|
|
Boost Testing : |
From: Vladimir Prus (ghost_at_[hidden])
Date: 2006-07-14 01:20:28
David Abrahams wrote:
> Aleksey Gurtovoy <agurtovoy_at_[hidden]> writes:
>
>> #1 is fixed and should be reflected in the reports within a few hours,
>> fixing #2 requires deciding whether we are droping v1 testing
>> altogether.
>
> We are. But I don't know if we're doing it _yet_; someone else will
> have to report on whether we're ready.
This has two parts:
1. Switching most regression runners.
2. Resolving regressions.
As far as (2) is concerned, I'm always here to comment on any issues that
look V2-related.
As far as (1) is concerned, it looks like most runners have switched, with a
notable exception of metacomm. IIRC, we got stuck because msvc-6.0 did not
work for metacomm -- the compiler rejected one of the options that works
elsewhere -- and Alexey did not have the time at the moment.
Alexey, can we revive that discussion and finally switch metacomm to V2?
If you're out of time for now, I think we might decide that metacomm won't
be switched for this release. That will leave borland tested only with V1,
but it's not a big deal, IMO -- I think borland is one of V2 toolsets
that's pretty well tested outside Boost.
- Volodya