[Coin-discuss] Problems compiling bonsaiG with gcc 2.96

[Ted Ralphs]

Lou,

I'm writing a paper on open source software for MILP and so I'm trying to 
compile bonsaiG to do some benchmarking and I'm failing. I modified the 
makefiles for Linux and everything seems fine, except that the routines you 
wrote for FORTRAN IO seem to have problems compiling and I can't quite see 
what's going on with them. I'm usng gcc 2.96, which I think you claim works fine 
(or at least 2.95 does). Anyway, I get the following errors when compiling:

make[1]: Entering directory `/home/tkr/src/Dist.010630/Libraries/IO'
Compiling errs.c to errs..o...errs.c: In function `errmsg_':
errs.c:542: invalid lvalue in assignment
errs.c:553: invalid lvalue in assignment
errs.c:556: invalid lvalue in assignment
errs.c:559: invalid lvalue in assignment
errs.c: In function `warn_':
errs.c:653: invalid lvalue in assignment
errs.c:664: invalid lvalue in assignment
errs.c:667: invalid lvalue in assignment
errs.c:670: invalid lvalue in assignment

Looking at those lines in the source file, they do in fact seem incorrect to me, 
but I can't really see what's happening easily. If I comment out the -DFORTRAN 
in the makefile, all comiles fine, but then of course it won't link. It 
complains about a bunch of undefined references in yla05.f. I seem to recal you 
saying something about someone replacing the Fortran routine with another C 
library at some point. Am I recalling correctly? What say you?

A more general question: Is it worthwhile for me to do this? Is bonsaiG still a 
viable code? Is it supported? If so, I may ask for your help reviewing my 
write-up about bonsaiG for the paper. Let me know what you think and thanks for 
your help.

Cheers,

Ted
-- 
Dr. Ted Ralphs
Assistant Professor
Industrial and Systems Engineering
Lehigh University
(610)758-4784
tkralphs at lehigh.edu
www.lehigh.edu/~tkr2

"An optimist is someone who thinks
'It doesn't get any better than this.'
A pessimist is someone who's afraid that's true."