Simon McVittie wrote:
> The reason I brought it up is that you mentioned detecting such bugs as
> a reason to prefer not to force TZ=UTC for related packages' builds;
> but if that makes a package's build fail, then it becomes "artificially"
> RC, whether it would have been RC on its own merits or not.
Just to be clear about the double-negatives here; setting TZ=UTC did
not make the timezone library's testsuite to fail, but being on "Tokyo
time" did.
Release critical or not it's still a bug that surely needs fixing and
playing severity wars in the BTS is, at best, just getting in the way of
fixing it.
I don't file issues automatically and exercise some judgement on their
severity level depending on how tortured the required environment is and
what the package actually is meant to do, etc. I'm not even certain I
filed this one as RC which makes this conversation even more academic… :)
Best wishes,
--
,''`.
: :' : Chris Lamb
`. `'` la...@debian.org / chris-lamb.co.uk
`-
