>Casper wrote:
>This is a known problem that affects most computer system including all
>Unices. A solution is to go 64 bit, something that is expected to happen
>in the next decade. Older machines will be a lot of trouble in 2035-2038.
>I hope, but do not expect, this will be solved in time. Solving it on
>other machines is IMHO harder than Y2K but doable.
-Bernie replied:
-And the problem is exactly what?
-We only change the start year from 1970 to 2038 (or whatever it is) if the
-year reported is less than 29 (for a program made after 1999 of course -
-this will change as the years passes by...)
-IMO "32 bits is enough for anyone" ;-)
-There are other problems, which will require a complete rewrite of the
-standards, SQL for instance stores the year like a string (something like
-"2000-02-08") and therefor has a Y10K bug.
-But I doubt anyone will use any SQL database I create now after 8 000 years ;)
I tested my machine on a variety of dates until I found the year where
I would have trouble. I think it was something like 2110.
