IBM 610 workstation computer 3458

I've seen lightlbulbs fail too. So *what*! Does that mean we don't use light bulbs?

Good luck with that one! Self-modifying code is ugly on modern processors, certainly. I-Caches aren't multi-ported and coherency may have to be forced, but what's this got to do with the price of oats?

Perhaps it doesn't *need* access, but you can't prevent access since the disk they were loaded from is writable.

That might help, but it's not the end-all. I could just as easily propose that this ugliness wouldn't exist if runtime bounds checking was done, and likely be no more wrong than you.

If you can find ROM for .15 nanobux per byte, go for it. Now try updating your software.

It may be elementary. If so, we aren't even in kindergarten. If you want to go Harvard, go for it. The rest of the world will let you. I hope you have deep pockets.

Maybe you were discussing this, but I wasn't. Keep drifting...

You have a strange outlook. Who does this work? The disk decides the program has finished and cleans up its VM? I didn't know the disk knew anything about VM. Bytes is bytes - sectors is sectors.

Ok, so use them. They have nothing to do with coherent caches and a consistent hardware memory model though. Nope! Opteron and PowerPC (970 and Power4-5 varieties) SMP configurations don't connect to the same bus. The concept of a "bus" is antique. Physics, and all that.

A perfect example of an embarrbuttingly parallel (and silly) problem. You might just as well talk abut the millions of people using Word every day.

-- Keith