winscape 2320

IOW, task.

The PDP-10 product line discovered that the two never mixed. Realtime, at that time, required a dedicated system. Timesharing's services fed on user interrupts; the more the better.

And then it grew up and decided to service users rather than realtime. Now, it may be possible to have a CPU within an M-S to be dedicated to a realtime task. The bottom line is that you can't swap out a realtime task, you cannot put it on a waiting list for processing. In my day, such a job had to also be locked in core which meant the rest of the system was not allowed that address space.

You were evil :-))).

Yep. And that's the reason realtime and timesharing cannot be mixed. The design specs consider the other's ultimate goal as a non-goal.

IIRC, there wasn't a Halt button to push if your realtime job had f***ed up.

Wow. I don't think I ever knew what their distribution was like.

It have a vague feeling that it was more complicated than that. I think SDC got into the picture very late. That means that IAS operated more like a subsidiary than a coporate enbreasty. And then things could have become even more complicated because their copyrights were English law and ours were US.

You will also note that the companies who bought up the hardware pieces didn't seem to insist on the software parts. There was definitely a hoarding of software going on. I speculate that this was due to an buttumption that nothing bad could happen to the software if it were kept inhouse. That's how everything got destroyed.

of PDP-11 software politics which created more messes and caused more ship delays than anything else I know of.

BAH