There's another alternative: Fixing your bug.
This is exactly what I want to do. It's not clear to me yet how to do that but I have some new ideas from this discussion. For example I will play with the idea of controlling the size of the the floats as mentioned by ikegami above. I have already tried explicit rounding rather than using sprintf but I will also try variations of that.
I should have said above that I may have to remove support for uselongdouble Perls until I can fix this bug. This bug has already been open three days.
In reply to Re^6: test fails on 64bit uselongdouble Perl
by frankcox
in thread test fails on 64bit uselongdouble Perl
by frankcox
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |