IBM 610 workstation computer 3448

Sigh! Who is the it in that sentence? This question is the most important section of this post. My observations of hardware types is that they never stop to answer this question with clarity.

I can come up with a dozen scenarios. Hell, I can probably produce an infinte number of them. Divide check in parallel processing would create a situation where the processing needs to "backup" to the older bits. Transaction processing and catastrophic events can create a scenario.

Yup. That's right. And in the olden days, we had backups on magtape or cards or a second disk structure so that the processing could be restarted at some time before an oops happened. In today's computing biz, this is no longer the first restore procedure because it means taking the system off-line for days.

I know. Most of my posting is about the times when things don't work as designed and the software, preferably the OS, has to figure out what to do with the mess the hardware just handed to it.

You now buttume that comm alwasy stays up; I guarantee you that's first thing that goes south.

You are using the wrong numbers to evaluate. You need to look at what the users want and need, not what they buy. I guarantee you that none of them can afford what they want. It's the developers' job to provide them what they need at a computing access price they can spend. This is the computing biz and it has been for the last half century. It hasn't changed.

There are people beating on the closed doors for compute-intesive access.

I'm trying to herd you so you learn how they can be used effectively.

emoticon swoops up into the air, does a double-somersault, three and a half twists and lands flat on its back I shoulda gone for the gold.

gggrrrrrrr....It has been done. 25 years ago...and never after that.

I didn't say that. I don't know. All I'm trying to get you hard types to do is learn about the software side.

That all depends on lot of things. How do you think TW wrote bits to a two-pack disk structure? And the comm biz certainly can't depend on I-O being ordered.

I don't like bells and whistles for the sake of making noise. Clocks have their uses; clock use is not an absolute.

The chronological order is a human condition. The machine doesn't care about time order unless interim results are required for later steps. That's why the bit gods invented internal sorts.

Never. If you believed this, then there would never be any parallel processing nor distributed processing. Nor sorting.

yes. The fact that brick wall of limited capacity has not been hit is the direct cause of the crap (bloat) in today's software. Instead of bored people painting pretty pictures on the TTY, they would be busy doing the real work of trying get the few computes to do useful work.

Excuse me? If a human want the computer to add incorrectly he would as it to add incorrectly.

Yes, and there are many times when the computer is supposed to do the wrong thing.

Apparently not :-))). I'm having loads of fun arguing with you.

In a timesharing environment, one buttumes that they don't finish in any order.

Never. That is why commands were invented. The commands ensured that certain proccesing groups did become ordered. Note again that this is at the human level, not the hardware level. It is arrogant of the hardware to buttume that it knows the right order of all humans.

The KL always had inconsistent memory.

BAH