Re: [igraph] Installing igraph on Solaris

[Jens Pfau]

Hi,

thanks for your quick response.

I was able to solve the problems with pottsmodel as you suggested. All
log() or exp() calls have to receive long doubles. The line numbers
where the change has to be made are 660, 839, 1232, 1234, 1235, 1236,
1239, 1240, 1241, 1242, 1243, 1918. In line 1663, the '0.0' has to be
changed into '0'.

Substituting the include of math.h by ieeefp.sys in math.c does not
help the finite-function warning.

The major problem, I guess, is the assembler failure. I don't have any
clue how to approach that problem. Do you have any idea?
It seems that the fault is triggered independently of whether I
configure with my own built gnu assembler or not.

Thanks a lot for your help.