#3: Mark Knopper, Dave Katz, and Bill Norton get dragged into the discussion

posted Apr 4, 2012, 4:51 PM by Jeff Ogden   [ updated Apr 5, 2012, 11:37 AM ]
On Mar 6, 2012, at 10:55 AM, Jeff Ogden <jeff.ogden@umich.edu> wrote:

Hello,

How good is your memory for things from 20 to 25 years ago?

A few folks are working to resurrect and preserve MTS. Versions from 1988 and 1996 are running under the Hercules S/370 emulator. Last July the U-M gave us permission to make the source and object freely available. And late last December we made MTS D6.0 (1988) available and some folks with no prior knowledge of MTS got it working within less than 48 hours. So you can now run MTS from 1988 in an emulated S/370 on a Mac, under Windows, or under Unix/Lunix.

Now we are working to make a more recent version (1996) available and to make it more complete.  One of the things we want to improve is the networking support, since right now that is pretty much limited to tn3270 terminal support. One effort is to create an emulated HIM device in Hercules (it is C code, no need for PDP-11 assembler). Tom Valerio has something that sort of / mostly works (it does FTP and Finger from MTS to the "Internet").  Mike Alexander just started to work with Tom on it a bit.

The problem we have is that none of us knows or remembers how all of this stuff hangs together.  In a recent e-mail note Dennis O'Reilly (retired, but still working as a consultant at UBC) explained that what U-M ran was based on, but different from what was used at UBC and implicated some of you in working with him and Mark Fox on this as part of the NSFNET effort.

So I'm writing to see if you'd be willing to help us dust off our collective memories about how things worked (and to see if I've got good e-mail addresses for you)? Let me know.  If you are, I will send more information about what we have figured out or remembered so far.

There are two e-mail messages included below for background, one from Gavin and one from me.

And, if you'd like to be added to the mts-interest e-mail group, let me know (Al, you are already on it). The list doesn't get a lot of mail, one message every few months at the most.

   -Jeff



On Mar 21, 2012, at 12:01 PM, Mark Knopper wrote:

Hi Jeff,

Long time no see! Hope things are going well, other than this obsession with ancient operating systems.

I never worked with the UBC technology. But Dave, Al and I could easily whip up a PCP with a Merit 370 interface and have your telnet VT100 access up and running in no time. (*)

Mark

(*) Just kidding. No way.



On Mar 21, 2012, at 2:17 PM, dennis.oreilly wrote:

Dave was the guy I think.



On Mar 21, 2012, at 2:13 PM, Jeff Ogden wrote to Mark Knopper:

You can try to deny any knowledge, but Dennis remembers differently (see note below, emphasis added :-).

Thanks for the offer, but I don't know where on my Mac I'd plug in the channel cables. (*)

Yes, it has been a long time.  I retired (quit) from PTC/Arbortext about two years ago.  Life is pretty good.  I'm enjoying retirement.

As to my obsession with ancient operating systems ….  Aren't most operating systems in use today ancient? Unix certainly is since it goes back to the same time as MTS in the late 1960s and early 1970s.  I guess IOS may not be, but isn't IOS based on Mac OS X (Darwin) and thus on Unix to some degree?  Windows is pretty ancient. The Android OS is Linux-based, so that goes back to Unix.  What else do you have?

How is the baseball statistics app world doing?  Are you involved in other stuff?  Still have an office in downtown AA?

   -Jeff



On Mar 21, 2012, at 4:26 PM, Mark Knopper wrote:

Hm, well yes I definitely remember going to a few bars with Dave while he designed and coded the project. But I give all the credit to him.

I'm semi-retired. The Baseball app is the only one that I've submitted and that has actual users. It has a small but dedicated following. I have written a number of apps for my own use but have not deemed any to be ready for prime time. I'd like to do something about that, or maybe get some other ideas for apps to write.

You are definitely right about all OS's being ancient. OSX is based on Darwin which is an evolution of BSD Unix. It's rather amazing that none of the rest of them are surviving. Perhaps MTS will now rise from the ashes and come to its full glory. I still find myself wishing for features of the mighty MTS Editor.

Mark



On Apr 3, 2012, at 1:13 PM, Dave Katz wrote:

It took me a bit to remember what the hell a HIM was...  ;-)

I did indeed work with Mark and Dennis in 1987 or so to get the HIM running (almost called it a NIM, which was something else, somewhere in my past...)  I do clearly remember when it first started working, as it led to the following exchange between them:  "Beauty, eh?"  "Woh!"  They also stood slack-jawed at one of the giant plate glass windows in the CC watching an incoming tornado before being encouraged to head for the basement.

When I was working on it, we were getting it ready to provide an interface between NOTIS and UMnet, via the venerable Auscom channel interface (buggy piece of shit, as I recall).  I think I added a second plane of screen memory for the ALA character set (diacritics, Polish L, Icelandic Thorn, etc.)  I learned waay more about 3270s and bus/tag than a guy with a CDC background should ever have known.

DOS/VSE was running on the mainframe, and the HIM was pretending to be a 3725 or whatever the hell those things were called.  It acted as a protocol translator between an unexpurgated 3270 controller feed and the Merit INP, which had been extended a bit by Al to support PC-Tie as a 3270 emulator, and also to support remarkably good field-based terminal access on vt100s (the SCP was told what the field width was;  when the cursor left the field, it would turn off echo and forward the input.  I think we were justifiably proud of that hack.)

Then Bill Norton and I added 3270 printer support (he called the PC driver "Fester" for reasons known only to him;  the splash screen was quite creative) when it became known, very late in the process, that they planned on using this to print checkout slips or somesuch.  I think we pulled it off in about two weeks, as it was perilously close to the date it was supposed to be turned on for the public.

I think the other near-disaster was that DOS/VSE only supported a limited (far too limited) number of terminals;  this was discovered only at the very end of the process as well.  I think the solution was to run multiple copies of it under VM.

Tim Prettyman was our sane contact within the libraries;  our respective management forbade our talking directly, so of course we did so in order to move things along.  I recall Mike Stolcarczk (not sure about the spelling) playing hardball on our side;  it was all terribly silly, but there it goes, or went.

I remember thinking that PLUS-11 was a great leap forward from PDP-11 assembly (I liked named blocks a lot;  still haven't seen that in a language since).  Now I'm a Scala fan, so I guess I'm still trainable.

Having said all that, there was no MTS interface at the time, so I haven't any idea how the HIM figured into that world.  I moved on to NSFnet in late '87 and had my hands full attempting (without success) to surgically implant a sense of humor into Hans-Werner, so I stopped paying attention.  (I recall trying to explain the premise of "Hogan's Heroes" to him, without success, at a Diverse Holiday party at Mirv's house).

Did this become a replacement for the Ontel-based scheme for MTS?  I could imagine that the PC-Tie and VT100 support might have been interesting for MTS as well.

All this is terribly interesting, but unfortunately contributes nothing to the discussion.

For what it's worth, I'd be happy to stay in the loop.  The source email on this message would be the one to use.

It would be cool if somebody could type "duderstadt" at the "#" prompt.  I want to know if it still responds, "Do you mean 'dunderhead'?"  (And was that serendipitous or was somebody being clever?)

--Dave



On Apr 3, 2012, at 2:43 PM, Jeff Ogden wrote:

Thanks Dave. This was fun reading, even if it wasn't directly helpful on the MTS question.

Mike Stolarchuck was the follow.  His initials were MTS, but he was more of a Unix guy. Shifrah and I were (are?) god parents to his son, not that we ever did much.  I believe that Mike is out in California now.

Tim Prettyman is still around and still working in the U-M Library's Info Tech office.

It is good for a CDC guy to have his perspective broadened. It was an IBM 3274-1D control unit.

It was after my time working on MTS, but I think the HIM supplemented rather than replaced the Ontel and other "MCP" programs. As best we can figure out, what was in use with MTS at UM was a hybrid between UBC's NIM and other UBC boxes (NIM, Node, …) whose names I haven't yet gotten straight.  It enabled true TCP/IP connections in and out of MTS and very possibly other more traditional "3270" sorts of things as well. I'll forward Dennis's note in a few minutes.

On Apr 3, 2012, at 1:13 PM, Dave Katz wrote among other things:
It would be cool if somebody could type "duderstadt" at the "#" prompt.  I want to know if it still responds, "Do you mean 'dunderhead'?"  (And was that serendipitous or was somebody being clever?)

I don't think this was at the MTS command (#) prompt, but rather something from the MTS $Message system when you entered a name on the To or CC field of a message that wasn't recognized as valid (not in the *Userdirectory database). That program used the Soundex algorithum to give people a list of alternatives. And so what you saw would depend on the names that were and weren't in the database. And the Soundex scheme certainly produced some surprising results, so this story could be true. I'm sure that it was serendipitous rather than deliberate.  We can check with Jim Sterken, the author of $Message, to see if he remembers. And for now you are considered the authority on this topic, see:  http://archive.michigan-terminal-system.org/discussions/anecdotes-comments-observations/doyoumeandunderhead.

  -Jeff

Note: The "dunderhead" discussion continues in the 'Do you mean "dunderhead"' discussion item.



On Apr 3, 2012, at 4:30 PM, Dave Katz wrote:

On Apr 3, 2012, at 11:43 AM, Jeff Ogden wrote:
It is good for a CDC guy to have his perspective broadened.

Little did I realize that I would learn all I needed to know about concurrent programming at the age of 18.  Now I rail around saying things like "we'll never get locks/mutexes to work, or at least not more than once, and when it fails we're all screwed!"  Thus my embracing of Scala, which has immutable data structures as first-class citizens...

 It was an IBM 3274-1D control unit.

Ah, right.  The 3725 is what we were convincing the libraries *not* to buy...

It was after my time working on MTS, but I think the HIM supplemented rather than replaced the Ontel and other "MCP" programs. As best we can figure out, what was in use with MTS at UM was a hybrid between UBC's NIM and other UBC boxes (NIM, Node, …) whose names I haven't yet gotten straight.  It enabled true TCP/IP connections in and out of MTS and very possibly other more traditional "3270" sorts of things as well. I'll forward Dennis's note in a few minutes.

Ah...

It would be cool if somebody could type "duderstadt" at the "#" prompt.  I want to know if it still responds, "Do you mean 'dunderhead'?"  (And was that serendipitous or was somebody being clever?)

I don't think this was at the MTS command (#) prompt, but rather something from the MTS $Message system when you entered a name on the To or CC field of a message that wasn't recognized as valid (not in the *Userdirectory database). That program used the Soundex algorithum to give people a list of alternatives. And so what you saw would depend on the names that were and weren't in the database. And the Soundex scheme certainly produced some surprising results, so this story could be true. I'm sure that it was serendipitous rather than deliberate.  We can check with Jim Sterken, the author of $Message, to see if he remembers. And for now you are considered the authority on this topic, see:  http://archive.michigan-terminal-system.org/discussions/anecdotes-comments-observations/doyoumeandunderhead.

I think I mentioned this on the Risks of Computing list 10 or 15 years ago (in a discussion on the risks of spell checking) and referred to MTS as "late, lamented" or something along those lines.  Mike contacted me privately to take exception to this description...

Speaking of people with too much time on their hands, in 2006 I went to the 50th anniversary celebration of the MSU Computer Laboratory (I was an employee from years 20-29).  Along with seeing pieces of hardware that I helped purchase in a special exhibit in the MSU Museum (ouch), somebody that had been a student programmer around the time I left showed me SCOPE/HUSTLER running on a PC under Windows (running at roughly the same speed as the Cyber 750).  They had not only emulated the mainframe, but also the disk and tape controllers (allowing them to deadstart the O/S) and the 6684 mux that tied async lines to a channel.  I don't think they went so far as to emulate the Interdata/Perkin-Elmer 7/32 that we used for terminal front-ends later (a truly fine machine).  I did receive complete source listings (full-width line printer images, natch) of FREND (the 7/32 code), and also 1SJ, which swapped Hustler jobs to disk.  It was interesting looking at code I wrote over 30 years ago--the style is recognizable, though probably because Richard Moore (RIP, died last year) had very strict coding standards, and I brought a lot of that with me to Cisco and then Juniper.  At the time it was irritating and seemed overly pedantic;  in retrospect it taught me a great deal about how to write code that doesn't totally suck, and how to code defensively.  Most of the ex-Systems-group folks at the reunion at Richard's house that night had similar things to say...

But then Len Bosack took his paltry nine figures of Cisco money and started a company building DEC-10 clones in the 90s;  that's taking it to the next level.

The efforts to build working versions of the Babbage Difference Engine, on the other hand, are cool...

--Dave



On Apr 3, 2012, at 4:36 PM, Dave Katz wrote:


Ah, you found my old message.  15 years ago?  Yikes.  At least my story is still consistent, meaning that either it's true or my hallucinations are repeatable...  ;-)

--Dave



On Apr 4, 2012, at 9:52 AM, Bill Norton wrote:

Thanks for the stroll down memory lane. :-)

I published a book last recently called "The Internet Peering Playbook: Connecting to the Core of the Internet" that talks about the way the Internet interconnection works. I've been doing the book promotion stuff lately and this guy just out of college spoke with me and said something striking:

"We all just assume that everything under HTTP, what you call the "core" will just be there. We have massive teams of people coding up apps that service the complex structures in layer 7.

That blew my mind - that the Internet is to the point where everything below layer 7 is considered the "core" that has and will always be there.

Pretty soon my grandkids will come over and laugh at me for the modern equivalent of the flashing 12:00s on my Facebook profile.

Bill



On Apr 4, 2012, at 9:07 AM, Jeff Ogden wrote:

"the PC driver "Fester" for reasons known only to him; the splash screen was quite creative"?

So, Bill, do you remember the reasons?  Does anyone still have a copy of the splash screen?

  -Jeff



On Apr 4, 2012, at 12:19 PM, Bill Norton wrote:

I was a dumb young kid on my first single person professional programming project. I had fun programming a spash page of Uncle Fester from the Addams Family, and indeed, the project festered.

Dave Katz was being kind in his description - he bailed me out and as project leader and re-programmed the whole thing himself in a few weeks after I failed to make robust code over a bunch of months. But the splash screen looked good ;-)

The project truly festered.

Bill



On Apr 4, 2012, at 6:42 PM, dennis.oreilly wrote:

I did indeed work with Mark and Dennis in 1987 or so to get the HIM running (almost called it a NIM, which was something else, somewhere in my past...)  I do clearly remember when it first started working, as it led to the following exchange between them:  "Beauty, eh?"  "Woh!"  They also stood slack-jawed at one of the giant plate glass windows in the CC watching an incoming tornado before being encouraged to head for the basement.

I had forgotten completely about that tornado situation.  Thanks for the memory Dave.

Dennis (in Tokyo)

Comments