Re: Year 2000.

On Tue, 10 Jun 1997, Jay Denebeim wrote:

> > I'm reasonably certain it's stupid to ask if Exim will have any problems
> > itself with the year 2000, but nevertheless I'll ask it :) I myself
> > can't see how it would, but I've been wrong before, and probably will be
> > again! Besides if our Year 2000 committee ends up asking me on it, I'd
> > rather have the opinion of the code maintainer to give them then simply
> > my own.
>
> I sorta doubt *any* unix application will have year 2000 problems. Times
> are stored in seconds since jan 1 1970 00:00, so nothing is going to roll
> over in 2000.

I doubt any would have problems with storing data or with calculating with
it; inputting it is another matter. I imagine lots of programs only have two
digit entry fields to save typing.

Also, RTC chips often only use two digit dates, which could cause problems.

Generally, what you say is correct, Cobol programs are the worst culprits,
but they're not the only ones, and any company would be stupid not to check
any mission critical software for any platform.

[don't quote me on any of this :)]

What I don't understand is why some people seem to think traffic lights will
have problems (as some newspapers seem to think)? Why should they need the
absolute time? All I can think of is that they might make a mistake over
whether it was a weekday or a weekend, if they currently take that into
account (do they?), but that just makes things sub-optimal, not dangerous,
and could easily be tweaked after the event.