AIX 1.x

From: Miles O'Neal <meo_at_netads.com>
Date: Sat Jan 9 12:31:25 1999

Bill Pechter said...
|
|> and don't get me started about the PS/2 AIX...
|
|That wasn't really AIX, though. Didn't it really start with Interactive
|Unix at the root.

That sounds right (I know it wasn't the AIX core).

Too late, you got me started...

IBM has had at kleast three , simultaneous, fairly
unrelated products called AIX. The workstation
version (RT, later RS series), the micro series
(originally only for PS/2), and the mainframe
series. None of them were very compatible at the
CLI level. (There may have been a separet version
for minis, I don't know.)

When the PS/2 came out, we had just recently gotten
away from the Gamma version of SunOS for the Roadrunner
(Sun's i386 box), which was more like a middling Beta
than a Gamma (although the FCS was OK). The PS/2 AIX
Beta made a Sun Beta (which usually felt like a late
Alpha) look like a mature product.

We had all kinds of nasty problems. The rand(3) function
always returned the same value. Some of the basic signal(2)
functions returned random results, and a couple didn't
work at all. The best part was that IBM swore up and down
that there were no known bugs! Eventually someone managed
to smuggle the known bug list out of IBM (one line per bug)
and post it on the net or send it via email to someone on
the project. Karl printed it out, and it was almost a half
inch thick.

IBM really wanted our software ported to the PS/2.
So did our management. We told them to look for it on
VMS first, and maybe even a Kaypro (which was definitely
impossible).

-Miles
Received on Sat Jan 09 1999 - 12:31:25 GMT