On Wed, Mar 1, 2017 at 2:45 AM, Ronald Natalie <ron@ronnatalie.com> wrote:

> On Feb 28, 2017, at 10:15 PM, Gregg Levine <gregg.drwho8@gmail.com> wrote:
> Hello!
> We (well most of us) all of us know about AIX. Well what about AIX/370?
> ——

AIX/370 was a real product.     One of the ones that I don’t ever think saw the light of day was the i860 AIX port.   IBM made two i860 add-in cards for the PS/2.   The single processor version was called the Wizard and there was a 4 processor version with an integral frame buffer called the W4.     We ported AIX to both.    The i860 version actually had more in common with the 370 version than it did with the 386.     All of these AIX versions came from the same source code and used the IBM TCF to allow you to transparently run executables across nodes in the cluster.     The only AIX that didn’t play was the completely independent (and in my opinion somewhat brain damaged) IBM/RT UNIX.

We had RTs where I was. By the time I came on the scene, they were being decommissioned in favor of RS/6k hardware (arguably, the RT was pretty low-powered even for its day), so the students were running around grabbing them and playing with them. We ran AOS on ours, which was a more-or-less straight port of 4.3BSD+NFS (maybe they started with Tahoe? I don't know), but IBM seemed to want to push AIX with them. The RT was my first exposure to "real" Unix source code.

What was interesting to me was all of the #ifdef's in the source that made it clear that someone at IBM had obviously tried to port 4.3 to the 370. I don't think that ever saw the light of day, but there were definitely vestiges of it in the kernel.

I had understood was that AIX/370 was actually OSF/1 based and totally separate from AIX 2.x (on the RT) and AIX 3.x (on the RS/6k)? From what you wrote, it sounds like that wasn't quite right.

If there was a TCF-based RT kernel, I never saw it, even inside the IBM labs.

Speaking of odd job control mechanisms.   The 386 side had a device that multiplexed the PS/2 console into multiple streams called the “High Function Terminal.”   When we wrote the virtual console for the Wizard/W4 add in card, we called it the “Low Function Terminal."

Did the HFT survive into the RS/6k? I seem to recall hearing about that. Perhaps it was an option on the RT, or somehow could be used with the "crossbow" card on the 6152?

        - Dan C.