Jason Rumney wrote:
Lennart Borgman wrote:
Does not that include calls to convert_time? At the end of
convert_time there is a line
return (time_t) (ret * 1e-7);
Could something has changed in the arithmetics on w32?
Are you suggesting that something in the source code for Emacs is
affecting how the compiler treats floating point multiplication?
I do not know enough to suggest something like that. It looked to me as
there could be rounding errors, but I might be totally misunderstanding
this.
_______________________________________________
emacs-pretest-bug mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
