Re: [boost] Bug reports - Showstoppers and Regressions

Subject: Re: [boost] Bug reports - Showstoppers and Regressions
From: Rene Rivera (grafikrobot_at_[hidden])
Date: 2012-01-17 10:17:00

On Tue, Jan 17, 2012 at 8:30 AM, Vicente Botet <vicente.botet_at_[hidden]>wrote:

>
> Rene Rivera-2 wrote
> >
> > On 1/17/2012 7:07 AM, Vicente Botet wrote:
> >>
> >> Marshall Clow-2 wrote
> >>>
> >>> There are a couple of useful reports on svn.boost.org:
> >>>
> >>> Active Showstoppers:        https://svn.boost.org/trac/boost/report/43
> >>> Active Regressions:         https://svn.boost.org/trac/boost/report/30
> >>>
> >>> IMHO, these bugs should be at least looked at before each release.
> >>>
> >>>
> >>
> >> Hi,
> >>
> >> Concerning the Boost.Thread showstoppers,
> >>
> >> #4521        Error using boost::move on packaged_task (MSVC 10)
> >> #4885        Access violation in set_tss_data at process exit due to
> invalid
> >> assumption about TlsAlloc
> >> #6174        packaged_task doesn't correctly handle moving results
> >> #6194        Adapt to Boost.Move
> >> #6222        Compile error with SunStudio: unique_future move
> >>
> >> Yesterday I made a commit in trunk  that should take care of all of them
> >> (I
> >> don't think #6222 will behave as expected after the Boost.Move
> >> adaptation,
> >> as SunStudio doesn't support well Boost.Move) except #4885.
> >>
> >> This commit contains much more that the fix for these Showstoppers. My
> >> idea
> >> was to merge the whole to release for 1.50. Should I merge the trunk
> once
> >> the tester results are good enough or should I merge only the code
> >> related
> >> to these Showstoppers?
> >>
> >> I don't think that I could have time to take care of #4885 or #6222.
> >> Anthony?
> >
> > You do realize the deadline for merging changes to release was yesterday?
> >
> >
>
> Yes, this is why I've planned to merge them on 1.50:)
>

D'oh. I didn't read you email carefully enough :-\

>
> If we consider that them are Showstoppers we should do something else. Note
> that these Showstoppers  were already for 1.48, ...
>

Well, they could be relabeled  to not showstoppers. Especially if it's a
problem for a "not release tested" platform.

-- 
-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com
-- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo