Newsgroups: alt.folklore.computers Reply-To: sarr@umich.edu Organization: University of Michigan Subject: The Dartmouth Time Sharing System From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 91 Message-ID: Date: Wed, 27 Dec 2000 02:08:14 GMT NNTP-Posting-Host: 207.75.146.69 X-Trace: news.itd.umich.edu 977882894 207.75.146.69 (Tue, 26 Dec 2000 21:08:14 EST) NNTP-Posting-Date: Tue, 26 Dec 2000 21:08:14 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!newsxfer.eecs.umich.edu!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr Xref: chonsp.franklin.ch alt.folklore.computers:71541 A long time ago Lars Poulsen wrote: > >How many of these other languiage systems were available to the >time-sharing users? I have the impression that the system had >a time-sharing personality and a batch processor personality >co-existing but quite different from each other. I would be >interested to heare more about those architectural aspects of >the system. I have some free time over the holidays, so let me take a crack at the original DTSS (also known as the GE-265 and as "Phase 1"; the system that ran for some 30 years on larger GE and Honeywell hardware was Phase 3). First, some ASCII art: +-------------+ | Dual Access | | 6mb disk | +-------------+ | | | | +------------+ +------------+ | GE-235 |<-CIU->| Datanet-30 | +------------+ +------------+ | | | | | | | | | | | Tapes, cards, Teletypes (tm) etc Yes, the disk was 6 megabytes. And weighed 2200 pounds. The Datanet-30 (D-30) was a simple real time machine that could support up to 128 very simple terminal interfaces (simple as in they required processor intervention 5 times per _bit_) and a single interrupt driven by a programmable timer. It collected input lines and wrote them to a terminal buffer on the disk (if they had line numbers) or executed the command (if it didn't). The "run" command was the only one that involved the 235 (a half truth, the text editing commands that started with a "$" were sort of faked run commands). Communication with the 235 was by leaving stuff on the disk, and through the Computer Interface Unit (CIU). The CIU allowed the D-30 to read and write anywhere in 235 memory (since the 235 had 20 bit parity memory and the D-30 had 18 bit no parity memory there was a 3 bit register for the extra 3 bits; yes the D-30 could write bad parity to the 235). Several addresses were reserved as mailboxes for communication. The 235 had 16k (20 bit words) of memory in two 8k banks. There was no memory protection, but by convention the top 2k of the upper bank was OS code and the bottom 2k was OS data. The remaining 6k pieces were "language system" (compiler and/or runtime) code (upper bank) and data (lower bank). The latter was christened the "6k area" and each terminal had a dedicated 6k region of the disk for swapping it. The "language system" code (high bank) was pure and didn't swap. Doing a "run" involved writing the current program text into the 6k area on disk and dropping a message to the 235 with the location of the 6k area on disk and the language. The 235 OS would read in the language system and the 6k area and start (at a specific address) the language system. Depending on the language the system would interpret (Lisp) or compile into RAM and then execute the user program. Or both, in the case of an assembler that generated real code by interpreted it during execution. Time slicing was managed by the Datanet-30. At the end of the time slice it would leave a message for the 235 (which checked 60 time/sec) telling it to swap out the current process. The 235 would write the 6k area back to disk and leave a reply message. The D-30 would look for output as a rubout (x7F) terminated string at a fixed address in the 6k area, and start the next user's time slice. If the 235 didn't respond to the end of a time slice the D-30 would fill 235 memory with branches to a fixed location and then store a boot program at that location. The "batch system" was a special time slice that had no terminal or 6k area associated with it and for which the 235 OS shrank itself down a bit more. Adding a language system involved getting permission, allocating disk space to hold the code, and adding it to some tables. I assume this will provoke more questions than it answers. Fire away; I'll respond as best I can. Do recall that I'm operating from 30+ year old memory here. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 JSTOR, University of Michigan http://www-personal.umich.edu/~sarr/ 301 E Liberty, Ann Arbor, MI 48109-2262 ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: Wed, 27 Dec 00 13:47:48 GMT Organization: UltraNet Communications, Inc. Lines: 37 Message-ID: <92cvuk$bd3$4@bob.news.rcn.net> References: <3A497C40.13DF8B9E@tnglwood.demon.co.uk> X-Trace: UmFuZG9tSVZukoh6y4zhYSMlhZxy9HsUGHU1aJ+VgmEXebldVA/kkO2oMlBI5FDa X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 27 Dec 2000 14:56:20 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!newspeer1.nac.net!netnews.com!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-157 Xref: chonsp.franklin.ch alt.folklore.computers:71494 In article <3A497C40.13DF8B9E@tnglwood.demon.co.uk>, "Bob Billing (AKA Uncle Bob)" wrote: >"Sarr J. Blumson" wrote: > >> Adding a language system involved getting permission, allocating disk space >> to hold the code, and adding it to some tables. > >I remember this thing from my student days. There was a machine in >London and we used to dial in with a teletype and 110 baud modem - Post >Office Telephones in those days, it was before BT. > >We had a thing called GC319 mnemonecode as well as the normal Fortran, >Algol and Basic. This was a simple assembler used for teaching. The >assembler and interpreter was a Fortran program. You editied your >sources as if they were Fortran, then ran the package and off it all >went. > >It took about two days to master it IIRC, then the rest of the year was >spent sitting at the teletype helping the girls^W other students run >their programs. I remember some of those days. The girls only wanted guy help (I could tell by counting the eye bat rate) and the guys didn't trust girl help if it was in front of other guys (unless there were other objectives involved--then they didn't care about public help displays). > >Nostalgia isn't what it used to be... Well, I'll never have to wrestle with that interpreter again. /BAH Subtract a hundred and four for e-mail. ###### From: "The Bakers" Newsgroups: alt.folklore.computers References: Subject: Re: The Dartmouth Time Sharing System Lines: 29 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Message-ID: Date: Wed, 27 Dec 2000 05:15:50 GMT NNTP-Posting-Host: 12.79.175.131 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 977894150 12.79.175.131 (Wed, 27 Dec 2000 05:15:50 GMT) NNTP-Posting-Date: Wed, 27 Dec 2000 05:15:50 GMT Organization: AT&T Worldnet Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!isdnet!newsfeed.icl.net!colt.net!news.maxwell.syr.edu!wn4feed!worldnet.att.net!135.173.83.71!wnfilter1!worldnet-localpost!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71493 Sarr, thanks for the great description of the time sharing system. That was very interesting. I've always been fascinated by the GE Datanet (tm?) communications controllers. Were there later versions with substantial enhancements, or were they all just variations on the basic scheme you described ? For example, some sort of interactive data entry involving DTMF touchtone detection perhaps [just blue-skying here --- I'm not sure how voice prompts might have been done]. I assume that, for connections involving more than just a hard-wired RS-232 interface....e.g. dial-up or leased line, the telco line termination/data modems were external, or were they integrated in somehow ? Was there ever support for higher speeds, bisync, etc. ? How were the Datanet boxes programmed .... presumably by GE, or perhaps some shops did their own systems programming ? Was there a specialized assembler language for this purpose, or maybe some higher-order language (or cross-compiler, cross-assembler, etc.) ? Was there an equivalent to TCAM or VTAM which could be used by, say, COBOL applications programs ? The Datanet equipment always seemed to be ahead of its time, compared to IBM 2705 etc. teleprocessing equipment. Or am I mistaken on this ? Did Univac or Burroughs or NCR or one of the other seven dwarfs have anything even comparable ? Perhaps some specialized stuff IBM used for the SABRE (or pre-SABRE) airline reservation & ticketing ? Oh well, you expected more questions :-) ###### From: "Bob Billing (AKA Uncle Bob)" Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: Wed, 27 Dec 2000 05:21:04 +0000 Message-ID: <3A497C40.13DF8B9E@tnglwood.demon.co.uk> References: NNTP-Posting-Host: tnglwood.demon.co.uk X-NNTP-Posting-Host: tnglwood.demon.co.uk:158.152.132.30 X-Trace: news.demon.co.uk 977907408 nnrp-04:13417 NO-IDENT tnglwood.demon.co.uk:158.152.132.30 X-Complaints-To: abuse@demon.net X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i586) X-Accept-Language: en MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 26 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.tele.dk!212.74.64.35!colt.net!dispose.news.demon.net!demon!news.demon.co.uk!demon!tnglwood.demon.co.uk!falstaff.tanglewood!nobody Xref: chonsp.franklin.ch alt.folklore.computers:71540 "Sarr J. Blumson" wrote: > Adding a language system involved getting permission, allocating disk space > to hold the code, and adding it to some tables. I remember this thing from my student days. There was a machine in London and we used to dial in with a teletype and 110 baud modem - Post Office Telephones in those days, it was before BT. We had a thing called GC319 mnemonecode as well as the normal Fortran, Algol and Basic. This was a simple assembler used for teaching. The assembler and interpreter was a Fortran program. You editied your sources as if they were Fortran, then ran the package and off it all went. It took about two days to master it IIRC, then the rest of the year was spent sitting at the teletype helping the girls^W other students run their programs. Nostalgia isn't what it used to be... -- I am Robert Billing, Christian, inventor, traveller, cook and animal lover, I live near 0:46W 51:22N. http://www.tnglwood.demon.co.uk/ "It burned me from within. It quickened; I was with book as a woman is with child." CS Lewis - Till we have faces, Ch 21. ###### From: lwinson@bbs.cpcn.com (lwin) Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: 28 Dec 2000 03:19:36 GMT Organization: The PACSIBM SIG BBS Lines: 22 Message-ID: <92ebg8$h7d@netaxs.com> NNTP-Posting-Host: bbs.cpcn.com Originator: root@bbs.cpcn.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!arclight.uoregon.edu!news-xfer.newsread.com!bad-news.newsread.com!netaxs.com!newsread.com!netaxs.newsread.com!bbs.cpcn.com!root Xref: chonsp.franklin.ch alt.folklore.computers:71681 x-no-archive: yes The GE Timesharing system was the first computer I ever programmed. If nothing else, I have nostalgia for it. I must admit it's pretty amazing to see how M/S took plain old BASIC and turned it into the basis shell for their Office software and even a fancy language.. And it's pretty amazing that the Teletype ASR we used to access it cost more to buy than a PC that had the BASIC loaded on it. When I got my first PC and brought up BASIC, I realized I didn't have to worry about someone else wanting to use the terminal, or waiting for a slow listing to print out (a dot matrix printer beats a teletype for speed). (But I missed the yellow paper, the chunka-chunk, and the flying paper tapes). I considering getting a Teletype to add to my PC (I found one available). But I'd have to get a "black box", and hope everything would work together. Since I had no way to mechanically adjust or maintain the TTY, I decided to let it go. And I moved on from GW BASIC to QuickBasic4.5, enjoying the power of a compiler. ###### From: "Michael A." Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: Thu, 28 Dec 2000 09:17:34 -0500 Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 X-Complaints-To: newsabuse@supernews.com Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!newshub2.rdc1.sfba.home.com!news.home.com!cyclone-sjo1.usenetserver.com!news-out.usenetserver.com!cyclone-west.rr.com!news.rr.com!news-west.rr.com!sn-xit-03!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71708 > If the 235 didn't respond to the end of a time slice the D-30 would fill > 235 memory with branches to a fixed location and then store a boot > program at that location. I always had the impression that the D-30 was the "master" of the system. During my high school years, sometimes our GE-255 (GECOS) would get hung, so one of us students would sneak over to the computer room and toggle the big bootstrap switch on the D-30, provoking a big buzz alarm and a full system reset. This always seem to fix *our* immediate problem, but of course we weren't the least bit concerned about the 30 or so teletype users that we just put into the ditch . A few of them would call in to complain, and then our instructor would get after us. Eventually the computer room was made off-limits, so we no longer had the fun of hearing the big buzz or typing "MON" on the master teletype to snoop on what everybody else was up to. We had to stay in the adjacent (boring) classroom and work on our GAP assembler programs :-( Mike ###### From: mwilson@the-wire.com (Mel Wilson) Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Message-ID: References: Lines: 96 X-Newsreader: VSoup v1.2.9.37Beta [95/NT] Date: Thu, 28 Dec 2000 10:24:58 -0500 NNTP-Posting-Host: 205.206.39.43 X-Trace: nnrp1.uunet.ca 978017212 205.206.39.43 (Thu, 28 Dec 2000 10:26:52 EST) NNTP-Posting-Date: Thu, 28 Dec 2000 10:26:52 EST Organization: UUNET Canada News Reader Service Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!isdnet!209.249.123.233.MISMATCH!xfer10.netnews.com!netnews.com!newsfeed.direct.ca!look.ca!news.uunet.ca!nnrp1.uunet.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71704 In article , "The Bakers" wrote: >I've always been fascinated by the GE Datanet (tm?) communications >controllers. I only knew them through the on-line banking system I worked on for CGE (in Canada.) I saw a 225/30 Timesharing system, in the computer room our project moved into, but it was soon to move out. Never used it. > Were there later versions with substantial enhancements, or >were they all just variations on the basic scheme you described ? For >example, some sort of interactive data entry involving DTMF touchtone >detection perhaps [just blue-skying here --- I'm not sure how voice prompts >might have been done]. I assume that, for connections involving more than >just a hard-wired RS-232 interface....e.g. dial-up or leased line, the telco >line termination/data modems were external, or were they integrated in >somehow ? Was there ever support for higher speeds, bisync, etc. ? I only ever saw one version -- three modules each roughly 2m high, 1m wide and less than 1m deep. The CPU, control panel and paper tape reader occupied the rightmost module, the comm line interfaces had the other two. CPU used 16K(?) 18-bit words and probably a clock around 1MHz, with 2? 3? 4? clocks per instruction. Probably implemented in discrete transistors, like the GE-400s were. I suspect the comm line interfaces could be patched to coexist with various kinds of modems. We used a bank of 2400-baud half-duplex synchronous modems on multi-drop lines to Olivetti TC349b banking terminals. In our application the DN-30 only assembled message packets and buffered them for the GE-425 that processed the transactions. The DN-30 was obsolescent by 1970 when I encountered it. It had done the heavy communications lifting for GE-200, GE-400 and GE-635 GECOS timesharing systems, but when the GE-655/H6000 was announced by Honeywell around 1972, the line included a DN-355 comm processor implemented with ICs, and DN-30s weren't seen much afterward. There were no DSPs then, and a 1MHz transistorized processor won't do a whole lot, so DTMF would probably be done in hardware, and would look like stuffing 127 touch-tone phones, sans cases, into the leftmost two modules. >How were the Datanet boxes programmed .... presumably by GE, or perhaps some >shops did their own systems programming ? Was there a specialized assembler >language for this purpose, or maybe some higher-order language (or >cross-compiler, cross-assembler, etc.) ? The DN-30 booted from paper tape, with a reader and a big tape bin on the front panel. It was programmed in assembler. When our Datanet programmer wanted a change, she shipped a deck to the plant in Oklahoma City. They had a DN-30 hooked to a DSS-204 disk[*] which ran the assembler, and they sent back a paper tape. Once you spliced the ends together, slid the first bit into the reader, and dropped the rest into the bin the change was installed. We hobnobbed quite a lot, while she was teaching me assembly programming, and I must have seen a bit of DN-30 assembly code, but I don't remember what it looked like. Like GE-200 code, perhaps. I'm surprised now that there was no cross-assembler, but I guess that was not how we did things then. Computers were rare. Cross-assembling on a GE-400 would have been easy for us (we had one), but in Ok City it would have meant begging a favour from a department that had a 400, booking time, getting bumped for emergency jobs, and finally getting assembly output after about a week. > Was there an equivalent to TCAM or >VTAM which could be used by, say, COBOL applications programs ? With anything I saw, datacomm was either done by one-off black magic (we used custom drivers with the Magnetic Tape Programming System) or was run through the operating system, as with timesharing and GECOS-III. >The Datanet equipment always seemed to be ahead of its time, compared to IBM >2705 etc. teleprocessing equipment. Or am I mistaken on this ? [...] GE was big on remote processing; remember the ad "Unite your people, Unite your systems ..."? and they gave a lot of attention to data comm. Somebody was complaining last year about a benchmark where GE submitted a system using timesharing that did the work in hours? days?, and the competition did a normal program that took minutes. I might sort of imagine a GE rejoinder ".. but will their program run in minutes *FROM TOKYO*?" (Although, most likely GE just didn't give enough effort to the benchmark.) Regards. Mel. [*] DSS-204: a fixed 20-platter movable-head monster roughly as big as two phone booths, back when they had phone booths. ###### Newsgroups: alt.folklore.computers References: Reply-To: sarr@umich.edu Organization: University of Michigan Subject: Re: The Dartmouth Time Sharing System From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 26 Message-ID: <20O26.11351$O5.310567@news.itd.umich.edu> Date: Thu, 28 Dec 2000 21:11:26 GMT NNTP-Posting-Host: 207.75.146.69 X-Trace: news.itd.umich.edu 978037886 207.75.146.69 (Thu, 28 Dec 2000 16:11:26 EST) NNTP-Posting-Date: Thu, 28 Dec 2000 16:11:26 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr Xref: chonsp.franklin.ch alt.folklore.computers:71734 In article , Michael A. wrote: > >I always had the impression that the D-30 was the "master" of the system. >During my high school years, sometimes our GE-255 (GECOS) would get hung, so >one of us students would sneak over to the computer room and toggle the big >bootstrap switch on the D-30, provoking a big buzz alarm and a full system >reset. If it really was a 255, then it wasn't a GECOS machine. GECOS was the standard OS for the GE-6x5 machines (and had it's name changed to GCOS when the Computer Department was sold do Honeywell). The 255 is a good story, though. As I said the 265 was a GE-235 plus a Datanet-30. The GE-225 was the 235's older, slower sibling, and GE sold a product using it, the 255. Unfortunately, they never could overcome some small instruction set differences between the two machines, so a 255 became a 265 with some delay code added to the OS to slow it down. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 JSTOR, University of Michigan http://www-personal.umich.edu/~sarr/ 301 E Liberty, Ann Arbor, MI 48109-2262 ###### Newsgroups: alt.folklore.computers References: Reply-To: sarr@umich.edu Organization: University of Michigan Subject: Re: The Dartmouth Time Sharing System From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 50 Message-ID: Date: Thu, 28 Dec 2000 21:21:33 GMT NNTP-Posting-Host: 207.75.146.69 X-Trace: news.itd.umich.edu 978038493 207.75.146.69 (Thu, 28 Dec 2000 16:21:33 EST) NNTP-Posting-Date: Thu, 28 Dec 2000 16:21:33 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!newsxfer.eecs.umich.edu!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr Xref: chonsp.franklin.ch alt.folklore.computers:71735 In article , Mel Wilson wrote: > >> Were there later versions with substantial enhancements, or >>were they all just variations on the basic scheme you described ? For >>example, some sort of interactive data entry involving DTMF touchtone >>detection perhaps [just blue-skying here --- I'm not sure how voice prompts >>might have been done]. I assume that, for connections involving more than >>just a hard-wired RS-232 interface....e.g. dial-up or leased line, the telco >>line termination/data modems were external, or were they integrated in >>somehow ? Was there ever support for higher speeds, bisync, etc. ? We had a bank of grey boxes supplied by New England Telephone that (I think, I'm real fuzzy on this part) converted the current loop from the Teletypes to something close to RS-232. Later on some early Bell 103 modems showed up, in similar packages. > I only ever saw one version -- three modules each roughly 2m >high, 1m wide and less than 1m deep. The CPU, control panel and >paper tape reader occupied the rightmost module, the comm line >interfaces had the other two. CPU used 16K(?) 18-bit words and >probably a clock around 1MHz, with 2? 3? 4? clocks per >instruction. Probably implemented in discrete transistors, like >the GE-400s were. The physical description is pretty accurate; on ours I think the disk interface and the interface to the 235 were also in the middle box. 16k 18 bit words, no parity, with a 7.9 (?) microsecond read cycle. I couldn't tell you what the clock was, but the 235 was actually made of somewhat newer technology and _it_ was only 1 mhz, IIRC. And an instruction on the 235 took at least 12 clocks. > The DN-30 booted from paper tape, with a reader and a big tape >bin on the front panel. It was programmed in assembler. When our >Datanet programmer wanted a change, she shipped a deck to the >plant in Oklahoma City. They had a DN-30 hooked to a DSS-204 >disk[*] which ran the assembler, and they sent back a paper tape. >Once you spliced the ends together, slid the first bit into the >reader, and dropped the rest into the bin the change was >installed. We had a disk interface, so we could do our own assemblies, and load most of the system from disk so the paper tape was only a boot loader. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 JSTOR, University of Michigan http://www-personal.umich.edu/~sarr/ 301 E Liberty, Ann Arbor, MI 48109-2262 ###### From: "Nico de Jong" Newsgroups: alt.folklore.computers References: <92ebg8$h7d@netaxs.com> Subject: Sv: The Dartmouth Time Sharing System Lines: 31 Organization: farum data aps MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Message-ID: <%c_26.43$dR2.2452@news.get2net.dk> Date: Fri, 29 Dec 2000 11:58:13 +0100 NNTP-Posting-Host: 130.228.39.97 X-Complaints-To: abuse@uni2.dk X-Trace: news.get2net.dk 978087867 130.228.39.97 (Fri, 29 Dec 2000 12:04:27 MET) NNTP-Posting-Date: Fri, 29 Dec 2000 12:04:27 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!neel.uni2.net!news.get2net.dk!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71758 > On 28 Dec 2000 03:19:36 GMT, lwinson@bbs.cpcn.com (lwin) wrote: > >And it's pretty amazing that the Teletype ASR we used to access it cost > >more to buy than a PC that had the BASIC loaded on it. --- In my younger days, I was an operator on a genuine RJE/3780 connection from Denmark to stockholm. The 3780 simulator was a danish mini called Regnecentralen (United used them for something) It wass in fact a Data General in disguise, with a TTY attached as operator console. I remember that I "programmed" the WRU drum to do a login sequence to Stockholm. As it was an unused drum, I just had to break off pieces of metal.... It reminds me of the days I programmed IBM punchcard systems like the 088 Jeez, those were the days... > >I considering getting a Teletype to add to my PC (I found one > >available). But I'd have to get a "black box", and hope everything > >would work together. Since I had no way to mechanically adjust or > >maintain the TTY, I decided to let it go. And I moved on from GW BASIC > >to QuickBasic4.5, enjoying the power of a compiler. > > The "black box" is an RS-232/20mA current loop converter, which > requires a maximum of 2 transistors - there are several versions > floating around the 'net and elsewhere. --- Not much of a problem. Almost any radio ham (apart from me) has diagrams floating around Nico ###### From: "The Bakers" Newsgroups: alt.folklore.computers References: Subject: Re: The Dartmouth Time Sharing System Lines: 15 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Message-ID: Date: Fri, 29 Dec 2000 03:30:31 GMT NNTP-Posting-Host: 12.79.175.73 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc07-news.ops.worldnet.att.net 978060631 12.79.175.73 (Fri, 29 Dec 2000 03:30:31 GMT) NNTP-Posting-Date: Fri, 29 Dec 2000 03:30:31 GMT Organization: AT&T Worldnet Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!news-peer.gip.net!news.gsl.net!gip.net!news.maxwell.syr.edu!wn4feed!worldnet.att.net!135.173.83.71!wnfilter1!worldnet-localpost!bgtnsc07-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71762 Thanks to Mel & Sarr for the detailed replies to my questions. After thinking about this a little more, it seems like the Datanet box's relationship with the main GE CPU was somewhat similar to the joint functionality of the Control Data 6600 PP (Peripheral Processor) to the CP (Central Processor). At least that is what comes to my mind. Of course IIRC, a CDC 6600 had multiple PPs connected to its CP. Were any of the GE systems equipped with multiple Datanet units interconnected with a single CPU, or was it always 1:1 ? ###### From: "Michael A." Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: Thu, 28 Dec 2000 23:44:16 -0500 Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <20O26.11351$O5.310567@news.itd.umich.edu> X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 X-Complaints-To: newsabuse@supernews.com Lines: 30 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news-fra1.dfn.de!gatel-ffm!newsfeed.wirehub.nl!newsgate.cistron.nl!fr.clara.net!heighliner.fr.clara.net!grolier!btnet-peer0!btnet-peer!btnet!feeder.qis.net!sn-xit-02!sn-post-02!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71768 > If it really was a 255, then it wasn't a GECOS machine. >... > a 255 became a 265 with some delay code added to the OS to slow it down. Hmmm...now I'm really confused . Originally it was a GE-225, and then there was a significant upgrade done. I remember an earlier model communications processor (Datanet-15?) which was replaced by a Datanet-30. At that point the staff began to refer to the system as a GE-255. Timesharing service was expanded to many more users, both local and remote, with BASIC, Fortran, and Algol available. I remember hearing the word GECOS at an early date (the reference manuals were green with a white border, if that means anything), but I could be confused since I also did some programming on their Honeywell 6000 (GCOS) a few years later. I know for sure that we weren't DTSS, although I remember hearing a lot about Dartmouth at that time and was curious about what that "other" timesharing system was like. There was never any knowledge of the interesting "delay code" imparted to us at that time (perhaps it was GE top-secret ), but the system seemed to run very well for that era, especially considering the amount of memory and disk. The staff didn't mind us getting into the games library, so we frittered away hours playing Football and Horses, printing banners, etc. The system also offered a rather large assortment of math, statistics, and engineering programs for use by timesharing users. Mike ###### From: David Razler Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Reply-To: david.razler@worldnet.att.net Message-ID: References: <92ebg8$h7d@netaxs.com> X-Newsreader: Forte Agent 1.8/32.548 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 44 Date: Fri, 29 Dec 2000 09:52:08 GMT NNTP-Posting-Host: 12.79.66.74 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 978083528 12.79.66.74 (Fri, 29 Dec 2000 09:52:08 GMT) NNTP-Posting-Date: Fri, 29 Dec 2000 09:52:08 GMT Organization: AT&T Worldnet Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.nextra.ch!newsfeed1.online.no!nextra.com!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn4feed!worldnet.att.net!135.173.83.71!wnfilter1!worldnet-localpost!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71785 On 28 Dec 2000 03:19:36 GMT, lwinson@bbs.cpcn.com (lwin) wrote: >x-no-archive: yes > >The GE Timesharing system was the first computer I ever programmed. >If nothing else, I have nostalgia for it. > >I must admit it's pretty amazing to see how M/S took plain old BASIC and >turned it into the basis shell for their Office software and even a >fancy language.. Well, if they can have Visual COBOL.... > >And it's pretty amazing that the Teletype ASR we used to access it cost >more to buy than a PC that had the BASIC loaded on it. When I got my >first PC and brought up BASIC, I realized I didn't have to worry about >someone else wanting to use the terminal, or waiting for a slow listing >to print out (a dot matrix printer beats a teletype for speed). >(But I missed the yellow paper, the chunka-chunk, and the flying paper >tapes). The problem is that TTYs are mechanical devices... >I considering getting a Teletype to add to my PC (I found one >available). But I'd have to get a "black box", and hope everything >would work together. Since I had no way to mechanically adjust or >maintain the TTY, I decided to let it go. And I moved on from GW BASIC >to QuickBasic4.5, enjoying the power of a compiler. The "black box" is an RS-232/20mA current loop converter, which requires a maximum of 2 transistors - there are several versions floating around the 'net and elsewhere. Someone (Doug Jones???) has the entire three-volume Teletype ASR-33 maintenance manual available for d/l. Buy 2 complete ASR-33s and also compile the unofficial Every User's ASR-33 TTY repair set (two paper clips, a roll of Scotch Magic Mending Tape and three rubber bands) and you should be set for life. dmr ###### Message-ID: <3A4CBF28.EF434B85@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System References: <92ebg8$h7d@netaxs.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 42 Date: Fri, 29 Dec 2000 16:48:47 GMT NNTP-Posting-Host: 158.252.50.17 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.prod.itd.earthlink.net 978108527 158.252.50.17 (Fri, 29 Dec 2000 08:48:47 PST) NNTP-Posting-Date: Fri, 29 Dec 2000 08:48:47 PST Organization: EarthLink Inc. -- http://www.EarthLink.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!logbridge.uoregon.edu!cyclone-west.rr.com!news.rr.com!news-west.rr.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!newsmaster1.prod.itd.earthlink.net!newsread1.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71790 David Razler wrote: > The "black box" is an RS-232/20mA current loop converter, which > requires a maximum of 2 transistors - there are several versions > floating around the 'net and elsewhere. And trivial to make. As recently as a couple years ago, I attached a 33 with one of these to an old 103 type acoustic coupler (Novation Cat) (I actually have a late 60's Anderson Jacobson acoustic coupler which supports both RS232 and current loop but its having some problems) and logged into my shell account over the packet network to Delphi. Unfortunately, Delphi has terminated their dial in access and although the shell accounts currently live, they're on a short string as the machines are not being maintained apparently (they're no longer charging for the existing accounts either). > Someone (Doug Jones???) has the entire three-volume Teletype ASR-33 > maintenance manual available for d/l. Buy 2 complete ASR-33s me know where you find them, I need a spare - also looking for Model > 1n ( RO-14, RO-19, KSR-14 KSR-19) TTYs (the wire service models, and a > BRPE tape punch, especially> and also compile the unofficial Every > User's ASR-33 TTY repair set (two paper clips, a roll of Scotch Magic > Mending Tape and three rubber bands) and you should be set for life. You forgot the electrical tape. My KSR 33's rubber "hammer" disintegrated and I replaced it with a tightly wound roll of "cloth" electrical tape held on with a couple of turns of the tape around both the roll and the metal bracket. Works fine. Unfortunately, before I got the unit, the bare metal bracket had hammered the type drum more or less flat and I've never found a replacement. Unfortunately its since developed other problems and although I have all the manuals, I have no spare parts and haven't had the time to look at trying to make them from scratch. Chris AN GETTO$;DUMP;RUN,ALGOL,TAPE $$ ###### From: mwilson@the-wire.com (Mel Wilson) Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Message-ID: References: Lines: 19 X-Newsreader: VSoup v1.2.9.37Beta [95/NT] Date: Fri, 29 Dec 2000 16:27:31 -0500 NNTP-Posting-Host: 205.206.39.114 X-Trace: nnrp1.uunet.ca 978127065 205.206.39.114 (Fri, 29 Dec 2000 16:57:45 EST) NNTP-Posting-Date: Fri, 29 Dec 2000 16:57:45 EST Organization: UUNET Canada News Reader Service Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!newsfeeds.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sunqbc.risq.qc.ca!torn!news.uunet.ca!nnrp1.uunet.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71808 In article , "The Bakers" wrote: > Were any of the GE >systems equipped with multiple Datanet units interconnected with a single >CPU, or was it always 1:1 ? On the GE-400, the Computer Interface Unit just connected to one of the peripheral channels. And this may have been true on the 635 as well... a spin through the 600/6000 Mainframes & Software booklet doesn't give a hint of any special treatment for the DN-30. The black magic software we used for the interface only did one datanet, but I spent some time breaking out the important data fields and adapting it for two; the old listings turned up a couple of months ago. IIRC the modification tested OK, but there were worries about the capacity of the 400. We wound up converting to H6000 with the Transaction Driven System instead. Regards. Mel. ###### From: Eric Chomko Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: 30 Dec 2000 04:19:29 GMT Organization: IDT Internet Services Lines: 60 Message-ID: <92jnoh$fl6@nnrp1.farm.idt.net> References: <92ebg8$h7d@netaxs.com> <3A4CBF28.EF434B85@earthlink.net> NNTP-Posting-Host: u2.farm.idt.net X-Newsreader: TIN [UNIX 1.3 unoff BETA release 961025] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!newsfeeds.belnet.be!news.belnet.be!news.stealth.net!newsfeed.wirehub.nl!newspeer1.nac.net!netnews.com!newsin.iconnet.net!news.idt.net!nntp.farm.idt.net!u2.farm.idt.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71832 jchausler wrote: : David Razler wrote: : > The "black box" is an RS-232/20mA current loop converter, which : > requires a maximum of 2 transistors - there are several versions : > floating around the 'net and elsewhere. : And trivial to make. As recently as a couple years ago, I attached a 33 : with one of these to an old 103 type acoustic coupler (Novation Cat) : (I actually have a late 60's Anderson Jacobson acoustic coupler which : supports both RS232 and current loop but its having some problems) and : logged into my shell account over the packet network to Delphi. : Unfortunately, Delphi has terminated their dial in access and although the : shell accounts currently live, they're on a short string as the machines are : not being maintained apparently (they're no longer charging for the : existing accounts either). Two things here. SWTPC had a serial card the MP-S that has both a 20mA current loop circuit AND a RS-232 circuit on the same card. That circuit (probably on www.swtpc.com, but I am not certain) might be a good starting place for the above application (black box). The second thing is my ISP plans on terminating dial-up access soon. Does anyone know of an ISP that still provides shell account service along with WWW service? : > Someone (Doug Jones???) has the entire three-volume Teletype ASR-33 : > maintenance manual available for d/l. Buy 2 complete ASR-33s me know where you find them, I need a spare - also looking for Model : > 1n ( RO-14, RO-19, KSR-14 KSR-19) TTYs (the wire service models, and a : > BRPE tape punch, especially> and also compile the unofficial Every : > User's ASR-33 TTY repair set (two paper clips, a roll of Scotch Magic : > Mending Tape and three rubber bands) and you should be set for life. : You forgot the electrical tape. My KSR 33's rubber "hammer" disintegrated : and I replaced it with a tightly wound roll of "cloth" electrical tape held : on with : a couple of turns of the tape around both the roll and the metal bracket. : Works fine. Unfortunately, before I got the unit, the bare metal bracket had : hammered the type drum more or less flat and I've never found a replacement. : Unfortunately its since developed other problems and although I have all the : manuals, I have no spare parts and haven't had the time to look at trying to : make them from scratch. What about those nice canary paper rolls? Are they available? : Chris : AN GETTO$;DUMP;RUN,ALGOL,TAPE : $$ Is the above sort of like: (read eval print) in Lisp? Eric ###### Sender: prep@k9.prep.synonet.com Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System References: <92ebg8$h7d@netaxs.com> <3A4CBF28.EF434B85@earthlink.net> <92jnoh$fl6@nnrp1.farm.idt.net> From: Paul Repacholi Date: 30 Dec 2000 20:44:45 +0800 Message-ID: <874rzm84ea.fsf@k9.prep.synonet.com> Lines: 17 User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 158.d01.pe.iqnet.net.au X-Trace: 30 Dec 2000 20:17:01 +0800, 158.d01.pe.iqnet.net.au Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!newshub2.rdc1.sfba.home.com!news.home.com!sjc1.nntp.concentric.net!newsfeed.concentric.net!newsfeed.ozemail.com.au!news.mel.connect.com.au!news.per.connect.com.au!newsfeed.iinet.net.au!news.waia.asn.au!usenet.per.paradox.net.au!127.0.0.1!nobody Xref: chonsp.franklin.ch alt.folklore.computers:71844 Eric Chomko writes: > : > Someone (Doug Jones???) has the entire three-volume Teletype ASR-33 > : > maintenance manual available for d/l. Buy 2 complete ASR-33s What about those nice canary paper rolls? Are they available? NC macine shops. Ask, and they will tell you where to get them. -- Paul Repacholi 1 Crescent Rd., +61 (08) 9257-1001 Kalamunda. West Australia 6076 Raw, Cooked or Well-done, it's all half baked. ###### Message-ID: <3A4E3F68.91B16FB3@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System References: <92ebg8$h7d@netaxs.com> <3A4CBF28.EF434B85@earthlink.net> <92jnoh$fl6@nnrp1.farm.idt.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 69 Date: Sat, 30 Dec 2000 20:09:05 GMT NNTP-Posting-Host: 158.252.49.68 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.prod.itd.earthlink.net 978206945 158.252.49.68 (Sat, 30 Dec 2000 12:09:05 PST) NNTP-Posting-Date: Sat, 30 Dec 2000 12:09:05 PST Organization: EarthLink Inc. -- http://www.EarthLink.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!news-stu1.dfn.de!news-mue1.dfn.de!news-was.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!cpk-news-hub1.bbnplanet.com!news.gtei.net!news-peer.gip.net!news.gsl.net!gip.net!news-out.usenetserver.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!newsmaster1.prod.itd.earthlink.net!newsread1.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71868 Eric Chomko wrote: > Two things here. SWTPC had a serial card the MP-S that has both a 20mA > current loop circuit AND a RS-232 circuit on the same card. That circuit > (probably on www.swtpc.com, but I am not certain) might be a good starting > place for the above application (black box). True, I have a number of these being a 6800 junkie but they all use 4N33 opto- isolators between the TTL and RS232/current loop sides of the circuit. (Actually I think they stole the circuit from Motorola as Moto used them on their D1 evaluation kit which I believe pre-dated the SWTPC designs.) Although the opto doesn't hurt, it isn't necessary unless you have some kind of grounding problem. All it takes is a couple of transistors some resistors, a capacitor and of course some power source with ground, plus and minus. If you're using TTL you can use the +5 for the positive 232 voltage. > The second thing is my ISP plans on terminating dial-up access soon. > Does anyone know of an ISP that still provides shell account service along > with WWW service? I asked Delphi that same question and did not get an answer. Although their service is still there, as soon as the machines fail, they say it will be gone for good. Tis a sad day. Real men use shell accounts ;-) > What about those nice canary paper rolls? Are they available? Don't know. I still have about a half dozen of them (some yellow, some white paper. More than I'll ever use. I also have quite a stash of ribbons, unfortunately, they're all dried out. (Adding a little light oil or lighter fluid to them can rejuvenate them for a short while. One used to be able to by bottles of re-inking solution but with dot matrix printers now rare (although still available as a customer or mine bought a number of Oki wide carriage printers a year or so ago.) It's just CompUSA or other "retail" places don't seem to carry it anymore. I haven't even seen ribbons for the popular Epson wide carriage printers lately. I have several including the original MX-100 but have not used them in the last several years either. > : AN GETTO$;DUMP;RUN,ALGOL,TAPE > : $$ > > Is the above sort of like: > > (read eval print) > > in Lisp? I don't know Lisp but the above is: 1. read the rest of the edit buffer into the scratch buffer 2. write out the scratch buffer as the new version of the file 3. start the Algol compile/load/go sequence using the scratch buffer for input. They were the last two lines (the following $$ was a job deck terminator) one would have either on punch cards or type into an rje TTY system for a "deck" which was a set of "batch" editing commands. The AN stood for AND, Alpha Numeric Director, which was a combination primitive file system and batch editor. Each user account got up to 63 files, stored by file number. IIRC the 0 files was like the .bak file for that user for his or her last edit. Chris AN GETTO$;DUMP;RUN,ALGOL,TAPE $$ ###### From: jsavard@ecn.ab.SBLOK.ca.nowhere (John Savard) Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Message-ID: <3a4ee9c2.23813277@news.powersurfr.com> References: X-Newsreader: Forte Free Agent 1.11/32.235 Lines: 15 Date: Sun, 31 Dec 2000 08:10:58 GMT NNTP-Posting-Host: 24.108.206.34 X-Complaints-To: abuse@powersurfr.com X-Trace: news-rep.ab.videon.ca 978250244 24.108.206.34 (Sun, 31 Dec 2000 01:10:44 MST) NNTP-Posting-Date: Sun, 31 Dec 2000 01:10:44 MST Organization: Videon CableSystems Alberta Inc. Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!torn!nntp.cs.ubc.ca!cyclone.bc.net!rover.ucs.ualberta.ca!news-rep.ab.videon.ca!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:71933 On Wed, 27 Dec 2000 02:08:14 GMT, sarr@engin.umich.edu (Sarr J. Blumson) wrote, in part: >Communication with the 235 was by leaving stuff on the disk, and through >the Computer Interface Unit (CIU). While that certainly seems odd, and slow and artificial, and all that... this is not only an intriguing way to set up a computer, but I wonder if doing things that way might make it easier to achieve a high degree of security against stuff like buffer overflow attacks. John Savard http://home.ecn.ab.ca/~jsavard/crypto.htm ###### Newsgroups: alt.folklore.computers From: weiner@world.std.com (Sam Weiner) Subject: Re: The Dartmouth Time Sharing System Message-ID: Date: Tue, 2 Jan 2001 04:12:07 GMT References: Organization: The World Public Access UNIX, Brookline, MA Lines: 37 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!fu-berlin.de!news.maxwell.syr.edu!wn4feed!worldnet.att.net!198.6.0.7!uunet!ash.uu.net!dfw.uu.net!ffx.uu.net!world!weiner Xref: chonsp.franklin.ch alt.folklore.computers:72015 In article , The Bakers wrote: >Sarr, thanks for the great description of the time sharing system. That was >very interesting. > >I've always been fascinated by the GE Datanet (tm?) communications >controllers. Were there later versions with substantial enhancements, or >were they all just variations on the basic scheme you described ? For >example, some sort of interactive data entry involving DTMF touchtone >detection perhaps [just blue-skying here --- I'm not sure how voice prompts >might have been done]. I assume that, for connections involving more than >just a hard-wired RS-232 interface....e.g. dial-up or leased line, the telco >line termination/data modems were external, or were they integrated in >somehow ? Was there ever support for higher speeds, bisync, etc. ? Others have already mentioned the DN-355 follow on. I don't know if they were compatible or not. One interesting thing I recall, though it would be nice to have someone else confirm it, is the DN-355 (or some DN-xxx) was used as a disk controller for GE-400/-600 systems at some point. >How were the Datanet boxes programmed .... presumably by GE, or perhaps some >shops did their own systems programming ? Was there a specialized assembler >language for this purpose, or maybe some higher-order language (or >cross-compiler, cross-assembler, etc.) ? Was there an equivalent to TCAM or >VTAM which could be used by, say, COBOL applications programs ? Not precisely on topic but the folks in Phoenix used a DN-30 as a standalone TRAC language timesharing system. Just the DN-30, no other computer. [snip] Sam (who's first hands on via TTY33 was to a GE-265 run by Call A Computer which was owned by Pillsbury - yes, the flour folks.) ###### From: Eric Chomko Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: 2 Jan 2001 16:55:45 GMT Organization: IDT Internet Services Lines: 91 Message-ID: <92t16h$ibi@nnrp3.farm.idt.net> References: <92ebg8$h7d@netaxs.com> <3A4CBF28.EF434B85@earthlink.net> <92jnoh$fl6@nnrp1.farm.idt.net> <3A4E3F68.91B16FB3@earthlink.net> NNTP-Posting-Host: u3.farm.idt.net X-Newsreader: TIN [UNIX 1.3 unoff BETA release 961025] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news.maxwell.syr.edu!news.idt.net!nntp.farm.idt.net!u3.farm.idt.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:72052 jchausler wrote: : Eric Chomko wrote: : > Two things here. SWTPC had a serial card the MP-S that has both a 20mA : > current loop circuit AND a RS-232 circuit on the same card. That circuit : > (probably on www.swtpc.com, but I am not certain) might be a good starting : > place for the above application (black box). : True, I have a number of these being a 6800 junkie but they all use 4N33 opto- : isolators between the TTL and RS232/current loop sides of the circuit. : (Actually I think they stole the circuit from Motorola as Moto used them on : their D1 evaluation kit which I believe pre-dated the SWTPC designs.) : Although the opto doesn't hurt, it isn't necessary unless you have some : kind of grounding problem. All it takes is a couple of transistors some : resistors, a capacitor and of course some power source with ground, : plus and minus. If you're using TTL you can use the +5 for the positive : 232 voltage. Do you have a diagram of it? : > The second thing is my ISP plans on terminating dial-up access soon. : > Does anyone know of an ISP that still provides shell account service along : > with WWW service? : I asked Delphi that same question and did not get an answer. Although : their service is still there, as soon as the machines fail, they say it will : be gone for good. Tis a sad day. Real men use shell accounts ;-) Agreed. I'd like to have that service still. : > What about those nice canary paper rolls? Are they available? : Don't know. I still have about a half dozen of them (some yellow, : some white paper. More than I'll ever use. I also have quite a : stash of ribbons, unfortunately, they're all dried out. (Adding a : little light oil or lighter fluid to them can rejuvenate them for a : short while. One used to be able to by bottles of re-inking : solution but with dot matrix printers now rare (although still Wait, I have a bottle of ink for resupplying the ink jet cartridges of HP type printers. That should work. : available as a customer or mine bought a number of Oki wide : carriage printers a year or so ago.) It's just CompUSA or other : "retail" places don't seem to carry it anymore. I haven't even : seen ribbons for the popular Epson wide carriage printers : lately. I have several including the original MX-100 but have : not used them in the last several years either. Yes, I have some older ink ribbons as well and they DO dry out! Some TTY paper as well. : > : AN GETTO$;DUMP;RUN,ALGOL,TAPE : > : $$ : > : > Is the above sort of like: : > : > (read eval print) : > : > in Lisp? : I don't know Lisp but the above is: : 1. read the rest of the edit buffer into the scratch buffer : 2. write out the scratch buffer as the new version of the file : 3. start the Algol compile/load/go sequence using the scratch buffer for : input. : They were the last two lines (the following $$ was a job deck terminator) : one would have either on punch cards or type into an rje TTY system for : a "deck" which was a set of "batch" editing commands. The AN stood : for AND, Alpha Numeric Director, which was a combination primitive : file system and batch editor. Each user account got up to 63 files, stored : by file number. IIRC the 0 files was like the .bak file for that user for : his or her last edit. On what system? Seems as encrypted as some Exec 8 stuff on UNIVAC 1100s with their '@' (at sign) preceeding each command, or worse, JCL from IBM 360/370. I heard a guy the other day saying 'thank God for Perl', as he was fed up with Korn Shell programming. I snickered and thought about the difference between JCL and Korn Shell vs. Korn Shell and Perl. Eric : Chris : AN GETTO$;DUMP;RUN,ALGOL,TAPE : $$ No, not at all like Lisp. ###### Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System References: Reply-To: sarr@umich.edu Organization: University of Michigan, College of Engineering From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 28 Message-ID: Date: Tue, 02 Jan 2001 19:04:03 GMT NNTP-Posting-Host: 207.75.146.69 X-Trace: news.itd.umich.edu 978462243 207.75.146.69 (Tue, 02 Jan 2001 14:04:03 EST) NNTP-Posting-Date: Tue, 02 Jan 2001 14:04:03 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!tungurahua!news-ge.switch.ch!news-fra1.dfn.de!news-lei1.dfn.de!news-was.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr Xref: chonsp.franklin.ch alt.folklore.computers:72062 In article , Sam Weiner wrote: > >Others have already mentioned the DN-355 follow on. I don't know >if they were compatible or not. One interesting thing I recall, >though it would be nice to have someone else confirm it, is the >DN-355 (or some DN-xxx) was used as a disk controller for GE-400/-600 >systems at some point. Maybe the 400s, although I don't think so. The 600s used this monster box called the GIOC (?) to connect to peripherals. And I'm pretty sure the 400 used the same peripherals. >Not precisely on topic but the folks in Phoenix used a DN-30 >as a standalone TRAC language timesharing system. Just the DN-30, >no other computer. The Dartmouth system would run quite happily without the 235 as long as nobody typed "run," and the D-30 could run a language like TRAC (tm) that required very little numeric computation quite nicely. But I think the particular D-30 you have in mind had some interesting one off hardware on it, including an I/O interface that could talk to 400/600 peripherals. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 JSTOR, University of Michigan http://www-personal.umich.edu/~sarr/ 301 E Liberty, Ann Arbor, MI 48109-2262 ###### From: Tom Van Vleck Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: Tue, 02 Jan 2001 21:54:56 -0500 Organization: MindSpring Enterprises Lines: 16 Message-ID: <3A529480.CB3536EE@multicians.org> References: NNTP-Posting-Host: 04.36.29.30 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Server-Date: 3 Jan 2001 02:51:30 GMT X-Mailer: Mozilla 4.76 [en] (Win98; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeeds.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!sn-xit-03!supernews.com!cyclone-west.rr.com!news.rr.com!news-west.rr.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!news.mindspring.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:72085 "Sarr J. Blumson" wrote: > Maybe the 400s, although I don't think so. The 600s used this monster box > called the GIOC (?) to connect to peripherals. And I'm pretty sure the 400 > used the same peripherals. 635 would have used an IOC. GIOC (Generalized I/O Controller) was only used on the GE-645 for Multics: it did terminals as well as tapes and disks. Only a few were made. See "Communications and Input/Output Switching in a Multiplex Computing System" by Ossanna, Mikus, and Dunten http://www.multicians.org/fjcc5.html for tons of detail about the GIOC. Later versions of Multics on the 6000 line used a standard IOM and DN-355, running custom Multics software. ###### Message-ID: <3A534E02.A5493AB7@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System References: <92ebg8$h7d@netaxs.com> <3A4CBF28.EF434B85@earthlink.net> <92jnoh$fl6@nnrp1.farm.idt.net> <3A4E3F68.91B16FB3@earthlink.net> <92t16h$ibi@nnrp3.farm.idt.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 103 Date: Wed, 03 Jan 2001 16:11:59 GMT NNTP-Posting-Host: 168.191.124.221 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.prod.itd.earthlink.net 978538319 168.191.124.221 (Wed, 03 Jan 2001 08:11:59 PST) NNTP-Posting-Date: Wed, 03 Jan 2001 08:11:59 PST Organization: EarthLink Inc. -- http://www.EarthLink.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!fu-berlin.de!newsfeed.direct.ca!look.ca!cyclone-sjo1.usenetserver.com!news-out.usenetserver.com!cyclone-west.rr.com!news.rr.com!news-west.rr.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!newsmaster1.prod.itd.earthlink.net!newsread2.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:72083 Eric Chomko wrote: > Do you have a diagram of it? Not at hand but from memory, TTL to current: resistor to base of NPN transistor, emitter to ground, collector thought resistor to base of PNP transistor. Its emitter to positive voltage supply and collector through resistor to one side of current loop. Other side of current loop to minus voltage. I recall needing a capacitor between base and collector of PNP to stop oscillations under certain conditions. current to TTL, one side of current loop to minus voltage the other to two resistors in series to positive voltage. The two resistors in series make a voltage divider. Base of PNP transistor to between the voltage divider resistors. Emitter of PNP transistor to plus voltage. Collector to resistor to base on NPN transistor and from base of NPN a second resistor to ground (another voltage divider). Emitter of NPN to ground, collector through resistor to +TTL (5 volts) and to output. Now just choose your resistor values appropriately :-) If the right plus and minus voltages are used, an almost identical circuit can be used for TTL to 232 with minor differences in the resistors and the circuit on the current loop /232 side. This is just for the output side the input side for 232 is different. Depending on the TTL circuit you might need another stage of logic inversion between it and these circuits and you definitely will need one in one case and not in the other if you are going to use if for both 232 and current loop. > Wait, I have a bottle of ink for resupplying the ink jet cartridges of HP > type printers. That should work. That would be worth the try for sure. > Yes, I have some older ink ribbons as well and they DO dry out! Some TTY > paper as well. The "ink" is still there, just not in a liquid form anymore. > : > : AN GETTO$;DUMP;RUN,ALGOL,TAPE > : > : $$ > : I don't know Lisp but the above is: > : 1. read the rest of the edit buffer into the scratch buffer > : 2. write out the scratch buffer as the new version of the file > : 3. start the Algol compile/load/go sequence using the scratch buffer for > : input. > : They were the last two lines (the following $$ was a job deck terminator) > : one would have either on punch cards or type into an rje TTY system for > : a "deck" which was a set of "batch" editing commands. The AN stood > : for AND, Alpha Numeric Director, which was a combination primitive > : file system and batch editor. Each user account got up to 63 files, stored > : by file number. IIRC the 0 files was like the .bak file for that user for > : his or her last edit. > > On what system? Seems as encrypted as some Exec 8 stuff on UNIVAC 1100s > with their '@' (at sign) preceeding each command, or worse, JCL from IBM > 360/370. A Bendix (later CDC) G-20, actually a pair of them, at CMU (CIT :-) in the 60's. However, the AND system not to mention the Algol compiler were written at the school and I believe the "monitor" was as well (or at least highly modified). Since I've never been able to determine any other G-20 installation, I don't know whether any of this was shared or in common with other G-20 sites. Never did use Exec 8 but spent a fair amount of time with Exec II on an 1108. A basic batch system with I/O stream spooling. This was needed at the 1108 could keep its two card readers and three printers running and still be idle a lot of the time. As a part time operator while a student, I would occasionally run a job on the adjacent 360/67 which would read in 1108 jobs and copy them to a tape. This tape could then be mounted on the 1108 as a third source of input. Although normally for use when one of the 1108 card readers (actually 1004's) were dead, I would also use it during "student project week" when the job load was large and most of the jobs did a "compile, load, crash" (if they got beyond the compile stage that is :-). At these times the 1108 was frequently done with the job before the readers had completed sucking in the next job card. An input spooler for the 1108 was also the best use I ever found for the 360/67 ;-) Exec II "JCL" was quite straight forward and I believe Exec 8 was similar. We only had one 64K word memory on the 1108 which wasn't really enough to run Exec 8 and further in the late 60's, Exec 8 wasn't really there yet. One would use the Exec II "Complex Utility Routines, i.e. XQT CUR, in much the same way as the AND system on the G-20. CUR provided a batch edit function which you could follow with a compile, load, go step in your favorite language. Instead of the large fixed disks (or the RCA RACE) of the G-20's, the file system for the 1108 was on a Fastrand II drum. Both also supported files on magnetic tape although the G-20's was a lot nicer as it was a "stringy disk" just like DECTape. > I heard a guy the other day saying 'thank God for Perl', as he was fed up > with Korn Shell programming. I snickered and thought about the difference > between JCL and Korn Shell vs. Korn Shell and Perl. The young'uns these days don't know what they have (and don't know what they've missed either). Chris AN GETTO$;DUMP;RUN,ALGOL,TAPE $$ ###### From: nouce@multics.ruserved.com (Richard Shetron) Newsgroups: alt.folklore.computers Subject: Re: The Dartmouth Time Sharing System Date: 4 Jan 2001 20:55:21 -0500 Organization: X-no-Archive: yes Lines: 71 Message-ID: <9339i9$4n5$1@multics.ruserved.com> References: <20O26.11351$O5.310567@news.itd.umich.edu> NNTP-Posting-Host: ruserved.com X-Original-NNTP-Posting-Host: ruserved.com X-Trace: 4 Jan 2001 20:55:24 -0500, ruserved.com XPident: nouce X-Original-NNTP-Posting-Host: 199.181.141.3 XPident: Unknown Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!merapi!news-ge.switch.ch!news.maxwell.syr.edu!newspeer.phoen-x.net!news.wizvax.net!multics.ruserved.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:72311 Several people mentioned varoius GE-600/H-6000 items about configuration. The following is from a Honeywell brochure (The Multics System (C) 1975,1976 Honeywell, part AK27, Rev2). Some of the information is from other sources that I don't have handy so it my flakey memory ;). I worked for HIS in the late 70's at RADC/Griffis AFB so this is based on the information/configurations at that time. Some of the items have different names from what have been mentioned, sometimes its because they are acutally different hardware doing the same job. The general descriptions should be correct for everything from the oldest 600 to the 'newest' L66/L68 systems. CPU's had 4 ports to connect the CPU to System Controllers. Later CPU's starting with the later 6000 models could have cache. I think the early cache was 4KW and later models could have 16KW. System Control Units had 8 ports that could be connected to either CPU's or IOM's (Input/Output Multiplexors). Each system controller sat on main memory. The 600 series and older 6000 ones had a max of 256KW (36 bit words or 1MB) of core (500ns cycle) per SCU. I know the later SCU's for semiconductor memory could have much more memory, at least 16MW/controller. IOM's had 4 ports for connecting to SCU's. They had plugin slots for connecting to other devices, such as DN355's, 6600FNPs, disk controllers, tape controllers, system console, etc. The original 600 series box was a GIOC IIRC. DN355's, 6600FNP's were 'terminal' processors connecting to many terminals (slow speed devices). tape and disk controllers were the same controller with different interface hardware specific to what they controlled. A typical Multics system might have 2 to 4 CPU's, 2 to 4 SCU's, 1 IOM, 1 system console, 1 disk controller, 1 tape controller, 1 unit record processor (printers, card readers, card punchs). NOTE that every CPU connected to every SCU and every IOM also connected to every SCU. The maximum 'size' of a system was a combinations of 8 CPUs and IOM's. GCOS could run on a Multics CPU, but Multics could not run on a GCOS CPU at this time. Multics CPU's included virtual memory and other additional hardware support. Under Multics, any system resource that was not needed, ie CPU, SCU, IOM, etc could be taken offline while the system was running. Some installations had a 'large' hardware system, but actually ran multiple smaller systems by allocating the hardware to run multiple systems. RADC where I worked for Honeywell had 2 CPU's, 2 IOM's, and the appropriate hardware to share tape drives between 2 systems so they ran 1 Multics system and 1 GCOS system at the same time. Hardware was 'switched' between the two OS's by taking it offline in one and then online in the other while both systems were running. SCU's hid the physical details of the memory. The orignal core boxes had a max of 256KW (1MB) of 500ns core (as seen by the processors/IOM's). When the first semiconducter memory was introduced, it had a 750ns cycle. There was an article on alt.os.multics that indicated that the later Multics systems could address up to 4TB of virtual memory and one site needed more so modified the OS to handle 6TB. Those wanting more information about Multics and the hardware should look at http://www.multicians.org/ -- Richard Shetron multics@ruserved.com multics@acm.rpi.edu NO UCE LEGAL NOTICE:Sender of UCE to this address agrees to pay me $500/email plus any and all costs of colleciton.