arvid.herzenberg wrote:
The GPC Test Suite in section 11.7 of the GNU Pascal Manual gives me 29 failed tests. This is 1.6% of the total of about 1800 tests; except for the 29 failures, the other tests all give 'OK'.
I would like your opinion whether I should simply ignore the 29 failures. Or do you want me to e-mail the titles of the 29 failed test to you?
The latter, as well as the error messages, i.e., the complete output from running the test suite (not in the long form, and overly long list of errors for a single program, if any, trimmed).
In trying out GPC, I have so far completely failed to get an answer to the GPC demo-program 'pi.pas'. All pi.pas gives me is an error message to say that linking was not possible. I presume that the answer should have been 3.14159...
It uses the GMP library to compute pi to arbitrary many digits. If you haven't installed it yet, it won't work (though I'm wondering why the error occurs at link time, and not at compile time already). There must be /usr/include/gmp.h and /usr/lib/libgmp.a (or .so*) for it to work.
Frank