[HECnet] VMS/RSX Guest accounts

Thomas DeBellis tommytimesharing at gmail.com
Thu Apr 23 15:15:28 PDT 2020


Sir, I stand *corrected* and do humbly admit my errors.  I didn't see 
anything Orange on the mac.io website, so I just assumed...  I didn't go 
to the Facebook site because I won't have a thing to do with their 
policies, stated or covert.

A memory surfaced and I realized that I was also completely wrong about 
the 2020's networking capabilities, too.  In 1979, DEC's Federal Systems 
group had a 2020 on the ARPAnet and I used it to send email to some of 
my pals at MIT LCS when I working 2nd and 3rd shift.  So the hardware 
existed to communicate with an IMP and MIT used it implement TCP/IP on 
ITS.  This is simulated by the KLH10 2020 implementation and that's how 
ITS communicates today.

Federal Systems was a real hike from manufacturing in Marlboro where I 
used to hang out; three buildings away.  It was also notable for having 
mil-spec VT100's.  Instead of plastic, they were made out of very thick 
metal, I think maybe machined aluminum.  The cables where sheathed in 
metal and the connectors were substantially enlarged and strengthened.  
The display glass also had a dense metal screen in front of it.  They 
really looked like they could withstand a hand grenade and weighed a 
ton... Well that's what I thought at the time, another alternative that 
didn't occur to me until decades later was that all this just could have 
been RF shielding.

I had left DEC and was at Columbia before the Internet role out, so I 
don't know whether Tops-20 4.1, the last official release for the 2020, 
supported TCP/IP.  I just can't remember, darn it...  I know that 
Tops-20 5.0 supported TCP/IP AND that MRC was able to port 5.0 to the 
2020.  LingLing was on the Internet from time to time.  That must have 
been some hack; by late version 3, it was seen that the monitor was 
running out of address space and when the hardware folks suggested 
eliminating the symbol table, the monitor folks flipped as debugging 
would have been effectively hatcheted.

The solution for 4.1 was one of the finest hacks I have ever heard of; 
while the 2020 doesn't support extended addressing, it does support 
multiple address spaces, so what they did was move all the symbols into 
a separate address space.  This was called 'hiding' symbols and I 
thought it was great because it made them harder to smash.  However, all 
of that went out the window with 5.0, which fully supported extended 
addressing.

DEC completely walked away from the 2020 and symbol hiding for the 
Tops-20 5.0 monitor.  In fact, I remember an SPR response scolding an 
acquaintance of mine for trying to turn symbol hiding back on in 5.0.  
Basically, it was a corporate, 'Fuggetaboutit'.  Pity; for user 
programs, Tenex and Tops-20 had something called IDDT (Invisible DDT) 
which kept the symbols in a completely different process space, 
effectively making them impossible to smash, no matter how sick your 
program got. Symbol hiding was cool.  Not that I'm going to try figure 
out how to turn it back on...

You know, I have been out to Pittsburgh a few times.  In the 1980's, 
Columbia flew me out there so that I could learn about CMU's 
modifications to LPTSPL to support the nearly entirely awesome Xerox 
9700.  I was out there again about a year ago for a conference.  Still, 
that's a real hike...  My relatives live in near Elizabethtown, not 
quite spitting distance from three mile island (!!), so that's another 
four hours to New Kensington after they're done with me.  Ouch...  Oh 
well, never say 'never'...

That's some pretty good swag you have by the way; minor suggestion, in 
addition to the anti-VAXer T-shirt, you might want to have another one 
that says the somewhat subtle, "If it doesn't have 36 Bits, it isn't a 
Digital computer".  I did know some people who were so anti-VAX that 
they positively would froth at the mouth.  I could understand the 
frothing, given what happened, but still, one prefers not wear one's 
froth on their sleeve.  It would have been a lot better for everybody 
had there been more and better communication and less NIH.
> ------------------------------------------------------------------------
> On 4/23/20 12:20 AM, Dave McGuire wrote:
>> ------------------------------------------------------------------------
>> On 4/22/20 11:42 PM, Thomas DeBellis wrote:
>>
>> Oh, I'm in Looooong Island, South Shore.
>    Ahhh, New Yawwwk.  You should head out this way at some point.
>
>> I actually have relatives not far from you in PA, but I hadn't gotten
>> around to noodging you yet for the distinct lack of 20's in your
>> collection.  I mean, you have a 4341 yet no 20?  Tisk, tisk...
>    You sir are misinformed.  There are three KS10s here.
>
>> Well, I'm pulling your leg a bit; a 20 is a seriously heavy lift.  A KL
>> has three purposes in life: 1) Generate Heat, 2) Suck Power, 3) Run
>> winning code.  A 2020 uses far less power, but doesn't have extended
>> addressing, so it is stuck at Tops-20 4.1, which leaves out a lot of
>> DECnet (and ARPAnet) stuff.  Apparently, you can squeeze 5.0 in; MRC did
>> this, but that was MRC.  Tops-20 is at version 7.0 now.
>    Indeed.
>
>> And both of them are seriously cranky beasts that needed care and
>> feeding from trained service personnel; wire wrap and all that (yech).
>> Remember, it's not a mainframe unless you can't fit it in your house and
>> have to take out a second mortgage to pay for the electricity and tons
>> (60 for a 20) of air conditioning.
>    We're no strangers to cranky around here. ;)
>
>> So Paul Allen had a KL based 20, which is now in the the Living Computer
>> museum.  But that cost was apparently less than a rounding error
>> compared to what he was worth.  A mere monetary blip, as it were.
>    Yep.
>
>            -Dave
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sonic.net/pipermail/hecnet-list/attachments/20200423/14958ae7/attachment.html>


More information about the Hecnet-list mailing list