Goswin von Brederlow wrote:
> > But there are not very many bugs with patches. If I read
> > <URL:http://bugs.debian.org/release-critical/> correctly, there are
> > only 123 of those. Send more, send more. :)
>
> And you think patch number 124 will be NMUed any more than the 123
> patches before that?
>
> We can't even find anyone to NMU the bugs with patches that are in the
> BTS already so how would that help?
If I count correctly[1], more than 410 bugs tagged patch were closed in
the past month. This would not seem to mesh with your claim that we do
not have enough developer time to deal with a mere 124 patches for
release critical bugs. Given recent BSP activity, which always starts
with easy stuff, like RC bugs with patches, if a RC bug has a patch that
is over a month old, the patch is probably not very good.
Which is not to say that we don't have far too many bugs in the BTS
tagged patch and not fixed. We have over two thousand of them, excluding
NMU patches.
But my experience as a maintainer and patch submitter is that good
patches are 95% likely to be eventually applied, and that they're 50%
likely to be applied within a month. My experience is also that 10% of
bugs tagged patch do not contain a patch, and >30% of patches are not
good patches. One of the most important jobs of a package maintainer is
filtering out bad ideas and bad patches.
--
see shy jo
[1] wget -O - 'http://bugs.debian.org/cgi-bin/pkgreport.cgi?which=tag&data=patch&archive=no&repeatmerged=no&pend-exc=pending&pend-exc=forwarded&pend-exc=pending-fixed&pend-exc=fixed&include=patch' | grep Done: |wc -l
This does not even include a count of patches applied in NMUs.
Attachment:
pgpwiQ79DqVCD.pgp
Description: PGP signature