Re: ia64/hppa porter NMUs for woody
On Sat, Jul 14, 2001 at 10:21:45PM +0100, Philip Blundell wrote:
> >My question is, shouldn't porter NMUs for non-frozen packages wait at least one
> >day, to give the maintainer a chance to integrate the fix with any pending
> >updates?
>
> What harm do you feel is being done by rapid NMUs? Any new version uploaded
> by the maintainer will replace an NMU'd package.
By that logic, any bug with a patch attached should be fixed with an immediate
NMU. I'm sure that such a practice would improve the quality and community
well-being of Debian. In fact, why bother filing the bug in the first place?
Just upload the fix.
Once an NMU is prepared, the work is done. There is not appreciably more
effort involved in waiting a day before uploading it, and it would only delay
the port a single day if it were done for all packages.
> The ia64 and hppa people certainly are being fairly aggressive, but I think
> that's justified given the circumstances.
The circumstances are that it was decided at the last minute to try to release
ia64 and hppa with woody. More importantly, it was decided at the last minute
to actually report bugs for packages which did not build on those
architectures. The bugs are not new, and the autobuilders I'm sure have known
about them for some time. If lower-severity bugs had been filed weeks or
months ago, the porters might not have so much work to do (with all due respect
to the work they are doing). If the bugs had been open for a long time, I'm
sure nobody would complain about a massive number of NMUs. The bugs in my
packages and others would have been fixed long ago.
--
- mdz
Reply to: