RE: time_t size: The year 2038 bug?

[David Schwartz]

> people do and will continue to do stuff like
>
>   time_t tx;
>   ...
>   time(&tx);
>   write(tx,sizeof(time_t),fd);
>
> and this gets embedded into fileformats like tar and various databases
> making it very hard to change.

	That is flat out broken. Nothing you can do will fix that. You obviously
can't fix that with a library or kernel change because any change in the
definition of a 'time_t' will _break_ that code. Code that has that in it
_deserves_ to break.

	DS