Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: (ASCEND) ascend software releases
At 06:18 PM 8/21/97 -0600, Marc Slemko wrote:
>Does anyone know what version of the k56 code is in 5.0Ai13?
0.526
>I can't find
>anything in the release notes except for the funny bit saying the hardware
>supports v.fc; not the software, I assume, just the hardware.
Actually at the time the doc was written, I believe we did think V.FC was
going to be in the Rockwell code for the K56 modems...
>The docs for 5.0p20 are also not there; they end at p19 and the p19
>changes don't include most of the p20 stuff.
The p20 stuff is all there. There were no "enhancements" added with p20,
that's
why the "enhancements" section ends at P19.
>Why has Ascend started making their docs impossible to use? If I want to
>know the difference between 5.0Ap6 and p16, I can't just print the notes
>for each revision between them.
We used to do that, and got more complaints - "why don't you guys have
cumulative release notes"....so we did!
>Not only do they insist on putting the
>notes for all the patch releases in the file, but now they have gone one
>step further and are mixing them together! The list of problems fixed is
>sorted by TR number, so to find what is fixed in a release you have to
>hunt through the whole list.
You can look at the list that I always post on this list when a release
is made. That is a simple list of what was fixed and enhancements added...
I'll also pass on this observation to the tech-pubs people, and maybe we
can have an index that does what you're asking for...
> If someone people want the full notes, that
>is fine; just also make the individual release notes available. We like
>hard copies of such things, and it is stupid to print out all the 5.0Ap
>notes every time a new release comes out.
OK, we'll look into it (I surrender, we can't win!)....
>It is also annoying that Ascend changes the release notes, removes
>versions from them and merges those changes into other versions after the
>fact, etc. on a regular basis. It shouldn't be that complicated to make
>the docs right the first time. Sure, mistakes are made but they are being
>made far too often.
Let me know when, and I'll pass it on. We can only improve things when
we know they are wrong.
>How about 5.0Ai3? We have one copy of release notes printed showing
>changes in it, another set printed showing that it doesn't exist and those
>changes were really Ai4.
Version Ai3 was *never* released. If you got release notes, then you got
them on a "special" agreement with TAC/Engineering. We had it on a closed
limited distribution, due to some significant frame-relay problems. Those
issues (and more) were fixed in Ai4 - which was subsequently released. The
fixes that went into Ai3 would not have appeared in the list so we added
them to the Ai4 release notes (they are in Ai4 too).
>It is nearly impossible to decide on a good release when half the notes
>don't include all the changes (eg. incremental releases sometimes include
>fixes from patch releases but they aren't included in the release notes),
>some of them are wrong (ie. obviously aren't even _looked_ at before
>release), and the actual release notes keep being updated for old
>versions.
Please detail any issues you have so that we can address them (I'll pass them
on to the Manager of Tech Pubs), since posting here will not have *that*
desired effect.
>All we are trying to do is find a good stable revision to upgrade to for
>our 56k modems before we have to ship the old ones back.
Ask here on the list, who is using what and their configuration....
>We are using
>5.0Ai8 and find it works well for us. We would upgrade to the latest Ai
>release, except it looks like it may have really old (in terms of k56flex)
>modem code and what version it has isn't documented anywhere I can find.
That is an error I think, I'll ask for something to be put in the notes
and/or on the FTP README files...
>We would use 5.0Ap20 but it is broken. 5.0Ap16 is fine except for the
>p20 note that multiple tcp-clear calls can cause p16 to reboot; we do have
>multiple tcp-clear sessions at once most of the time. Earlier 5.0Ap
>revisions may be fine, but since Ascend doesn't have a decent bug database
>available we have no idea when this bug was introduced, how frequent it
>is, etc. All too often the descriptions in the release notes of bugs are
>just wrong.
>
>We have long since abandoned OSPF because it just doesn't work. Sure, you
>can make it work. But if you look at it wrong it will suddenly start
>rebooting.
>
>Not impressed with Ascend's software releases. Very not impressed.
>
>++ Ascend Users Mailing List ++
>To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
>To get FAQ'd: <http://www.nealis.net/ascend/faq>
>
Kevin
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>
References: