhicks I want the longer prompt to show up since we use a couple of different logins on our various hosts, and I want a hint at which password I should be typing in. Does anyone know how to get these FreeBSD hosts to act like the others? Er, well, without going through the process of upgrading lots and lots of systems? -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/5554c9b0/attachment.pgp From marc at e-skinner.net Thu Feb 28 09:16:35 2008 From: marc at e-skinner.net (Marc Skinner) Date: Thu, 28 Feb 2008 09:16:35 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: Message-ID: <47C6D053.9040208@e-skinner.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 you might need to flash your wireless router/wpa device. i know my dlink only did WEP when i purchased it, but after 2 firmware flashes i was able to add WPA and WPA2 respectively over the course of about a year. Chuck Cole wrote: > >> -----Original Message----- >> From: tclug-list-bounces at mn-linux.org >> [mailto:tclug-list-bounces at mn-linux.org]On Behalf Of Yaron >> >> >> I recently put Linux on my old Powerbook. It's pretty cool, every piece of >> hardware is detected and works, including the wireless card. But I use >> WPA2 and I have NO idea how to get that in there. The WiFi setup tool that >> comes with Yellowdog Linux asks for a "key", but apparently that's not the >> actual WPA passphrase. >> >> I've done some googling and it looks like I need to generate the key >> somehow... but I'm not too sure how. No links to key generation utilities >> actually worked. >> >> Anyone? >> >> >> -Yaron >> > > Linksys routers can do this. I copy the resulting key to a text file that I keep in a flash drive so I can merely paste in the long > key when I need it. > > > Chuck > > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) iD8DBQFHxtBTvE9HrEfeE4cRApMzAKDwz1jRVcEDpuCcJHe5UyRdnrV+DwCg04uE oxwXCSK/hdK98bIhoYEgAHQ= =NE9Q -----END PGP SIGNATURE----- From daniel.armbrust.list at gmail.com Thu Feb 28 09:30:59 2008 From: daniel.armbrust.list at gmail.com (Dan Armbrust) Date: Thu, 28 Feb 2008 09:30:59 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204208293.15788.22.camel@localhost> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> Message-ID: <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> Thanks for all the feedback - this turned out to be quite informative. So, I seem to have the following 2 methods: P1=$(\dirname "$(\which "$0")") P2=$(cd "$(\dirname "$0")"; pwd) And they both seem to work in all the cases that I'm worried about, anyway. And if I followed the thread correctly, method two is better, because it doesn't depend on any behavior from "which" - and that is good, because which is not necessarily consistent across all platforms. However, since I barely know enough shell scripting to be dangerous, can someone please explain to me why the second method works? Why doesn't the cd command leave me in a different place than where I started, if I execute the script from another folder? Does cd get executed twice? Thanks, Dan From tclug at freakzilla.com Thu Feb 28 09:43:31 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 09:43:31 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: <47C6D053.9040208@e-skinner.net> References: <47C6D053.9040208@e-skinner.net> Message-ID: On Thu, 28 Feb 2008, Marc Skinner wrote: > you might need to flash your wireless router/wpa device. > Chuck Cole wrote: >> Linksys routers can do this. I copy the resulting key to a text file that I keep in a flash drive so I can merely paste in the long >> key when I need it. I have a Linksys WRT54GL and it's setup to do WPA2. I have two Macs that talk WPA2 and that works just fine. My question is how do I get Linux to talk WPA2. I've got other OSes to, because they all let you actually say "Use WPA2", but the Linux utils just ask for a SSID and a "key". The router lets you set up a "WPA Shared Key:" in the form of a passphrase. When I use this passphrase on Linux, it does not work. Either because it wants a hex-based key, or because there's no place to tell it tat we're using WPA2 as opposed to, say, WEP. This is what I'm asking: How do I tell Linux we're using WPA2? -Yaron -- From daniel.armbrust.list at gmail.com Thu Feb 28 10:12:42 2008 From: daniel.armbrust.list at gmail.com (Dan Armbrust) Date: Thu, 28 Feb 2008 10:12:42 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: <47C6D053.9040208@e-skinner.net> Message-ID: <82f04dc40802280812j557af545tfed468033defc794@mail.gmail.com> I dunno - I just use Ubuntu - it works there :) I know in older version of Ubuntu there were still some manual steps, and they shipped a key generation tool. But I can't remember what it was called. Sorry, Dan From strayf at freeshell.org Thu Feb 28 10:13:16 2008 From: strayf at freeshell.org (Steve Cayford) Date: Thu, 28 Feb 2008 10:13:16 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> Message-ID: <47C6DD9C.50604@freeshell.org> Dan Armbrust wrote: > Thanks for all the feedback - this turned out to be quite informative. > > So, I seem to have the following 2 methods: > > P1=$(\dirname "$(\which "$0")") > > > P2=$(cd "$(\dirname "$0")"; pwd) > > And they both seem to work in all the cases that I'm worried about, > anyway. And if I followed the thread correctly, method two is better, > because it doesn't depend on any behavior from "which" - and that is > good, because which is not necessarily consistent across all > platforms. > > However, since I barely know enough shell scripting to be dangerous, > can someone please explain to me why the second method works? Why > doesn't the cd command leave me in a different place than where I > started, if I execute the script from another folder? Does cd get > executed twice? > Commands inside the $() are executed in a subshell (I'm not sure if it actually runs a separate process or just behaves like one.) Once the subshell is done executing all the environment changes get discarded. -Steve From marc at e-skinner.net Thu Feb 28 10:19:36 2008 From: marc at e-skinner.net (Marc Skinner) Date: Thu, 28 Feb 2008 10:19:36 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: <47C6D053.9040208@e-skinner.net> Message-ID: <47C6DF18.4020801@e-skinner.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 i know with fedora, centos and red hat, they all use the network manager application, which supports wep, wpa and wpa2. i'm guess your version of yellow dog doesn't have that app or it isn't installed. i would start there. Yaron wrote: > On Thu, 28 Feb 2008, Marc Skinner wrote: > >> you might need to flash your wireless router/wpa device. > >> Chuck Cole wrote: > >>> Linksys routers can do this. I copy the resulting key to a text file that I keep in a flash drive so I can merely paste in the long >>> key when I need it. > > > I have a Linksys WRT54GL and it's setup to do WPA2. I have two Macs that > talk WPA2 and that works just fine. My question is how do I get Linux to > talk WPA2. I've got other OSes to, because they all let you actually say > "Use WPA2", but the Linux utils just ask for a SSID and a "key". > > The router lets you set up a "WPA Shared Key:" in the form of a > passphrase. When I use this passphrase on Linux, it does not work. Either > because it wants a hex-based key, or because there's no place to tell it > tat we're using WPA2 as opposed to, say, WEP. > > This is what I'm asking: How do I tell Linux we're using WPA2? > > > -Yaron > > -- > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) iD8DBQFHxt8XvE9HrEfeE4cRAlvFAJ936u5SfljpwE6DBBhCG6SvjDilngCg10rh 4xXF7cvXEhfEgkE99M5U9jc= =JTl+ -----END PGP SIGNATURE----- From tclug at freakzilla.com Thu Feb 28 10:19:29 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 10:19:29 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: <82f04dc40802280812j557af545tfed468033defc794@mail.gmail.com> References: <47C6D053.9040208@e-skinner.net> <82f04dc40802280812j557af545tfed468033defc794@mail.gmail.com> Message-ID: On Thu, 28 Feb 2008, Dan Armbrust wrote: > I dunno - I just use Ubuntu - it works there :) And you know, I'd have NO problem using ubuntu on this thing, except there's no PPC version. I tried debian on the thing but that ust failed to install, so I'm on yellowdog now. -Yaron -- From cncole at earthlink.net Thu Feb 28 10:43:09 2008 From: cncole at earthlink.net (Chuck Cole) Date: Thu, 28 Feb 2008 10:43:09 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: Message-ID: > -----Original Message----- > From: tclug-list-bounces at mn-linux.org > [mailto:tclug-list-bounces at mn-linux.org]On Behalf Of Yaron > Sent: Thursday, February 28, 2008 9:44 AM > > On Thu, 28 Feb 2008, Marc Skinner wrote: > > > you might need to flash your wireless router/wpa device. > > > Chuck Cole wrote: > > >> Linksys routers can do this. I copy the resulting key to a text file that I keep in a flash drive so I can merely > paste in the long > >> key when I need it. > > > I have a Linksys WRT54GL and it's setup to do WPA2. I have two Macs that > talk WPA2 and that works just fine. My question is how do I get Linux to > talk WPA2. I've got other OSes to, because they all let you actually say > "Use WPA2", but the Linux utils just ask for a SSID and a "key". > > The router lets you set up a "WPA Shared Key:" in the form of a > passphrase. When I use this passphrase on Linux, it does not work. Either > because it wants a hex-based key, or because there's no place to tell it > tat we're using WPA2 as opposed to, say, WEP. > > This is what I'm asking: How do I tell Linux we're using WPA2? > > > -Yaron I'd guess that the Linus driver must know which encryption to use, and know to use the right one. My answer was only about key generation. I believe that the key generation algorithm is independent of the encryption choice, but that's only a guess and probably not a good one. Updating your wireless device sounds like the best and maybe the only option. FWIW, I have some Proxim Orinoco Gold PCMCIA cards and a PCI for sale that do a/g/b and supposedly have Linux drivers and WPA. These are posted on Craig's list, but haven't gotten around to posting here yet. Chuck From mbmiller at taxa.epi.umn.edu Thu Feb 28 10:51:51 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Thu, 28 Feb 2008 10:51:51 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> Message-ID: On Thu, 28 Feb 2008, Dan Armbrust wrote: > Thanks for all the feedback - this turned out to be quite informative. > > So, I seem to have the following 2 methods: > > P1=$(\dirname "$(\which "$0")") > > > P2=$(cd "$(\dirname "$0")"; pwd) > > And they both seem to work in all the cases that I'm worried about, > anyway. And if I followed the thread correctly, method two is better, > because it doesn't depend on any behavior from "which" - and that is > good, because which is not necessarily consistent across all platforms. Right. As the inventor of "P1," I would definitely recommend that you stick with "P2". Also, "P2" works even with a newline in the path name (admittedly a very weird phenomenon). Just remember to run it early in the shell script before any cd commands or else it can fail. > However, since I barely know enough shell scripting to be dangerous, can > someone please explain to me why the second method works? Why doesn't > the cd command leave me in a different place than where I started, if I > execute the script from another folder? Does cd get executed twice? There are two issues here: (1) Like Vegas, a lot of what happens in the shell script stays in the shell script. If you change your default directory in the script, it changes only in the script and not in your login shell -- so when the script is done, you will always "be" where you "were." This is because the script runs in its own shell. (2) Kathryn put the cd command within parentheses so that it would run in a subshell which means that it would not affect the default directory for the rest of the script. You changed it slightly by putting a dollar sign in front of the parenthesis -- I think that is good and it is doing just what you want because the $() enclosure also creates a subshell. Mike From tclug at freakzilla.com Thu Feb 28 10:59:11 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 10:59:11 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: References: Message-ID: On Thu, 28 Feb 2008, Chuck Cole wrote: > Updating your wireless device sounds like the best and maybe the only option. I still don't see how this could possibly be the router, considering WPA2 works just fine with other devices. I've used it on several Macs, a few Windows laptops, a Nokia N800 and an iPod touch. All of those happily let you choose WPA2 and enter the passphrase. I just don't know where you make that selection on Linux. The network config Yellowdog came with (I think it was Gnome's network manager) did not have an option. The firmware on my router is fairly uptodate. I'd try a newer one except that I'm expecting a new router tomorrow, anyway. > FWIW, I have some Proxim Orinoco Gold PCMCIA cards and a PCI for sale that do a/g/b and supposedly have Linux drivers and WPA. > These are posted on Craig's list, but haven't gotten around to posting here yet. Heh, this is on a Powerbook with no PCI or PCMCIA slots. -Yaron -- From jima at beer.tclug.org Thu Feb 28 12:04:25 2008 From: jima at beer.tclug.org (Jima) Date: Thu, 28 Feb 2008 12:04:25 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: References: <47C6D053.9040208@e-skinner.net> <82f04dc40802280812j557af545tfed468033defc794@mail.gmail.com> Message-ID: On Thu, 28 Feb 2008, Yaron wrote: > On Thu, 28 Feb 2008, Dan Armbrust wrote: >> I dunno - I just use Ubuntu - it works there :) > > And you know, I'd have NO problem using ubuntu on this thing, except > there's no PPC version. > > I tried debian on the thing but that ust failed to install, so I'm on > yellowdog now. Try Fedora? They're upstream for Yellow Dog anyway, and a release or two ahead. Jima From jeremy at rosengren.org Thu Feb 28 11:55:49 2008 From: jeremy at rosengren.org (Jeremy Rosengren) Date: Thu, 28 Feb 2008 11:55:49 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: <47C6D053.9040208@e-skinner.net> <82f04dc40802280812j557af545tfed468033defc794@mail.gmail.com> Message-ID: <47C6F5A5.7020003@rosengren.org> Yaron wrote: > On Thu, 28 Feb 2008, Dan Armbrust wrote: > > >> I dunno - I just use Ubuntu - it works there :) >> > > And you know, I'd have NO problem using ubuntu on this thing, except > there's no PPC version. > Sure there is, it's just no longer a "first-class citizen" in Ubuntu-land: http://cdimage.ubuntu.com/ports/releases/gutsy/release/ -- j -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/3770aa9d/attachment.htm From tclug at beitsahour.net Thu Feb 28 12:38:12 2008 From: tclug at beitsahour.net (Munir Nassar) Date: Thu, 28 Feb 2008 12:38:12 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: Message-ID: On 2/28/08, Yaron wrote: > I'll admit, last time I tried to do wireless on Linux WEP was barely even > around. > > I recently put Linux on my old Powerbook. It's pretty cool, every piece of > hardware is detected and works, including the wireless card. But I use > WPA2 and I have NO idea how to get that in there. The WiFi setup tool that > comes with Yellowdog Linux asks for a "key", but apparently that's not the > actual WPA passphrase. The Key you are trying to enter is probably the wepkey, iwconfig does only works with web keys, if you want to use the mode advanced security of WPA you must you the wpa_supplicant to authenticate. From tclug at freakzilla.com Thu Feb 28 13:07:51 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 13:07:51 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: <006f01c87a37$ea895d80$bf9c1880$@com> References: <006f01c87a37$ea895d80$bf9c1880$@com> Message-ID: On Thu, 28 Feb 2008, Dana Gerard wrote: > Try searching on wpa_supplicant. NOW we're talking. Thanks! You too. Munir (: -Yaron -- From josh at radkeland.org Thu Feb 28 13:21:00 2008 From: josh at radkeland.org (Joshua Radke) Date: Thu, 28 Feb 2008 13:21:00 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <4e291ac30802280620s6b3603b1mf69011376e68681b@mail.gmail.com> References: <47B5F1EF.3050407@radkeland.org> <47BB979E.2000609@beitsahour.net> <47BC4C0F.8010907@radkeland.org> <4e291ac30802280620s6b3603b1mf69011376e68681b@mail.gmail.com> Message-ID: <47C7099C.2060509@radkeland.org> Thomas Lunde wrote: > (Sorry for resurrecting an old thread...) > > On 2/20/08, Joshua Radke wrote: > >> The punch line >> is that I've retired the drive for live use, and will be using it for >> backups instead. >> > > > I've been bitten HARD by this in the past and strongly urge you not to do this. > > When you need your backup, you'll really need it. Do you want to find > out at that point that it was a bad drive and you didn't _really_ have > a backup? > > My $0.02 > thomas > Perhaps an update is in order. As it turns out, the other external disk I had started with the USB resets after it was a bit more full, so I'm chalking it up to a problem with the motherboard ... or something. Since the other disk is formatted directly as ext3 (instead of ext3 on LVM), the reset has no effect on running applications other than them suffering an approximately 10 second pause. Also, the backup in this case is less critical than in an enterprise situation (home MythTV setup). Speaking of reliability of backup media, does anybody have suggestions on verifying the integrity of external disks? Unfortunately, I cannot check SMART status on either of these disks (afaik ... I've tried, and there seems to be some problems with smartmontools on USB disks?). Does running an occasional e2fsck --force and badblocks give a good indication of disk health? From florin at iucha.net Thu Feb 28 13:41:44 2008 From: florin at iucha.net (Florin Iucha) Date: Thu, 28 Feb 2008 13:41:44 -0600 Subject: [tclug-list] USB Bus Resets In-Reply-To: <47C7099C.2060509@radkeland.org> References: <47B5F1EF.3050407@radkeland.org> <47BB979E.2000609@beitsahour.net> <47BC4C0F.8010907@radkeland.org> <4e291ac30802280620s6b3603b1mf69011376e68681b@mail.gmail.com> <47C7099C.2060509@radkeland.org> Message-ID: <20080228194144.GM3247@iris.iucha.org> On Thu, Feb 28, 2008 at 01:21:00PM -0600, Joshua Radke wrote: > Speaking of reliability of backup media, does anybody have suggestions > on verifying the integrity of external disks? Unfortunately, I cannot > check SMART status on either of these disks (afaik ... I've tried, and > there seems to be some problems with smartmontools on USB disks?). Does > running an occasional e2fsck --force and badblocks give a good > indication of disk health? You could create md5 checksums for all the files and verify them periodically. Sort of a poor man's ZFS scrubbing. Cheers, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/4d1ea121/attachment.pgp From thecubic at thecubic.net Thu Feb 28 14:01:17 2008 From: thecubic at thecubic.net (Dave Carlson) Date: Thu, 28 Feb 2008 14:01:17 -0600 Subject: [tclug-list] SSH's "Password:" prompt In-Reply-To: <1204210125.15788.32.camel@localhost> References: <1204210125.15788.32.camel@localhost> Message-ID: <200802281401.20465.thecubic@thecubic.net> On Thursday 28 February 2008 08:48:45 am Mike Hicks wrote: > I want the longer prompt to show up since we use a couple of different > logins on our various hosts, and I want a hint at which password I > should be typing in. > > Does anyone know how to get these FreeBSD hosts to act like the others? > Er, well, without going through the process of upgrading lots and lots > of systems? "user at host's password:" <- password authentication "Password:" <- keyboard-interactive authentication the default preference output (man ssh_config) is: PreferredAuthentications ?gssapi-with-mic, hostbased, publickey, keyboard-interactive, password? if you override that (in your ~/.ssh/config) to: PreferredAuthentications ?gssapi-with-mic, hostbased, publickey, password, keyboard-interactive? it will prefer the longer over the shorter. -Dave -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/fb134bbc/attachment.pgp From mbmiller at taxa.epi.umn.edu Thu Feb 28 18:13:20 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Thu, 28 Feb 2008 18:13:20 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <47C5F97B.7050702@mtu.net> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> Message-ID: I thought I was done with all this, but then I went back to Firefox to close out some tabs and took a look at this: http://www.linuxquestions.org/questions/programming-9/discovering-script-directory-path-531088/ I think I sent that URL to the list earlier. I think that web page was where I got some of my incorrect ideas, such as: For example, by issuing ../../my_script the result of "dirname $0" will be ../.. That is certainly false wherever I have tried it. Is it *ever* true or was the author just deluded? (As I was!) Then there is the claim that in bash, all external commands (e.g., dirname) can be avoided using this tactic: case $0 in */*) EXEC_DIR=${0:%/*} ;; *) EXEC_DIR=$PWD ;; esac Does that ever work? I think the problem is more than just missing quotes. It gives me this error: 0: %/*: syntax error: operand expected (error token is "%/*") Strangely, on that LinuxQuestions.org page, I didn't find anything that really worked for me despite the fact that the writers there seem to know a lot about bash scripting. Mike From tclug at freakzilla.com Thu Feb 28 18:22:33 2008 From: tclug at freakzilla.com (Yaron) Date: Thu, 28 Feb 2008 18:22:33 -0600 (CST) Subject: [tclug-list] WPA2 encryption In-Reply-To: References: Message-ID: On Thu, 28 Feb 2008, Chuck Cole wrote: > I agree. Who said "router"? I meant whatever is in your poorbook. Oh yeah. Oops (; Either way, it did WPA2 just fine under OS X > Hmmm.. no option means no option. > > Apple is good at compatibility.. ask their support :-) I'm not sure how well they suppot people putting Linux on their laptops (: I could ask Yelowdog, or just wait for the next version of that which should be freely avaiable Any Day Now. >> Heh, this is on a Powerbook with no PCI or PCMCIA slots. > Ain't compatibility wunnerful? *smiles* Iwouldn't expect a PCI slot in a laptop. The bigger Powerbooks did have PCMCIA slots, but the 12" ones did not. -Yaron -- From brockn at gmail.com Thu Feb 28 18:23:38 2008 From: brockn at gmail.com (Brock Noland) Date: Thu, 28 Feb 2008 18:23:38 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> Message-ID: <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> > Then there is the claim that in bash, all external commands (e.g., > dirname) can be avoided using this tactic: > > case $0 in > */*) EXEC_DIR=${0:%/*} ;; > *) EXEC_DIR=$PWD ;; > esac > > Does that ever work? I think the problem is more than just missing > quotes. It gives me this error: > > 0: %/*: syntax error: operand expected (error token is "%/*") Note the missing colon in the second example: [noland at a90 ~]$ ./mike.sh ./mike.sh: line 3: 0: %/*: syntax error: operand expected (error token is "%/*") EXEC_DIR = -------------- EXEC_DIR = . [noland at a90 ~]$ cat mike.sh #!/bin/bash case $0 in */*) EXEC_DIR=${0:%/*} ;; *) EXEC_DIR=$PWD ;; esac echo EXEC_DIR = $EXEC_DIR echo '--------------' case $0 in */*) EXEC_DIR=${0%/*} ;; *) EXEC_DIR=$PWD ;; esac echo EXEC_DIR = $EXEC_DIR This is using parameter substitution. http://tldp.org/LDP/abs/html/parameter-substitution.html Brock From mbmiller at taxa.epi.umn.edu Thu Feb 28 19:40:18 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Thu, 28 Feb 2008 19:40:18 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> Message-ID: On Thu, 28 Feb 2008, Brock Noland wrote: > Note the missing colon in the second example: > [noland at a90 ~]$ ./mike.sh > ./mike.sh: line 3: 0: %/*: syntax error: operand expected (error token is "%/*") > EXEC_DIR = > -------------- > EXEC_DIR = . > [noland at a90 ~]$ cat mike.sh > #!/bin/bash > case $0 in > */*) EXEC_DIR=${0:%/*} ;; > *) EXEC_DIR=$PWD ;; > esac > echo EXEC_DIR = $EXEC_DIR > echo '--------------' > case $0 in > */*) EXEC_DIR=${0%/*} ;; > *) EXEC_DIR=$PWD ;; > esac > echo EXEC_DIR = $EXEC_DIR > > This is using parameter substitution. > > http://tldp.org/LDP/abs/html/parameter-substitution.html Thanks Brock! In other words, the problem with the script was that there was a single colon where it didn't belong. I've tested the following script and it seems that this will do what we want: case "$0" in */*) EXEC_DIR="${0%/*}" ;; *) EXEC_DIR="$PWD" ;; esac EXEC_DIR="$(builtin cd "$EXEC_DIR" && builtin echo "$PWD")" builtin echo "$EXEC_DIR" I think that is probably the best we've found so far (even though I don't completely understand it!). It uses no external commands. The command "builtin" forces the command that follows to be the shell builtin rather than an alias or external command. I'm not sure that that trick is needed but I think it can't hurt. I'm not sure that it handles every kind of special character, but I have tested it with a newline and spaces in the path and it worked fine. Mike From hick0088 at tc.umn.edu Thu Feb 28 20:19:38 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 20:19:38 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> Message-ID: <1204251578.15788.36.camel@localhost> On Thu, 2008-02-28 at 19:40 -0600, Mike Miller wrote: > Thanks Brock! In other words, the problem with the script was that there > was a single colon where it didn't belong. I've tested the following > script and it seems that this will do what we want: > > case "$0" in > */*) EXEC_DIR="${0%/*}" ;; > *) EXEC_DIR="$PWD" ;; > esac > > EXEC_DIR="$(builtin cd "$EXEC_DIR" && builtin echo "$PWD")" > builtin echo "$EXEC_DIR" I think we've shown through previous discussions of how $0 works that the second option in that case statement would never get hit (meaning, there's always a '/' in there somewhere) Couldn't you take that down to EXEC_DIR="$(builtin cd "${0%/*}" && builtin echo "$PWD")" -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/245af775/attachment.pgp From hick0088 at tc.umn.edu Thu Feb 28 20:49:39 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 20:49:39 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204251578.15788.36.camel@localhost> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> Message-ID: <1204253379.15788.49.camel@localhost> On Thu, 2008-02-28 at 20:19 -0600, Mike Hicks wrote: > EXEC_DIR="$(builtin cd "${0%/*}" && builtin echo "$PWD")" The only problem I've found so far is with a directory with a '-' as the first character. 'cd -' changes to $OLDPWD. 'cd -P' and 'cd -L' handle following symlinks in different ways (I believe 'cd -P' is the default). But, in order to even execute when the first character in $0 will be a dash, you have to start the script with an extra '--', as in: #!/bin/bash -- If you want to check for this, then I'd suggest: EXEC_DIR="$(builtin cd -P -- "${0%/*}" && builtin echo "$PWD")" Similarly, you *might* have to pass '-E' to echo to disable some escape sequence handling, but I'm not quite sure if that's necessary. Bash's builtin echo won't ignore '--', but we don't need to worry about accidentally passing a parameter to echo since the first character in $PWD is always going to be a slash. I think... -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/c818a5e4/attachment.pgp From hick0088 at tc.umn.edu Thu Feb 28 22:38:40 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 22:38:40 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204253379.15788.49.camel@localhost> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> Message-ID: <1204259921.15788.54.camel@localhost> On Thu, 2008-02-28 at 20:49 -0600, Mike Hicks wrote: > EXEC_DIR="$(builtin cd -P -- "${0%/*}" && builtin echo "$PWD")" Okay one last thing, and I think I'm done :-p Turns out quotes aren't necessary around the ${0%/*}, and 'pwd' is also a builtin, so: EXEC_DIR="$(builtin cd -P -- ${0%/*} && builtin pwd) -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/2de58b1c/attachment.pgp From hick0088 at tc.umn.edu Thu Feb 28 22:46:06 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Thu, 28 Feb 2008 22:46:06 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204259921.15788.54.camel@localhost> References: <82f04dc40802271413xd255042s5ebd402f842f149b@mail.gmail.com> <30983.192.168.0.7.1204161993.squirrel@www.flyballdogs.com> <1204171807.15788.17.camel@localhost> <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> <1204259921.15788.54.camel@localhost> Message-ID: <1204260366.15788.57.camel@localhost> Nope, I was wrong, you need double quotes. But, with this, I'm definitely done: #!/bin/bash -- EXEC_DIR=$(builtin cd -P -- "${0%/*}" && builtin pwd) echo "$EXEC_DIR" -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/1f81533f/attachment.pgp From florin at iucha.net Thu Feb 28 22:48:23 2008 From: florin at iucha.net (Florin Iucha) Date: Thu, 28 Feb 2008 22:48:23 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204259921.15788.54.camel@localhost> References: <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> <1204259921.15788.54.camel@localhost> Message-ID: <20080229044823.GQ3247@iris.iucha.org> On Thu, Feb 28, 2008 at 10:38:40PM -0600, Mike Hicks wrote: > On Thu, 2008-02-28 at 20:49 -0600, Mike Hicks wrote: > > EXEC_DIR="$(builtin cd -P -- "${0%/*}" && builtin echo "$PWD")" > > Okay one last thing, and I think I'm done :-p > > Turns out quotes aren't necessary around the ${0%/*}, and 'pwd' is also > a builtin, so: > > EXEC_DIR="$(builtin cd -P -- ${0%/*} && builtin pwd) Now, can a kind soul wait a couple more days to see if the horse is well done or still twitching, then write up a concise and correct reply (preferably to the original message) so it is archived for posterity? Thanks, florin -- Bruce Schneier expects the Spanish Inquisition. http://geekz.co.uk/schneierfacts/fact/163 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080228/f1957830/attachment.pgp From mbmiller at taxa.epi.umn.edu Fri Feb 29 01:33:54 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 01:33:54 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <20080229044823.GQ3247@iris.iucha.org> References: <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> <1204259921.15788.54.camel@localhost> <20080229044823.GQ3247@iris.iucha.org> Message-ID: On Thu, 28 Feb 2008, Florin Iucha wrote: > Now, can a kind soul wait a couple more days to see if the horse is well > done or still twitching, then write up a concise and correct reply > (preferably to the original message) so it is archived for posterity? Maybe we should continue to include the latest and greatest in every reply: #!/bin/bash -- EXEC_DIR=$(builtin cd -P -- "${0%/*}" && builtin pwd) echo "$EXEC_DIR" I think this thread will be a great learning experience for someone who wants to understand what the issues are. There is a lot in these messages. It's nice to have an answer to the original question, but how that answer was arrived at is also important. If you search the web for the answer, you will find lots of answers, many of which don't work at all. Other suggested answers will fail under common conditions. Some are just not optimal. Regarding the above script, is the "--" needed after "bash"? I don't think it is because there is nothing else on that line. Mike From hick0088 at tc.umn.edu Fri Feb 29 01:54:22 2008 From: hick0088 at tc.umn.edu (Mike Hicks) Date: Fri, 29 Feb 2008 01:54:22 -0600 Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: References: <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> <1204259921.15788.54.camel@localhost> <20080229044823.GQ3247@iris.iucha.org> Message-ID: <1204271662.15788.66.camel@localhost> On Fri, 2008-02-29 at 01:33 -0600, Mike Miller wrote: > #!/bin/bash -- > EXEC_DIR=$(builtin cd -P -- "${0%/*}" && builtin pwd) > echo "$EXEC_DIR" [...] > Regarding the above script, is the "--" needed after "bash"? I don't > think it is because there is nothing else on that line. The '--' is there for the outside case that you've executed the script from a directory that starts with a dash. If I create a directory '-P' and put my script in it, I could see this: [mike at 3po][~]$ -P/execdir /bin/bash: -/: invalid option Usage: /bin/bash [GNU long option] [option] ... /bin/bash [GNU long option] [option] script-file ... GNU long options: --debug --debugger [...] My command had been expanded to be 'bash -P/execdir'. If you add the two dashes at the top of the script, the actual command executed will instead be 'bash -- -P/execdir'. Of course, having a directory with a dash as the first character is really unlikely, and having actual executables in that directory is a few orders of magnitude even less likely. It was basically an exercise in making the script as bombproof as possible. -- Mike Hicks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part Url : http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/f2caf0d9/attachment.pgp From mbmiller at taxa.epi.umn.edu Fri Feb 29 08:38:50 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 08:38:50 -0600 (CST) Subject: [tclug-list] shell script - find shell script parent directory? In-Reply-To: <1204271662.15788.66.camel@localhost> References: <47C6B58B.1010200@mtu.net> <1204208293.15788.22.camel@localhost> <82f04dc40802280730m3acdce06pef9f223b583ce83b@mail.gmail.com> <741dcbb80802281623s17e084d6l2741cf8ce2c0c71a@mail.gmail.com> <1204251578.15788.36.camel@localhost> <1204253379.15788.49.camel@localhost> <1204259921.15788.54.camel@localhost> <20080229044823.GQ3247@iris.iucha.org> <1204271662.15788.66.camel@localhost> Message-ID: On Fri, 29 Feb 2008, Mike Hicks wrote: > On Fri, 2008-02-29 at 01:33 -0600, Mike Miller wrote: >> #!/bin/bash -- >> EXEC_DIR=$(builtin cd -P -- "${0%/*}" && builtin pwd) >> echo "$EXEC_DIR" > > [...] > >> Regarding the above script, is the "--" needed after "bash"? I don't >> think it is because there is nothing else on that line. > > The '--' is there for the outside case that you've executed the script > from a directory that starts with a dash. > > If I create a directory '-P' and put my script in it, I could see this: > > [mike at 3po][~]$ -P/execdir > /bin/bash: -/: invalid option > Usage: /bin/bash [GNU long option] [option] ... > /bin/bash [GNU long option] [option] script-file ... > GNU long options: > --debug > --debugger > [...] > > My command had been expanded to be 'bash -P/execdir'. If you add the > two dashes at the top of the script, the actual command executed will > instead be 'bash -- -P/execdir'. That is very interesting. I did not know that was how it worked. I guess the same argument could apply to any script -- maybe they all should have two dashes after bash on the first line. On the other hand, you can type this on the command line to execute the script: bash -- -P/execdir That's good to know because think of how many scripts you have written without the double hyphen on line one! Luckily, we don't name many directories that way, so it probably won't be needed often. Mike From chewie at wookimus.net Fri Feb 29 10:11:44 2008 From: chewie at wookimus.net (Chad Walstrom) Date: Fri, 29 Feb 2008 10:11:44 -0600 Subject: [tclug-list] SSH's "Password:" prompt In-Reply-To: <1204210125.15788.32.camel@localhost> References: <1204210125.15788.32.camel@localhost> Message-ID: <3247.1204301504@skuld.wookimus.net> This isn't an answer to your question, rather a question itself. Why not use ssh keys? Just concatenate your id_dsa.pub/id_rsa.pub/identity.pub file(s) into a file called authorized_keys. Place that in your ~/.ssh directory on each host. Then fire up ssh-agent, type ssh-add, and enter your password for your private key. Now, you won't have to type your password in whenever you go to a host that has your authorized_keys file. It makes life SOOOO much simpler. ;-) See also: ssh(1), ssh-keygen(1), ssh-agent(1), ssh-add(1). I plan on home-cooking something with cfengine2 to push out key files to a /etc/ssh/authorized_keys.d directory and changing this directive in sshd_config: AuthorizedKeysFile /etc/ssh/authorized_keys.d/%u.keys That way, users don't need home directories to log in, and I've got a central location in which to manage access. This works well in a corporate environment where sysadmins/engineers are in charge of system security. We allow people to update their own authorized_keys files on a primary shell server, then copy them to a central location (i.e. cfengine master server/CVS repository) to be distributed to individual hosts. There are other security tie-ins you can use. For example, if you're using some form of Kerberos, there is a Kerberized SSH daemon you can install, bypassing the need for password authentication or ssh keys entirely. Chad From slushpupie at gmail.com Fri Feb 29 10:16:35 2008 From: slushpupie at gmail.com (Jay Kline) Date: Fri, 29 Feb 2008 10:16:35 -0600 Subject: [tclug-list] WPA2 encryption In-Reply-To: References: Message-ID: <47C82FE3.7090705@gmail.com> Yaron wrote: > I recently put Linux on my old Powerbook. It's pretty cool, every piece of > hardware is detected and works, including the wireless card. Ive been running ubuntu on mine, and I cant get the wireless to work correctly. Which card do you have and which kernel are you running? Ive got 2.6.22 (2.6.22-14-powerpc for the ubuntu-specific versions) and the BCM4306 (I think that is labeled "airport extreme" or something). Might be time for a kernel upgrade if you have it working on a newer kernel. Jay From dalan at visi.com Fri Feb 29 12:16:21 2008 From: dalan at visi.com (Don Sparish) Date: Fri, 29 Feb 2008 12:16:21 -0600 Subject: [tclug-list] exact match using grep In-Reply-To: <47C82FE3.7090705@gmail.com> References: <47C82FE3.7090705@gmail.com> Message-ID: An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/248d29af/attachment.htm From samir.nassar+tclug at steamedpenguin.com Fri Feb 29 12:30:52 2008 From: samir.nassar+tclug at steamedpenguin.com (Samir M. Nassar) Date: Fri, 29 Feb 2008 12:30:52 -0600 Subject: [tclug-list] exact match using grep In-Reply-To: References: <47C82FE3.7090705@gmail.com> Message-ID: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> On Friday 29 February 2008 12:16:21 Don Sparish wrote: Many brackets, unformatted Every time you send HTML mail the gods kill a small kitten and make more Caffeine-Free Diet Pepsi [1]. Samir M. Nassar 1- http://shakespearessister.blogspot.com/2008/02/caffeine-free-diet-pepsi-is-most.html From john.meier at gmail.com Fri Feb 29 12:37:21 2008 From: john.meier at gmail.com (John Meier) Date: Fri, 29 Feb 2008 12:37:21 -0600 Subject: [tclug-list] exact match using grep In-Reply-To: References: <47C82FE3.7090705@gmail.com> Message-ID: <65293fcc0802291037j2c420a2fr17b856e5ea56f77a@mail.gmail.com> On Fri, Feb 29, 2008 at 12:16 PM, Don Sparish wrote: > Is Solaris Ass Backward or what. > > I can get an exact string match in AIX, Linux, HPUX but not in SlowLaris. > -w option maybe??? > > > > _______________________________________________ > TCLUG Mailing List - Minneapolis/St. Paul, Minnesota > tclug-list at mn-linux.org > http://mailman.mn-linux.org/mailman/listinfo/tclug-list > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/48a98b09/attachment.htm From dalan at visi.com Fri Feb 29 13:11:30 2008 From: dalan at visi.com (Don Sparish) Date: Fri, 29 Feb 2008 13:11:30 -0600 Subject: [tclug-list] exact match using grep In-Reply-To: <65293fcc0802291037j2c420a2fr17b856e5ea56f77a@mail.gmail.com> References: <47C82FE3.7090705@gmail.com> <65293fcc0802291037j2c420a2fr17b856e5ea56f77a@mail.gmail.com> Message-ID: An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/c87674be/attachment.htm From bdunnette at gmail.com Fri Feb 29 13:41:24 2008 From: bdunnette at gmail.com (Brian Dunnette) Date: Fri, 29 Feb 2008 13:41:24 -0600 Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> References: <47C82FE3.7090705@gmail.com> <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> Message-ID: <47C85FE4.5080908@gmail.com> > Every time you send HTML mail the gods kill a small kitten and make more > Caffeine-Free Diet Pepsi [1]. > Every time you use a mail client that doesn't gracefully deal with HTML, another Ron Paul story appears on Reddit. -Brian D. From admin at lctn.org Fri Feb 29 12:42:54 2008 From: admin at lctn.org (Raymond Norton) Date: Fri, 29 Feb 2008 12:42:54 -0600 Subject: [tclug-list] Mail archiving follow up Message-ID: <47C8522E.4010509@lctn.org> I made a couple posts in the recent past on mail archiving. Postfix was pretty simple, but sendmail was more evasive. I have it working now, and wrote a simple howto, based on "copyuser". I can email it or post it for whoever is interested. We are using Mailarchiva (free version) to archive all mail on the server now. Raymond From samir.nassar+tclug at steamedpenguin.com Fri Feb 29 14:12:51 2008 From: samir.nassar+tclug at steamedpenguin.com (Samir M. Nassar) Date: Fri, 29 Feb 2008 14:12:51 -0600 Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <47C85FE4.5080908@gmail.com> References: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> <47C85FE4.5080908@gmail.com> Message-ID: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> On Friday 29 February 2008 13:41:24 Brian Dunnette wrote: > Every time you use a mail client that doesn't gracefully deal with HTML, > another Ron Paul story appears on Reddit. My mail client deals just fine with HTML. I don't have my mail client set to automagically parse HTML. I also make sure that when I use a mail client that I don't send out HTML mail. Mail clients read and display text just fine. They even have decent conventions for quoting mail. Attached are html.txt and message.txt message.txt weighs in at 553 bytes html.txt weighs in at 256 bytes If the formatting had added anything of worth, fine. But in this case the text was not marked up. You know, the Markup in Hyper-text Markup Language? Why would you need to markup anyway? Since you are throwing around graceful clients, how graceful is a mail client when it can parse HTML but can't turn a lonely URL into a link? If you notice, the HTML itself sucks sweaty goat balls, all repetitive break tags. If you are sending HTML mail, why again is the mail client using font declarations? Lastly it is a pretty long standing convention to not abuse mailing lists by sending HTML mail. Samir M. Nassar PS- Thank you for changing the subject line, something I neglected to do. -------------- next part --------------






   &aps



&aps

&aps<>&aps &aps

&aps











 
-------------- next part -------------- Is Solaris Ass Backward or what. I can get an exact string match in AIX, Linux, HPUX but not in SlowLaris. What I have tried so far, grep -x returns a usages messages /usr/xpg4/bin/grep -x doesnt work grep -e or -E returns a usages message /user/xpg4/bin/grep -e or -E doesnt work grep \string\ doesnt work grep ^"string"$ doesnt work. What I am trying to do is search an environment variable such as PATH to see if a directory already exist in the variable. Anybody figured out how to find exact matches within a string on SlowLaris? Thanks Don S. From dalan at visi.com Fri Feb 29 14:43:20 2008 From: dalan at visi.com (Don Sparish) Date: Fri, 29 Feb 2008 14:43:20 -0600 Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> References: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> <47C85FE4.5080908@gmail.com> <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> Message-ID: An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/5d6cfb6e/attachment.htm From mbmiller at taxa.epi.umn.edu Fri Feb 29 14:44:31 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 14:44:31 -0600 (CST) Subject: [tclug-list] exact match using grep In-Reply-To: References: <47C82FE3.7090705@gmail.com> Message-ID: On Fri, 29 Feb 2008, Don Sparish wrote: > Is Solaris Ass Backward or what. I've put all GNU utils on my Solaris box. I'll be retiring it soon. > I can get an exact string match in AIX, Linux, HPUX but not in SlowLaris. > > What I have tried so far, > > grep -x? returns a usages messages > /usr/xpg4/bin/grep -x??? doesn't work I don't know why that is happening to you. On my Solaris 8 box, the old /usr/xpg4/bin/grep is there and working and it accepts -x. Man page: man -M /usr/man grep That includes info for both the xpg4 and /usr/bin versions. The xpg4 version can be used in a few ways: /usr/xpg4/bin/grep [ -E | -F ] [ -c | -l | -q ] [ -bhinsvwx ] -e pattern_list ... [ -f pattern_file ] ... [ file ... ] /usr/xpg4/bin/grep [ -E | -F ] [ -c | -l | -q ] [ -bhinsvwx ] [ -e pattern_list ... ] -f pattern_file ... [ file ... ] /usr/xpg4/bin/grep [ -E | -F ] [ -c | -l | -q ] [ -bhinsvwx ] pattern [ file ... ] > grep -e or -E returns a usages message > > /user/xpg4/bin/grep -e or -E doesn't work Not because of misspelling "user"? It works for me. > grep '\'? doesn't work > > grep ^"string"$ doesn't work. > > What I am trying to do is search an environment variable such as PATH to > see if a directory already exist in the variable. > > Anybody figured out how to find exact matches within a string on > SlowLaris? This works for me on Solaris 8 SPARC from the command line in both bash and tcsh: echo $PATH | tr ':' '\n' | /usr/xpg4/bin/grep -x '/usr/openwin/bin' That works here, so I'm thinking your /usr/xpg4/bin/grep was replaced by something nonstandard. Check into it. Make sure your system wasn't attacked and altered. If you are using SPARC and want to try my grep file, I can send it to you. Mike From bbaptist at iexposure.com Fri Feb 29 14:54:52 2008 From: bbaptist at iexposure.com (Bret Baptist) Date: Fri, 29 Feb 2008 14:54:52 -0600 Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: References: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> Message-ID: <200802291454.52733.bbaptist@iexposure.com> That is very nice of you and shows great mailing list etiquette. Bret. On Friday 29 February 2008 2:43:20 pm Don Sparish wrote: > Since it seems to annoy him so much I'll just keep on using it.? > > > On Fri, 29 Feb 2008 14:12:51 -0600 > > "Samir M. Nassar" wrote: > > On Friday 29 February 2008 13:41:24 Brian Dunnette > > > >wrote: > >> Every time you use a mail client that doesn't gracefully > >>deal with HTML, > >> another Ron Paul story appears on Reddit. > > > > My mail client deals just fine with HTML. I don't have > >my mail client set to > > automagically parse HTML. I also make sure that when I > >use a mail client that > > I don't send out HTML mail. Mail clients read and > >display text just fine. > > They even have decent conventions for quoting mail. > > > > Attached are html.txt and message.txt > > > > message.txt weighs in at 553 bytes > > html.txt weighs in at 256 bytes > > > > If the formatting had added anything of worth, fine. But > >in this case the text > > was not marked up. You know, the Markup in Hyper-text > >Markup Language? Why > > would you need to markup anyway? Since you are throwing > >around graceful > > clients, how graceful is a mail client when it can parse > >HTML but can't turn > > a lonely URL into a link? > > > > If you notice, the HTML itself sucks sweaty goat balls, > >all repetitive break > > tags. If you are sending HTML mail, why again is the > >mail client using font > > declarations? > > > > Lastly it is a pretty long standing convention to not > >abuse mailing lists by > > sending HTML mail. > > > > Samir M. Nassar > > > > PS- Thank you for changing the subject line, something I > >neglected to do. -- Bret Baptist Senior Network Administrator bbaptist at iexposure.com Internet Exposure, Inc. http://www.iexposure.com (612)676-1946 x17 Providing Internet Services since 1995 Web Development ~ Search Engine Marketing ~ Web Analytics Network Security ~ On Demand Tech Support ~ E-Mail Marketing ------------------------------------------ From mbmiller at taxa.epi.umn.edu Fri Feb 29 14:58:51 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 14:58:51 -0600 (CST) Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> References: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> <47C85FE4.5080908@gmail.com> <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> Message-ID: On Fri, 29 Feb 2008, Samir M. Nassar wrote: > If the formatting had added anything of worth, fine. But in this case > the text was not marked up. You know, the Markup in Hyper-text Markup > Language? Why would you need to markup anyway? HTML attachments are a constant bother -- everybody is using them. It must be hard to find a mailer these days that doesn't do HTML by default. The users are mostly completely clueless. I often receive messages that use up 50 KB of space but would have been less than 10% the size with no loss of information if formatted as plain text. > If you notice, the HTML itself sucks sweaty goat balls, all repetitive > break tags. If you are sending HTML mail, why again is the mail client > using font declarations? Microsoft is a great offender. They try to generate HTML that can be automagically converted to Word format with no loss of formatting info -- all margins, fonts, etc., are to be maintained. It's a total waste of bytes in about 95% of cases and it is usually a lot of bytes. > Lastly it is a pretty long standing convention to not abuse mailing > lists by sending HTML mail. Wouldn't it be great if people would do what we wanted them to do! Unfortunately, it isn't in their interest to care about this issue. If the system could be configured to send an automatic reply saying: WARNING: You have sent an HTML attachment to the email list. We will hold your message in queue for 15 minutes in case you wish to send a version of this message without the attachment. If we receive a very similar text-only message in the next 15 minutes, we will forward that message to the list immediately and we will return your original message to you without delivering it to the list. Many readers will not enjoy receiving an HTML attachment instead of a plain text message. (It could go on to explain how to send plain text in a few of the common MUAs.) That would give a little incentive to the habitual HTML users. Mike From dalan at visi.com Fri Feb 29 15:02:00 2008 From: dalan at visi.com (Don Sparish) Date: Fri, 29 Feb 2008 15:02:00 -0600 Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <200802291454.52733.bbaptist@iexposure.com> References: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> <200802291454.52733.bbaptist@iexposure.com> Message-ID: An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/346c8766/attachment.htm From mbmiller at taxa.epi.umn.edu Fri Feb 29 15:14:38 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 15:14:38 -0600 (CST) Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: <200802291454.52733.bbaptist@iexposure.com> References: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> <200802291454.52733.bbaptist@iexposure.com> Message-ID: On Fri, 29 Feb 2008, Bret Baptist wrote: > That is very nice of you and shows great mailing list etiquette. Another really cool thing that people can do for us is to retain only the parts of the previous posting that are relevant to the reply. That helps the reader to quickly understand the point of the message. It also helps later when searching archives -- you won't see a lot of extra, useless hits. Mike From cschumann at twp-llc.com Fri Feb 29 15:30:07 2008 From: cschumann at twp-llc.com (Chris Schumann) Date: Fri, 29 Feb 2008 15:30:07 -0600 (CST) Subject: [tclug-list] exact match using grep In-Reply-To: References: Message-ID: <27101.192.28.2.17.1204320607.squirrel@alpha.twp-llc.com> Those of us who get the digest get this: > Message: 1 > Date: Fri, 29 Feb 2008 12:16:21 -0600 > From: "Don Sparish" > Subject: [tclug-list] exact match using grep > To: > Message-ID: > Content-Type: text/plain; charset="us-ascii" > > An HTML attachment was scrubbed... > URL: > http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/248d29af/attachment-0001.htm So it's not just the client. It's mailman too. Chris From mbmiller at taxa.epi.umn.edu Fri Feb 29 15:28:00 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 15:28:00 -0600 (CST) Subject: [tclug-list] HTML mail (was: exact match using grep) In-Reply-To: References: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> <200802291454.52733.bbaptist@iexposure.com> Message-ID: On Fri, 29 Feb 2008, Don Sparish wrote: > Since Im using a web based mail client provided by the ISP. I have no > way to shut it off anyway, etiquette has nothing to do with it. Allow me to recommend Gmail. It truly is a nice system and it is especially good for storing list traffic -- you can filter incoming messages, add a label appropriate to the list, skip the inbox if you wish, find all list messages with a single click (on the "label" on the left side) and the threads are automagically accumulated as "discussions." In addition, Gmail allows you to choose between "rich formatting" (which sends HTML) and "plain text" (which does not send HTML). Whatever email format you chose last is the default the next time you compose a message. Later, if you change ISP, Gmail is still working. You can access Gmail by web, IMAP or POP3. I use the web interface for cell phones to get to it from my Treo device. I can access Gmail from Pine via IMAP on my Solaris box and my sister accesses it from Outlook via IMAP on her laptop. Gmail -- it's all good. (TM) Mike From bdunnette at gmail.com Fri Feb 29 15:34:46 2008 From: bdunnette at gmail.com (Brian Dunnette) Date: Fri, 29 Feb 2008 15:34:46 -0600 Subject: [tclug-list] HTML mail In-Reply-To: <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> References: <200802291230.52810.samir.nassar+tclug@steamedpenguin.com> <47C85FE4.5080908@gmail.com> <200802291412.51926.samir.nassar+tclug@steamedpenguin.com> Message-ID: <47C87A76.2060102@gmail.com> > If the formatting had added anything of worth, fine. But in this case the text > was not marked up. You know, the Markup in Hyper-text Markup Language? Why > would you need to markup anyway? Since you are throwing around graceful > clients, how graceful is a mail client when it can parse HTML but can't turn > a lonely URL into a link? > Point taken -- a /good/ mail client should be able to differentiate between a message that actually /uses/ markup (which would be sent as HTML) and a truly plain-text message (which would automagically be sent as plain text, just to simplify matters.) > Lastly it is a pretty long standing convention to not abuse mailing lists by > sending HTML mail. > For many years, it was also convention for all computer games to require a text parser, and/or represent players with an "@" symbol. Conventions /change/. -Brian D. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/91872a27/attachment.htm From mbmiller at taxa.epi.umn.edu Fri Feb 29 15:36:53 2008 From: mbmiller at taxa.epi.umn.edu (Mike Miller) Date: Fri, 29 Feb 2008 15:36:53 -0600 (CST) Subject: [tclug-list] exact match using grep In-Reply-To: <27101.192.28.2.17.1204320607.squirrel@alpha.twp-llc.com> References: <27101.192.28.2.17.1204320607.squirrel@alpha.twp-llc.com> Message-ID: On Fri, 29 Feb 2008, Chris Schumann wrote: > Those of us who get the digest get this: > >> Message: 1 >> Date: Fri, 29 Feb 2008 12:16:21 -0600 >> From: "Don Sparish" >> Subject: [tclug-list] exact match using grep >> To: >> Message-ID: >> Content-Type: text/plain; charset="us-ascii" >> >> An HTML attachment was scrubbed... >> URL: >> http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20080229/248d29af/attachment-0001.htm > > So it's not just the client. It's mailman too. Wow. That is crazy. Why does mailman abuse the HTML that way? So the HTML starts out like this...