On Thu, 29 Apr 1999, Riley Williams wrote:
> > Are you sure that it is 190 ? Not 100 ? The latter would be easy
> > to explain with "struct tm". It has a member tm_year that is
> > often mistakingly taken as representing the year and decade as a
> > two digit number when it actually is years since 1900 (see man
> > mktime). The fix would be to get two digits with (tm.tm_year %
> > 100) or better the correct year with (tm.tm_year + 1900).
>
> That 190 is almost certainly correct - I would imagine the relevant
> code looks something like the following:
>
> Q> printf( "%02u.%02u.19%u", tm_mday, tm_mon, tm_year % 100 );
Aah, ok, well that would explain the 190. However that would be
somewhat... err... funny coding. The coder knows that tm_year can be > 99
but doesn't realise that if it really is, it means the century has
changed...
> Try it, and you'll see you get just that for a date of 1st Jan 2000,
> and the fix would be to use the following instead:
>
> Q> printf( "%02u.%02u.%u", tm_mday, tm_mon, tm_year + 1900 );
Yes. Just like I said above.
--
--... Tomi Manninen / [EMAIL PROTECTED] / OH2BNS @ OH2RBI.FIN.EU ...--
