more squid

[Chewie <chewie@wookimus.net>]

NOTE: These were sent yesturday, but since the email server I use was 
misconfigured and since Ben somehow lost my secondary email address on the list
server (chewie@nerp.net...hint hint...we set this up a while ago), it didn't go through..

On Wed, Mar 15, 2000 at 06:13:21AM -0600, Eric M. Hopper wrote:
> 	*grin* One potential problem with this is that this makes it
> even more likely that the music industry will go after the U of MN.
> After all, copyrighted material will now be cached on their servers.

I actually was talking about caching in general.  Napster, web surfing, ftp
downloads.  It doesn't matter really.  I did realize one problem with this,
however.  Napster incorporates a set of central servers to exchange IP
addresses and song title/location information between users.  Once the IP
address of a "song-pusher" is identified, the Napster client connects directly
to that machine.  The songs are not downloaded from one server in particular.
Squid, IIRC, stores files in relation to the website (IP) it gleaned the
information from.  It will not be smart enough to realize that 10 Napster
"song-pushers" have the same song and only cache one copy for all ten source
targets.

So, with Napster, the proxy idea is really a bust, unless multiple people are
accessing the cache server to leach from the same "song-pusher".  With as many
people that use Napster, this wouldn't help.  In effect, it would simply take
up space on the cache disks with no where to go.

-- 
^chewie

PGP signature