Re: Formal objection: Changing how the testing of potato works would invalidate the whole test. So please don't change it.
This discussion started in the context of my message about bug 63946
against kernel-image-2.2.15, which _is_ release critical. I think sending
out a release which we _know_ will cause kernel panic on a significant
number of Adaptec SCSI controllers and calling it "stable" would be a
really bad idea. Whether the kernel patch to fix this is applied during
or after this immediate test cycle is not the major issue: there is no
point to a test cycle unless we use the opportunity to fix things which
are discovered to be broken.
In my opinion, it is extremely important that the newest version of the
Adaptec SCSI driver, which fixes this known problem, be patched into the
potato kernel before final release.
On 2000-05-14 at 06:32 -0700, Jim Lynch wrote:
> I hope that during the evaluation phase, it is -decided- what packages
> get in, and potato does -not- effectively become unfrozen, i.e., by
> dark taking Josip's suggestion, that dark would "accept uploads to
> potato" at any time during freeze that is in any way new features or
> code, or unimportant fixes; that is, not release critical. I hope very
> careful thought is applied and the release manager is involved in that
> thinking process when it is decided whether a package gets in or does
> not get in. Debian will not be truely stable without that.