[LEAPSECS] Some definitions --> practically stated

lennox at cs.columbia.edu lennox at cs.columbia.edu
Wed Mar 9 11:51:33 EST 2011


On Wednesday, March 9 2011, "Warner Losh" wrote to "leapsecs at leapsecond.com" saying:


> Out of curiosity, does anybody know of a system that chose to implement

> time_t as a float/double?


I believe early versions of BeOS did, but it wreaked so much havoc for
code portability that they quickly shifted to using an integer type.


More information about the LEAPSECS mailing list