Boost logo

Boost-Build :

From: Rene Rivera (grafikrobot_at_[hidden])
Date: 2006-10-25 16:41:42

Eric Niebler wrote:
> Rene Rivera wrote:
>> Eric Niebler wrote:
>>> Jürgen Hunold wrote:

> Not sure what (..) part you're referring to, but if I execute "g++ -v"
> from cygwin, this is what I get:

It's usually the last part of the output when doing "g++ --version".

> gcc version 3.4.4 (cygming special) (gdc 0.12, using dmd 0.125)

Which I think in the "-v" it's the "(cygming special)".

> That's for the default g++ that comes with cygwin.

Thanks :-)

>> For now... I've changed it as Volodya suggested. For now the flavor only
>> gets added for MinGW. Could you check to see if this helps.
> Thanks, that fixed it.

OK. Volodya mentioned that the bug is likely not in the toolset code I
change but in one of the regex functions :-(

> I don't suppose you and Volodya have a way of testing out the impact of
> these toolset changes before you commit them, do you? Broken toolsets
> are pretty disruptive. I guess it'd be hard to test, since everybody's
> machine is configured differently. Thoughts?

No way other than testing ourselves on as many platforms as possible.
But it's hard to do this as we don't personally have access to that many
setups. And especially because some problems only show up after trying
to use the tools, rather than when running the regression tests BBv2
has. Now if we had a way of checking in changes and having others test
them without inflicting on everyone that would likely be ideal. Only way
I can think of doing that without too much pain is to have the
regression testers (i.e. do an initial BBv2 test with the
  new code. But for doing the rest of the testing have it use a known to
work tagged version. That is use something like a BUILD_TESTED CVS tag.

-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. -
-- rrivera/ - grafik/
-- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

Boost-Build list run by bdawes at, david.abrahams at, gregod at, cpdaniel at, john at