2010/8/26 Reinhard Tartler <siret...@tauware.de>:
> I'm still waiting for an answer on this question. Is this issue solved
> with the latest upload, or do you prefer me to upload the patch posted
> above?

Sorry, should have sent that to the bug itself, too…
(we discussed that a bit on irc last week i think)

It would be nice if you could fix it in your package
(which at least feels a bit cleaner anyway: liba2 breaks liba1
sounds saner than libpartlyunrelated breaks liba1).

The thing is, fixing it for APT is a bit more complicated than it looks:
The patch changes a line which is old. And if i say old than i mean
that this one-line patch circumvents the score-consideration introduced
in rev. 8 of the source. Call me nostalgic, but i am pretty sure that
changing code which is from 1998 and survived until now shouldn't
be changed to upgrade one package shortly before a release if
it can be worked-around easily…
(okay, the breaks are newer - they are from 2006…)

Not because this couldn't be an old bug, but because changing it
now without in-deep testing will very likely break something else.
(scoring is an important concept in the resolver, so my personal guess
is that we can safely place bets on the possibility that it will break…)

We can consider this after squeeze then we have ~2 years of
testing before shipping it…

Best regards

David Kalnischkies

pkg-multimedia-maintainers mailing list

Reply via email to