Hi John, > > > And to follow up on that, it appears that POSIX used the C > definition for > gmtime() and it does not *explicitely* require support for > negative time_t > values. It appears that DEC implementated negative time_t > support and other > vendors added it as well (AIX 3.2.5 has it for example). > > Dopes! If you aren't going to support negative numbers, why > use long integer > (and give rise to the Y2038 issue)? > > John
What I find strange is that timegm work correctly!! And in the documentation for timegm it states: These routines are the inverse of built-in perl functions localtime() and gmtime. >From that statement one would think it would work both ways! Just my 2 cents. Ron