On Tue, Nov 6, 2012 at 11:12 AM, Andreas Koenig wrote: > Shmuel Fomberg writes: > > > I remember that with the Data::ParseBinary I had to dig around an find > which > > compilation option caused my tests to fail... that was not fun at all. > > Can you explain what the specific problem was? Maybe we can learn from > it a bit. >
My memory is a bit fuzzy, (it was about three years ago) but it was about handling 64 bit numbers. The compilation switches were use64bitint and use64bitall, that if the former was true and the letter was false it caused Math::BigInt to give wrong results. So I talked to the Math::BigInt guys, and they already knew about that bug and it was fixed on the next release. Versions were bumped and case close. Shmuel.
