🚀 go-pugleaf

In article <57592@microsoft.UUCP>, jamesth@microsoft.UUCP (James THIELE) writes:
> In article <22...@opus.cs.mcgill.ca> pet...@opus.cs.mcgill.ca (Peter Deutsch) writes:
> |In article <900919072...@lilac.berkeley.edu>, DLV@CUNYVMS1.BITNET writes:
> |> >       Talking about how-to-turn-off-machine I forgot to say
> |> >       that BESM has a software feature to block power turnoff
> |> >       (I think it's a quite valuable idea, especially for Unix
> |> >       machines; what's about self-turning-on machines? ;-).

Actually, I didn't pick up on this bit of the posting at
first, but the Vax 11/780 had a separate CPU monitor which
was able to initiate a reboot when it thought it
necessary (Sound of me banging my head aginst the side of
the monitor: "Whack! Whack!" Nope, sorry, I can't remember
what this thing was called. I'm sure to be told 137 times
by tomorrow night).

Anyways, I remember that power glitches (we get a LOT of
these here in Montreal) in our terminal room was always
fun. Imagine a ten second glitch. It looked like this"

The IBM mainframe 3270 terminals would "flash -
blink" and then go off until the machine could be IPL'ed.
"Go have coffe, gang!"

Our ASCII terminals plugged into our Masscomp's (GAWD!
THere was a piece of shit. Sorry, I don't star out the
'i', they were shit.) would blink and then scramble a few
characters, then go dead. Even hand starting them took at
_least_ an hour. If you were having a bad day,the fsck
would fail and the disk was probably gorfed for good.
Maybeit would be back before the IBM.

Our Vaxen would go "huh!?!" The machine would reset,
reboot and then fsck itself. It would recover and we could
do an "vi -r file" and go back to work. No wonder
everybody bought Vaxen. The competition was so awful.


Actually, someone pointed out to me today that the NeXTen
have serious problem with power outages. Their fsck will
probably clean things up okay, but just restoing power to
the box wont bring it back. Someone (or something) has to
physically push the "power" button before it will restart.
Sort of limits their use in mission critical apps! :-)


> |> i have seen such a unix machine---the off switch just started a software
> |> routine that does synch and then turns off the power via software.
> 
> First generation Apollo workstations did that in 1981.

Of course, I don't claim NeXT was first for much, just
that they've done a pretty good integration job. Pity
their marketing "sucks dead gophers through a garden
hose", to repeat someone else's colorful phrase.