On Wed, May 19, 2004 at 10:26:36PM -0700, Jesse Keating wrote: [snip...] > As we move forward, streamlining updates is absolutely necessary. In > order to streamline, the bottlenecks need to be addressed, and today > these road blocks (aside from me and my personal time management > issues) are RHL 7.2 and RHL 8.0. [snip] > Please provide your (relevant) feedback. Thanks. IME it's possible to "upgrade" a machine from RHL 7.2 to recompiled RHEL 2.1 clones (such as CentOS-2) relatively easily. This is not the case for RHL 7.3; many major RHL 7.3 components are newer than what shipped with RHL 7.2 and RHEL 2.1 (e.g. KDE 3.0 vs. 2.2, glibc 2.2.5 vs. 2.2.4). However, it's better to support a few releases well than many releases poorly, and it seems like many more people are running 7.3 than 7.2, so dropping 7.2 might be a good idea. FWIW, I don't yet know whether I'll need support for RHL 8.0. I know that at work, some people had various software compatibility problems due to NPTL in RHL 9. I don't remember the exact details, except that LD_ASSUME_KERNEL worked for some software, but not all (even booting into an NPTL-incapable kernel didn't fix all problems); vendors had to provide updates for the rest of the software. I *assume* there are now updates available for everything that can't be fixed with LD_ASSUME_KERNEL, but I won't know for a fact until I find and upgrade all the 8.0 boxes... However, I can't justify spending time on build or QA work for RHL 8.0 until I find a box that absolutely cannot be upgraded -- if such a box even exists -- so if there aren't many other people who need 8.0 for glibc-related reasons, then I have no problem with 8.0 being dropped. -Barry K. Nathan <barryn@xxxxxxxxx> -- fedora-legacy-list@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-legacy-list
