>The main archive site tends to stop responding when a stable
>kernel is released. Mirror sites are ineffective because they
>are not updated promptly and are hard to choose. Most people
>don't learn to use patches and bzip2 files. Some ideas:
Yes. I've noticed this. Linux users are VERY impatient when it comes to
new kernels... even the development kernels. I am one of the chanops in
EFnet #Linux. It is EXTREMELY annoying to have impatient lusers come into
the channel every 10 seconds asking:
a) WHERE DO I GET THE NEWEST KERNEL!!?!? (Even though it's in the /topic)
Not much we can do about these people but /kick them. ;)
b) ftp.kernel.org is too slow? What's another site that has the newest
kernel [that came out 2 hours ago]?
This one's my pet peeve.
>3. Mirror sites get a push feed. Mirror sites that do not respond
> to the push feed are considered dead and do not get redirects.
> The main site does not make files directly available until
> after it has tried to feed all mirror sites.
Like Mr. Cahalan wrote, wouldn't it make sense to have kernel.org push
the patches to a few mirrors before making it public? It would help ease
the load on kernel.org.
Also, wouldn't it be prudent to list some mirrors on a "too many anon
users" error? Especially a list of mirrors that have already been pushed.
Or, in the least, put a list of 5 good mirrors in kernel.org's finger...
>1. Files are made available in this order:
> patch.bz2 tarball.bz2 patch.gz tarball.gz
> There would be an intentional delay of 1 day between each file
> release for development kernels, and 3 days for stable kernels.
> This would strongly encourage use of the bzip2 patch files.
> Non-http access may get an additional delay, to support idea #2.
Excellent idea!!!
-Sir Woody Hackswell balint@udayton.edu woody@woodynet.siscom.net
(The ArchFool) http://woodynet.siscom.net/woody
Sordo of Chui believed the Song, and destroyed the two boys of
Inoperativi. If for the faith of this lie, so that it corresponds to be,
it does not ask you if the man puts in too much, because He saw it.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html