|
|
Boost Testing : |
From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2005-06-06 16:14:23
Stephen W. Carson writes:
> Rene Rivera wrote:
>
>> Jeff Garland wrote:
>>
>>> So...we really need to standardize toolset naming for regression
>>> testers.
>>
>>
>> It is standardized as: [toolset]-[version]-[platform]-[cpu]-[build]
>
> First time I've seen that. I think my toolsets are named fine
> because I did what I was told to do.
And they are fine. "[cpu]" part is necessary only if you are running a
multiplatform OS, and "[build]" is needed only if you are testing
release builds.
>
>> As far as the release and Beman's gcc tests. Don't worry about it,
>> i.e. don't do anything about it. Only toolsets marked as required
>> for release are important right now. And *only* toolsets that are
>> properly versioned are marked for release.
>
> Which toolsets are marked as required for release? How do I know?
They are listed at the top of
http://cvs.sourceforge.net/viewcvs.py/boost/boost/status/explicit-failures-markup.xml,
highlighted in bold on the report pages, and are the only ones that are
displayed in the "Release View".
>
> You wrote (in a different message) that all a toolset requires is
> a commitment from one person to support the platform. Does that
> mean running the regressions faithfully as I've been doing for gcc
> and cw on the Mac? Or more than that?
Quoting an old posting of mine:
> Currently, the toolset is marked as "required" if either of the
> following holds:
>
> 1) The compiler/configuration is known to be widely used, and/or the
> majority of Boost developers is committed to supporting it. The
> current versions of MSVC, Intel, GCC and CodeWarrior all fall into
> this category.
>
> or
>
> 2) The compiler/configuration is not widely used, but there are one or
> more Boost delevopers who consider it important enough to be
> included in the release criteria, and therefore volunteer to run
> regression tests for it, fix/mark up its failures, and generally
> take responsibility for driving things towards the release at
> reasonable speed. "intel-win32-7.1-vc6-stlport-4.5.3" is one such
> example.
-- Aleksey Gurtovoy MetaCommunications Engineering