From: Beman Dawes (bdawes_at_[hidden])
Date: 2003-09-05 15:36:38
At 06:20 AM 9/5/2003, Toon Knapen wrote:
>Indeed case does matter. 'boost_root' however is defined arount line
>number 10 in the run_tests.sh. Now I start to wonder, would it be better
>to just rely on BOOST_ROOT (the env.var.) instead of defining boost_root
>and providing this var to the command line of the 'compiler_status' exe
>? Beman ?
Sorry, I don't have enough context to answer that, and just this morning I
again shot myself in the foot on a Unix system, proving yet again that I
know only enough about Unix shell scripts to be dangerous.
Maybe part of the answer is to add some debugging options to the regression
test programs. Any ideas?
>I tried to do that but. On Sun this works fine for instance.
>On IBM however, I get
>/home/tk/boost_regression/boost/status> bjam config_info
>...found 35 targets...
>...updating 1 target...
>mkdir: 0635-358 Cannont create ...
>...: Do not specify an existing file.
>This is really weird. I have not set ALL_LOCATE_TARGET. Anybody any idea
>what could be going wrong ?
Um... That looks bjam related. Maybe ask on jamboost_at_[hidden]
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk