Need discussion, Re: Latest contrib perl

[Jim Popovitch]

I agree with Eric's and Jesse's premise that we test and release each identified fix rather than patching an in-process patch.  It is easier to track (as we generally leverage work done by other distros), and it is easier to QA (as our tests only need to be done for a specific issue not various multiple issues.  Ideally I would like to see our release cycle to be nothing more than an import/merge of the RHEL/FCx fix and then a simple QA rubber-stamping by a few members.

-Jim P.

----- Original Message ----
From: Eric Rostetter <rostetter at mail.utexas.edu>
To: fedora-legacy-list at redhat.com
Sent: Tuesday, January 03, 2006 12:19:35 PM
Subject: Re: Need discussion, Re: Latest contrib perl

Quoting Jesse Keating <jkeating at j2solutions.net>:

> My opinion on the matter is that if we already have some QA work done on
> a bug (like we do in this case) we shouldn't interrupt the process to
> add more fixes to the bug, unless other problems are found during the QA
> process.  If no QA has been done then it is easy to add the new fix to
> the sources.
> 
> This is only my opinion, I welcome others.

I agree.  I'm tired of doing QA and having it thrown out as more
bug fixes are added, and then people complaining about the patches
being so late to arrive...
 
Once we have a "sufficient" amount of QA done, we should finish it
and release it.  New bugs that appear go into the next update of the
package.

Just my opionion though...

-- 
Eric Rostetter
The Department of Physics
The University of Texas at Austin

Go Longhorns!

--
fedora-legacy-list mailing list
fedora-legacy-list at redhat.com
https://www.redhat.com/mailman/listinfo/fedora-legacy-list