Numbering conventions

Howard Owen hbo at egbok.com
Tue Mar 29 01:08:11 UTC 2005


Unless you want to dig in and understand what patches are being applied
in detail, I'd really recommend against doing that. I'd stick with one
or the other for consistency's sake. It's possible that an assumption
made in one package, the kernel say, could be coupled with a userland
change, in the NFS client stuff, for instance. If you have the kernel
from one project and the userland from another, you could have a
problem. 

I don't know of a specific instance where this is true for either the
Progeny or FL patches, so it may just be normal sysadmin paranoia. But
I'd worry about it if they were my servers.

On Mon, 2005-03-28 at 18:44 -0600, Mike Vanecek wrote:
> On Mon, 28 Mar 2005 11:55:04 -0800, Howard Owen wrote
> > Progeny is the Linux systems company founded by Ian Murdock of Debian
> > fame. They run a "transition service" for RH 7.2 7.3, 8 and 9. Their
> > numbering scheme is name-base-version.progeny.n .. So the indicated
> > kernel:
> > 
> > 	kernel-2.4.20-31.9.progeny.7.athlon.rpm
> > 
> > is the 7th Progeny patch to the base RHL 9 kernel. There is no way to
> > easily compare this with the legacy kernels, because they are the
> > product of two different patching processes. The only way to tell is 
> > to grab the source RPMs and compare the patches between the two. 
> > It's likely that they have chosen similar approaches to similar 
> > problems, just as it's likely that they have diverged on some.
> 
> That is what I thought, but wanted someone else's feedback. Progeny sometimes
> has updates not available via legacy. Hence, my approach has been legacy first
> and progeny second. Progeny had just released their newest kernel whereas the
> legacy one was released in Feb. However, even if I force install the progeny
> version, yum is going to want to install the current legacy version. Things do
> not seem to be broke, I think I will just let them be.
> 
> Thank you.
> 
> --
> fedora-legacy-list mailing list
> fedora-legacy-list at redhat.com
> http://www.redhat.com/mailman/listinfo/fedora-legacy-list
> 
-- 
Howard Owen        hbo at egbok.com "Even if you are on the right
EGBOK Consultants Linux Architect track, you'll get run over if you
fwd:50279    pstn:+1-650-218-2216 just sit there." - Will Rogers




More information about the fedora-legacy-list mailing list