> Wouldn't the user interface port require a replacement or repair in the hardware itself to make it compliant? (I'm not a Y2k programmer myself, I'm just asking.)

Probably not. The vast majority of Y2k issues are
in software only, and ALL hardware Y2k issues can
be worked around in software (although unless the
hardware is on an orbiting satellite, it's probably
easier just to replace the hardware).

> But, um... isn't our Government's most common computer is the old 486? (see link)

But where do you think those 486's are? In charge
of vitally important automated tasks, or the
personal work computers of Joe Blow Government
Worker? So Joe Blow's log of office supply purchases
gets sorted incorrectly. Not the end of the world.

> I've never programmed an OS before in my life, so I can't speak for what it would take to repair one. Nevertheless, the OS isn't the only part of a computer that can be non-compliant. (see link)

No, but it's a representative example.

> Actually, I wasn't thinking about them. I was thinking about how much of a pain it would be to us. What are we going to do if the FAA isn't compliant in time? Flying a 747 is definately out of that picture...

Why? Planes don't use the date to calculate their
altitude. I'd be surprised if plane computers even
*store* the date, let alone use it for any purpose
other than possibly displaying it on a digital screen.
That won't mess anybody up.

> and I'm sure that would have serious consequences for our economy if shipping gets all kinked up.

Nah. This, though, is at least a situation that
might be affected. But what can happen? Miami
port authorities say, "Hmmm, our shipment of sugar
cane from Brazil hasn't shown up yet," so they
make a phone call, and Brazil says, "Oops," and
sends the ship out, and the incident is over.

> Yeah, just a "few" instances... you know, like 15% of American businesses
>9% of UK's businesses
> the nation of Australia
> ...nothing much, really.

But how many of those instances of non-compliance
are actually going to lead to significant problems?
What, are companies going to announce ZERO earnings
for 1st quarter 2000 before it occurs to them that
perhaps the computer screwed up?

> > [all you have to do is reboot]
>
> Is it really that easy on a mainframe? I mean, we're not talking about some desktop computer here.

Sure it's that easy. Mainframes can reboot. It
just takes longer (15-20 minutes, maybe). Actually
it's probably a lot easier than that. No computer
I know (even ancient onces, like 8088s and Apple
IIs and Commodore 64s) grinds to a halt on a
divide by zero error. The *program* crashes (after
which you simply rerun it), but the computer
won't. Plus, software can be programmed *not* to
crash on a divide by zero error. For computers
where it is important that they stay up and
running, there are all kinds of safeguards ensuring
that they stay up. A divide by zero error is an
easy one to trap. But like I say, even if it
*isn't* trapped, the whole computer isn't going
to go down, just that program will.

> (again, a disclaimer of my ignorance into the willy world of mainframes, but I thought that it some of the larger ones hours to reboot)...

Maybe, but it's possible if the computer has
to do a lot of maintenance tasks during boot time.
These computers, however, are VERY hard to bring
down (short of cutting off the power), as they
are built with the expressed intention of performing
tasks where it is important they don't crash.
A divide by zero error isn't going to do it.

> Yess! Some good news! That was one of the things I was hoping I'd hear! Thank you!

If we *hadn't* made our software Y2k compliant,
the worst that would have happened would be that
a few customers would have been mad. That's it.
It would have hurt us, but not the world at large.
A percentage of our customers just would have gone
to our competitors. So yes, that's good news to
the world, but not because we're taking steps to
be compliant -- but rather because it wouldn't make
a bit of difference to anyone but us if we didn't.

Again, 99% if not more of the unresolved Y2k issues
are NOT going to screw anything up on a global scale.

> Got any more good news? (you wouldn't happen to know *how* close you guys are, would you?)

Within weeks.

Anyway, to sum up. The instances of Y2k
non-compliance that will actually affect anybody
on a global scale to a significant degree are
few in number. Those that are are the ones that
people are actively fixing. Those few that don't
get fixed are most likely running on robust systems
that aren't going to up and die. Those few that
do will be sorted out rather efficiently after
2000 hits.

Remember, when you're reading these articles,
that they are not written by software engineers
but by media persons who make money by getting
people concerned about what they write enough to
*read* what they write.

I'm a software engineer. I don't know everything,
but I am in the business. And you have my opinion:
there's going to be some incidents when 2000 hits,
but I'd be very surprised if they were incidents
that caused any lasting global problems. More
likely, it will cause temporary headaches for
some accountants and clerks who will have upgrade
their obsolete systems in a hurry and "make do"
in the meantime.