|
|
Boost Testing : |
From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2005-06-25 23:18:37
David Abrahams writes:
> Aleksey Gurtovoy <agurtovoy_at_[hidden]> writes:
>
>> David Abrahams writes:
>>>> Okay; I hadn't read Misha's log in that thread carefully before, but
>>>> there's something screwy going on in the BBv1 code. There should be
>>>> no const_argument.pyd targets there at all; the only pyds should be
>>>> const_argument_ext.pyd. So I'm trying to track that down now and
>>>> hopefully will be able to check in a fix sometime soon.
>>>
>>> Darn, I hate working on BBv1! Well, it's done.
>>
>> Thank you for addressing this! It's been an annoying problem.
>>
>>> We'll see how it works out tonight.
>>
>> Keeping our fingers crossed.
>
> So, the results are interesting. If you look at the Python tests, all
> the white rectangles are gone except for one.
Err, not quite. Quoting your origial post:
David Abrahams writes:
> Can anyone explain to me why in
> http://engineering.meta-comm.com/boost-regression/CVS-HEAD/developer/python.html
> the vc-6_5-stlport tests that fail show up as failures, but the msvc
^^^^^^^^^^^^^^ ^^^^
> tests that fail show up as white rectangles?
So we have at least one white rectangle in the place where before your
changes process_jam_log was able to trace the dependency, and unknown
status for the "msvc" toolset (most likely because the Martin Slater's
run was affected by the issue you've fixed earlier today).
> I think the cause of that one must be different from that of all the
> others. Can you explain it?
I'll try to take a deeper look at what's up here later tonight.
-- Aleksey Gurtovoy MetaCommunications Engineering