Error:
[cut lots of config output]
checking how to hardcode library paths into programs... immediate
checking whether stripping libraries is possible... yes
checking for f2c bin/lib/include locations... checking for nag-fortran bin/lib/include locations... ./configure: bad substitution
Error:
[cut lots of config output]
checking how to hardcode library paths into programs... immediate
checking whether stripping libraries is possible... yes
checking for f2c bin/lib/include locations... checking for nag-fortran bin/lib/include locations... ./configure: bad substitution
Any idea what I am doing wrong?
Nothing. The test is coded incorrectly. Or at least the shell being used doesn't like it. The "bad substitution" message comes from the shell executing the configure script while its trying to figure out how to access the f2c program. I take it you're running this on a sparc machine. The logic seems to work okay on Linux but it uses bash. Probably there is some bash specific logic in the configure script that sh doesn't like.
Nothing. The test is coded incorrectly. Or at least the shell being used doesn't like it. The "bad substitution" message comes from the shell executing the configure script while its trying to figure out how to access the f2c program. I take it you're running this on a sparc machine. The logic seems to work okay on Linux but it uses bash. Probably there is some bash specific logic in the configure script that sh doesn't like.
Ah.. ok. I don't think its the logic figure out how to access the f2c program, but the nag-fortan compiler. The f2c stuff has always worked, unless its changed since June.
Can whoever added the nag-fortran checks please fix this?
I tried using bash instead of sh, and that worked so I can continue testing. Thanks.
I have some experience (read: a lot) with pesky systems that can't
handle Bash shell commands. I can take a look at the configure tonight
to see if there's anything glaring.
Odd...this one seems to indicate that it exitted without error...
Hmm. I think that it does default to bash after all (found something in my .cshrc that was resettting CONFIG_SHELL). Its still an error with /bin/sh.. so it should be fixed, but not a release blocker.
