Subject: Re: [Boost-build] Boost.Build test "default_toolset" broken on trunk ?
From: Vladimir Prus (ghost_at_[hidden])
Date: 2008-12-20 03:48:11
On Saturday 20 December 2008 10:55:29 Steve M. Robbins wrote:
> Every single test machine building trunk is failing the
> regression test for Boost.Build, specifically the
> test "default_toolset".
> See http://www.boost.org/development/tests/trunk/developer/build.html
> The test output is as follows. Is Boost.Build broken or the test?
It depends on whether we think that the msvc toolset should be able to
default-initialize itself on Linux. Probably not, in which case it's the
test that's broken.
BTW, it's a bit early to look at boost.build tests in the matrix -- there
some integration issues that make analysis hard.
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk