|
Boost Testing : |
From: Rene Rivera (grafik.list_at_[hidden])
Date: 2005-02-27 03:35:06
John Maddock wrote:
>>> try running a small test by:
>>>
>>> 1. cd [boost-root]/libs/config/test
>>> 2. bjam -sTOOLS=cw test
>>
>>
>> I have attached the results of doing this.
>
>
> Looks like the most pressing problem is the fact that ranlib is failing,
> does the Metrowerks toolset actually need to run ranlib at all anyway (I
> don't know, but I'm suspecting that maybe ranlib expects gcc-style
> object files in the archive)?
Oh, if all things where that simple... So after reviving my Beige-G3
from storage I made some progress..
I did enough changes to disable running ranlib. It fixed that problem.
But there where a variety of others dealing with what kind of runtime
one is targeting. As it stands I got it working by targeting the
MSL/Carbon/Mach-O combination, and that's the *only* one I could get
working. I could not get straight MSL/Carbon working from the command
line, even though it works from the IDE.
So hopefully things will show improvement now. By the way I did the
changes based on CW-8.3, as that's what I have. And I wasn't able to do
full testing as it would take a few days to run the regressions on my
lowly 233Mghz Mac :-(
-- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq