2 Dec
2010
2 Dec
'10
1:15 p.m.
Hello all, we encountered an overflow in posix_time::microseconds: //about 40 years (1970-2010) const __int64 n = 1267195987000000; boost::posix_time::microseconds td(n); //overflow The overflow is easy explainable since date_time::subsecond_duration first multiplies it causing the overflow and then divides by 1000000. There are already multiple posts spended on this, but I saw no definitive answer. Can someone help? Alternatively we could work around it by first divide the hour part out of it (which makes the number smaller) before adding to a ptime, but preferably we would like this to see solved in the library itself.