Article 1979 of alt.sys.pdp10: Path: nntp1.ba.best.com!news2.best.com!news-hog.berkeley.edu!ucberkeley!enews.sgi.com!fido.engr.sgi.com!rigden.engr.sgi.com!rpw3 From: rpw3@rigden.engr.sgi.com (Rob Warnock) Newsgroups: alt.sys.pdp10 Subject: Re: Q: Why not (2^n)-bit? Date: 16 Nov 2000 12:33:13 GMT Organization: Silicon Graphics Inc., Mountain View, CA Lines: 141 Message-ID: <8v0k69$hveg7$1@fido.engr.sgi.com> References: <8ti24q$rks$1@nntp1.ba.best.com> <8tsb1f$1c9s$1@nntp1.ba.best.com> NNTP-Posting-Host: rigden.engr.sgi.com X-Trace: fido.engr.sgi.com 974377993 18856455 163.154.34.115 (16 Nov 2000 12:33:13 GMT) X-Complaints-To: news@fido.engr.sgi.com NNTP-Posting-Date: 16 Nov 2000 12:33:13 GMT Xref: nntp1.ba.best.com alt.sys.pdp10:1979 Joe Smith wrote: +--------------- | Toots Mutant wrote: | >...misfortune of writing a DZ-11 driver for 20F...[bad] the way a DZ-11 | >worked. It was made even worse when that software was marketed by a company | >that liked to blame the software for hardware problems in their DZ | >emulators. | | DCA - Digitial Communications Associates of Alpharetta, Georgia. | | We used their terminal multiplexors on the Colorado School of Mines | campus. Clusters of 4 to 8 terminals in the various buildings were | connected to a single box, that in turn connected the terminal to either | our KS-2020 or our KL-1091. The hardware that plugged into the KS's | Unibus emulated a DZ11. +--------------- Yes, indeed. Actually, it emulated up to *16* DZ-11's!! (128 serial ports) I should know, since I designed that board! Or rather, I architected it and managed the 3-person team that designed/built it. It was basically a Zilog Z-80 8-bit microprocessor, with a portion of its main RAM used as 16 sets of emulated DZ-11 registers. Unibus PIO reads or writes to "DZ-11 registers" caused a DMA on the Z-80's bus into the corresponding portion of the Z-80's RAM, as well as an interrupt to the Z-80, which then compared the emulated-register portion of RAM with a backup shadow copy to see what had changed. Most of the "DZ-11 register" state transitions could be emulated in software by the Z-80, but a few things had to be handled "RIGHT NOW!" during the Unibus PIO (a.k.a. Z-80 DMA). Those were done with a small amount of external "hard" logic that masked the RAM data. That is, we didn't bother doing a read-modify-write DMA cycle to the RAM, we just stored an external hint in hard logic that modified the next Unibus PIO read or write. For example, it is an absolute requirement that if one Unibus read gets the last character in the receive FIFO, then another immediately following Unibus read of the receive FIFO location must *NOT* return the "valid" bit (bit <15>). So any such read had bit <15> AND'd with an external hard-logic "receive FIFO ready bit" for that DZ-11, turned off the "ready" bit at the end of the DMA... and also interrupted the Z-80. After processing the interrupt, dequeueing the next input byte (if any) and putting it in the "receive FIFO" offset in RAM, the Z-80 would turn on the external "receive FIFO ready bit" for that DZ-11 again (which was, of course, also coupled to the DZ-11 command/status word and the Unibus interrupt request). That was the DCA "SmartMux/205", if I recall the model number correctly. A single quad-connector Unibus card, which emulated 16 DZ-11's (128 ports), and which connected with the outside world through a single RS-232c synchronous serial cable running DCA's proprietary terminal multiplexing protocol. [Which, note, was based on DEC's DDCMP protocol, which we had implemented from the public spec and shipped to customers before DEC did! (*Tee-hee!*) And by the way, a tip-o'-the-hat to DEC's Stu Wecker for writing a *great* protocol document!!] We sold a whole bunch of those to DEC-20 and VAX customers, and later, to a few TOPS-10 KL customers as well (after "PDP-10" I/O busses stopped being available on KLs). Other miscellaneous trivia: - It was the first board DCA had ever built using the then-brand-new "Multiwire" technology, wherein you gave a chip layout and a netlist to the board vendor, and an automated machine laid insulated *wires* down on the board (pushed into an adhesive coat). Since they were insulated, wires could cross over/under each other. After all the wires were laid down, some more glue was spread on, baked hard, and then the holes for the chip pins and other discrete components were drilled *through* the board, the glue (epoxy), and the wires, and copper was additively plated onto the inside of the holes (thus connecting to the wire ends exposed by the drilling). Finally, parts were stuffed, and the board was wave-soldered. Anyway, it was a new (to us) technology, and the schedule was tight, so we were *very* careful about checking the design and the netlist we gave them. Wonder of wonders, the very first board we got back worked the first time!!! (The 2nd failed, but it was a Multiwire process glitch.) All in all, some 11 or 12 of the first 15 just worked! - At the time we designed the /205, DEC had a patent on the Unibus bus arbitration logic, and if you wanted to build a Unibus device, you *had* to infringe on the patent -- it was fundamental to Unibus arbitration. But the per-instance (or per-item) license fee they wanted was simply *outrageous* [we thought], $100 or 10% of the item's retail price, whichever was higher (IIRC). However... At the same time, the DECdirect catalog openly sold DEC's own Unibus arbitration chip -- which contained the patented circuitry inside it -- as a "spare part" to anyone who wanted to buy it. And it was cheap, just a few bucks each. Since patents cover *manufacture* of the patented item, by buying DEC's own chip to use in the SmartMux/205, we could completely legally avoid paying the outrageous patent fee. [Or to look at it another way, we really "paid the fee", and got the chip for free!] Oh, and AFAIK, DEC's catalog sales people never bothered to ask why we kept buying so many of those "spares"... ;-} ;-} +--------------- | But for the KL connection, we were still using the 680I, a PDP-8 based | solution that was the successor to the DC68. +--------------- A real DEC 680I? ...or a DCA "SmartMux/250", which used an OEM'd PDP-8 all right, but with DCA-designed UART-based serial ports instead of the 680I's "software bit banging", and as a result could be populated with up to 128 9600-baud lines! That was DCA's first product, and the mainstay of the business for a long time. It interfaced to the PDP-10 I/O bus through a "bus converter" I had designed while I was working at Emory University, which design DCA picked up (along with me, as it happened). Originally we used the normal DC68 driver on the host, but as we got asked for more features, I wrote a new driver from scratch that made SCNSER think we were a DC76. Well, sort of a DC76/DC68 hybrid, actually. Anyway, close enough that we were able to support "remote echo" on terminals at the other end of our synchronous multiplexer links... +--------------- | If I remember correctly, DCA did not have a functional DH11 emulator | suitable for plugging in to the KL's console front end. +--------------- Kee-rect. The DZ-11 emulator was the only Unibus device we ever built (AFAIK, but I left in '79, so they might have done something later). +--------------- | The kludge of putting a DZ11 equivalent into RSX-20F was not satisfactory. +--------------- Perhaps not for you (sorry!), but some customers found it quite acceptable. -Rob ----- Rob Warnock, 31-2-510 rpw3@sgi.com Network Engineering http://reality.sgi.com/rpw3/ Silicon Graphics, Inc. Phone: 650-933-1673 1600 Amphitheatre Pkwy. PP-ASEL-IA Mountain View, CA 94043 Article 1980 of alt.sys.pdp10: Path: nntp1.ba.best.com!news1.best.com!newsfeed.mathworks.com!portc03.blue.aol.com!cyclone2.usenetserver.com!news-out.usenetserver.com!news01.optonline.net!news02.optonline.net.POSTED!not-for-mail Message-ID: <3A15842A.5CE3FCFC@bartek.net> From: Arthur Krewat X-Mailer: Mozilla 4.7 [en] (X11; I; SunOS 5.8 i86pc) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.sys.pdp10 Subject: Re: Q: Why not (2^n)-bit? References: <8ti24q$rks$1@nntp1.ba.best.com> <8tsb1f$1c9s$1@nntp1.ba.best.com> <8v0k69$hveg7$1@fido.engr.sgi.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 58 Date: Fri, 17 Nov 2000 19:21:36 GMT NNTP-Posting-Host: 24.190.213.53 X-Trace: news02.optonline.net 974488896 24.190.213.53 (Fri, 17 Nov 2000 14:21:36 EST) NNTP-Posting-Date: Fri, 17 Nov 2000 14:21:36 EST Organization: Optimum Online Xref: nntp1.ba.best.com alt.sys.pdp10:1980 Rob Warnock wrote: > That was the DCA "SmartMux/205", if I recall the model number correctly. At BOCES III in Dix Hills, NY, I worked on five KS10's, each with a DCA 305. Like you said, capable of 128 ports, Unibus based. > A single quad-connector Unibus card, which emulated 16 DZ-11's (128 ports), > and which connected with the outside world through a single RS-232c > synchronous serial cable running DCA's proprietary terminal multiplexing > protocol. [Which, note, was based on DEC's DDCMP protocol, which we had > implemented from the public spec and shipped to customers before DEC did! > (*Tee-hee!*) And by the way, a tip-o'-the-hat to DEC's Stu Wecker for > writing a *great* protocol document!!] What speed was that serial link running at? We had setup 10 cross-over ports at 19200 between the 5 systems (4+3+2+1). I wrote a file transfer protocol and away she went! We never noticed any degradation of performance (of our 9600 baud video terminals). The file transfer daemon did a dance, though boy! > Other miscellaneous trivia: > > +--------------- > | But for the KL connection, we were still using the 680I, a PDP-8 based > | solution that was the successor to the DC68. > +--------------- > > A real DEC 680I? ...or a DCA "SmartMux/250", which used an OEM'd PDP-8 > all right, but with DCA-designed UART-based serial ports instead of the > 680I's "software bit banging", and as a result could be populated with > up to 128 9600-baud lines! That was DCA's first product, and the mainstay > of the business for a long time. It interfaced to the PDP-10 I/O bus > through a "bus converter" I had designed while I was working at Emory > University, which design DCA picked up (along with me, as it happened). My first experience with a PDP-10 was with a KA10 with a DCA/PDP-8 front-end. I am told that it had 128 ports in it, and says DCA on the front (my friend still has it in his basement). Based on a PDP-8/i? It's serial # 45. Interesting to learn that the UARTs in it are not standard DEC equipment :) > +--------------- > | The kludge of putting a DZ11 equivalent into RSX-20F was not satisfactory. > +--------------- > > Perhaps not for you (sorry!), but some customers found it quite acceptable. I found the DCA equipment quite revolutionary in terms of mux'ing the serial lines. When all the bugs were worked out of the OS, it was rock-solid. There were many problems at first, but I think it was mainly the Z80 rebooting itself for no reason. Eventually, field service decided to replace the CPU and it worked flawlessly ever-after. I don't remember the model # of the base unit, but the boards in the KS10's were definitly 305's (I think). IIRC, TOPS-10 6.03 had to be modified to recognize the 305's, or otherwise clear up some conflict. ak Article 2000 of alt.sys.pdp10: Path: nntp1.ba.best.com!news1.best.com!nntp.primenet.com!nntp.gblx.net!enews.sgi.com!fido.engr.sgi.com!rigden.engr.sgi.com!rpw3 From: rpw3@rigden.engr.sgi.com (Rob Warnock) Newsgroups: alt.sys.pdp10 Subject: DCA muxes [was: Re: Q: Why not (2^n)-bit? ] Date: 19 Nov 2000 08:49:08 GMT Organization: Silicon Graphics Inc., Mountain View, CA Lines: 90 Message-ID: <8v8464$5dtco$1@fido.engr.sgi.com> References: <8tsb1f$1c9s$1@nntp1.ba.best.com> <8v0k69$hveg7$1@fido.engr.sgi.com> NNTP-Posting-Host: rigden.engr.sgi.com X-Trace: fido.engr.sgi.com 974623748 5698968 163.154.34.115 (19 Nov 2000 08:49:08 GMT) X-Complaints-To: news@fido.engr.sgi.com NNTP-Posting-Date: 19 Nov 2000 08:49:08 GMT Xref: nntp1.ba.best.com alt.sys.pdp10:2000 D. R. Banks wrote: +--------------- | "Rob Warnock" wrote: | > Joe Smith wrote: | > +--------------- | > | The kludge of putting a DZ11 equivalent into RSX-20F was not | satisfactory. | > +--------------- | > | > Perhaps not for you (sorry!), but some customers found it quite | acceptable. | | No, it wasn't satisfactory, and I wrote the driver to enable people | to do it. | | There were severe memory limitations in the RSX-20F front end that forced | some pretty unacceptable tradeoffs. +--------------- As I said, I'm sorry you had trouble with it. As another posted said, some *did* find it useful, though. +--------------- | The 205 board also had this maddening little habit of hanging, particularly | if you put two or more of them in adjacent slots on a unibus. Believe it or | not, many (most?) problems with these boards were solved by putting physical | space between the boards on the unibus. +--------------- I could easily believe it. IIRC, there was some sort of metastability problem with some versions of the DEC Unibus arbitration chip that we used (that we bought as "spares", as I mentioned in my previous article). I had thought it eventually had gotten worked out, but maybe not. The symptom *would* have been a "hang", and, yes, it would have been *very* sensitive to where you placed the card in the system. [If my vague memory is correct -- and apologies to DEC if it's not!! -- the problem was a synchronizer metastability issue in the Unibus interrupt grant logic, and such things were just barely beginning to be generally understood in those days, with a lot of misinfomation & myth floating around. We had a *bear* of a time figuring out what the problem was, and an even worse problem getting DEC to admit anything. IIRC, the later, better version of the chip had a "metastability-hardened" flop in the synchronizer role.] +--------------- | Why put more than one 205 board into a CFE when the board would emulate up | to 128 lines, while the (hacked) RSX-20F only supported 80? The link coming | out of the 205 would run, at best, 19.2K. +--------------- Actually, IIRC the original 205 could only do 9600 on the sync port, but another poster said the "305"(?) could do 19,200. I don't remember the latter, so I don't know when the speed bump occurred (in the 205 or 305). +--------------- | Try muxing 32 active terminal sessions across one 19.2K line and you'll | know why multiple boards were common. +--------------- Well, you certainly have a valid point, but application loads do vary. The "data entry clerks + line printer" I mentioned in my other reply ran a dozen 2400-baud terminals and a line printer on a single 2400-baud sync line, and they were perfectly happy with it. Whereas I'm more like you, I suspect. I wouldn't be happy sharing a 19,200-baud line with more than one or two other people, if I was doing heavy editing. Go figure... +--------------- | I wrote the DZ driver for the service bureau I worked for at the time. | We went into it with our eyes open, and were able to work around the | restrictions, the two worst being that the front end would only support 80 | lines (rather than the original 128), and that "Send all" support would be | broken. Unfortunately, a decision was made to recoup some of our costs by | selling the driver back to DCA, who in turn resold the software to others. | Full disclosure of the limitations was not always made, and there was a | pattern of DCA support people blaming faulty boards on the 11's DZ driver. +--------------- Was that after Sept.'79? If so, I was long gone. But I'll still apologize for my former co-workers if they treated you badly. (*sigh*) -Rob ----- Rob Warnock, 31-2-510 rpw3@sgi.com Network Engineering http://reality.sgi.com/rpw3/ Silicon Graphics, Inc. Phone: 650-933-1673 1600 Amphitheatre Pkwy. PP-ASEL-IA Mountain View, CA 94043 Article 1998 of alt.sys.pdp10: Path: nntp1.ba.best.com!news1.best.com!newsfeed.mathworks.com!newsxfer.eecs.umich.edu!logbridge.uoregon.edu!news.indiana.edu!news.iupui.edu!haystack!mhwood From: "Mark H. Wood" Newsgroups: alt.sys.pdp10 Subject: Re: Q: Why not (2^n)-bit? Date: Tue, 21 Nov 2000 02:27:25 GMT Organization: La Petite Hackerie Lines: 23 Message-ID: References: <8ti24q$rks$1@nntp1.ba.best.com> <8tsb1f$1c9s$1@nntp1.ba.best.com> <8v0k69$hveg7$1@fido.engr.sgi.com> NNTP-Posting-Host: mhw.ulib.iupui.edu User-Agent: tin/1.4-19991113 ("No Labels") (UNIX) (Linux/2.2.14 (i586)) Xref: nntp1.ba.best.com alt.sys.pdp10:1998 Rob Warnock wrote: [lots of interesting stuff about DCA 205s] For what it's worth, we used to have quite a large herd of DCA gear, tying terminal clusters all over campus back to 355 switches in the machine room. Then it gets weird: we switched the traffic onto individual serial ports, 128 of which were cabled to cabinets full of DH11s to feed a pair of 2060s (and later to sixteen DECserver 100s feeding a VAX 8800, later still a 7000-620). (Other ports were hooked to a bank of IBM 7171 ADACUs to provide pseudo-3270 access to a 4381, later a 3090.) I've never again seen so much cable in so little space. I had actually wanted to condense all of those sessions onto two X.25 lines run into a pair of DMP11s, but the plan was vetoed as having too many things in it with which we had no experience. What we wound up with greatly resembled what we'd done earlier with Computer Transmission Corporation equipment (the "TRAN gear") and those same DH11s. -- -- Mark H. Wood, radical centrist OpenPGP ID 876A8B75 mhwood@ameritech.net Why is it that, of all the species on Earth, only the presence of Man is considered an intrusion? Article: 15371 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!paloalto-snf1.gtei.net!news.gtei.net!newsfeed.stanford.edu!logbridge.uoregon.edu!gestalt.direcpc.com!cyclone2.usenetserver.com!usenetserver.com!news01.optonline.net!news02.optonline.net.POSTED!not-for-mail From: Arthur Krewat Newsgroups: alt.sys.pdp10,alt.sys.pdp.11,alt.sys.pdp8 Subject: Re: SIMH 2.9 beta release Organization: Kilonet.net Lines: 17 Message-ID: <3C4DFC9E.1BF0C88E@bartek.dontspamme.net> References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.78 [en] (X11; U; SunOS 5.8 i86pc) X-Accept-Language: en Date: Wed, 23 Jan 2002 00:02:40 GMT NNTP-Posting-Host: 24.186.100.134 X-Trace: news02.optonline.net 1011744160 24.186.100.134 (Tue, 22 Jan 2002 19:02:40 EST) NNTP-Posting-Date: Tue, 22 Jan 2002 19:02:40 EST Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15371 alt.sys.pdp8:5225 Bob Supnik wrote: > > Some number of additional terminals won't be a problem, but the PDP-8 > only has 64 device numbers, each discrete terminal interface takes 2, > so 32 individual interfaces aren't possible. I thought the maximum > number supported was 19. Not to argue, but somehow DCA got at least 128 terminals in a PDP-8/i (serial # 45 - probably DCA's serial #) I understand that the UART's or whatever equivalent were custom on this box. This was used as a front end to a KA10 - BOCES/LIRICS - my friend still has the PDP-8/i in his father's basement. And that PDP-8/i combined with the KA10 was faster for terminal I/O than a KS10 ever was. And our KS10's used DCA 355's (?) aak Article: 15384 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!nycmny1-snh1.gtei.net!news.gtei.net!news-out.visi.com!hermes.visi.com!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-245-9 From: jmfbahciv@aol.com Newsgroups: alt.sys.pdp10,alt.sys.pdp.11,alt.sys.pdp8 Subject: Re: SIMH 2.9 beta release Date: Wed, 23 Jan 02 10:55:26 GMT Organization: UltraNet Communications, Inc. Lines: 30 Message-ID: References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> <3C4DFC9E.1BF0C88E@bartek.dontspamme.net> X-Trace: UmFuZG9tSVaOCWTivgWBanNC3rnfZiKuQ8dCJih+oC+QXxiHuq9ySvFAIfYAyVsU X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 23 Jan 2002 13:05:46 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15384 alt.sys.pdp8:5229 In article <3C4DFC9E.1BF0C88E@bartek.dontspamme.net>, Arthur Krewat wrote: >Bob Supnik wrote: >> >> Some number of additional terminals won't be a problem, but the PDP-8 >> only has 64 device numbers, each discrete terminal interface takes 2, >> so 32 individual interfaces aren't possible. I thought the maximum >> number supported was 19. > >Not to argue, but somehow DCA got at least 128 terminals in a PDP-8/i >(serial # 45 - probably DCA's serial #) I understand that the UART's >or whatever equivalent were custom on this box. > >This was used as a front end to a KA10 - BOCES/LIRICS - my friend >still has the PDP-8/i in his father's basement. And that PDP-8/i >combined with the KA10 was faster for terminal I/O than a KS10 >ever was. And our KS10's used DCA 355's (?) > What was LOGMAX for that KA system? I bet it wasn't 128. My guess is that it may have been built for 40..maybe 60 jobs. That only "allows" 15-30 connections through that 8/I. If you want to look at the code that we shipped for PDP-8s, look for DC72s. I can't remember is we put the 680I stuff on the Customer Supported tape. Looking for PAL8.EXE would help, too. /BAH Subtract a hundred and four for e-mail. Article: 15418 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!washdc3-snh1.gtei.net!news.gtei.net!feeder.qis.net!sunqbc.risq.qc.ca!falcon.america.net!cyclone2.usenetserver.com!usenetserver.com!news01.optonline.net!news02.optonline.net.POSTED!not-for-mail From: Arthur Krewat Newsgroups: alt.sys.pdp10,alt.sys.pdp8 Subject: Re: SIMH 2.9 beta release Organization: Kilonet.net Lines: 42 Message-ID: <3C4F5576.475D89BD@bartek.dontspamme.net> References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> <3C4DFC9E.1BF0C88E@bartek.dontspamme.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.78 [en] (X11; U; SunOS 5.8 i86pc) X-Accept-Language: en Date: Thu, 24 Jan 2002 00:37:40 GMT NNTP-Posting-Host: 24.186.100.134 X-Trace: news02.optonline.net 1011832660 24.186.100.134 (Wed, 23 Jan 2002 19:37:40 EST) NNTP-Posting-Date: Wed, 23 Jan 2002 19:37:40 EST Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15418 alt.sys.pdp8:5235 Jim Thomas wrote: > > >>>>> "Arthur" == Arthur Krewat writes: > > Arthur> Bob Supnik wrote: > >> > >> Some number of additional terminals won't be a problem, but the PDP-8 > >> only has 64 device numbers, each discrete terminal interface takes 2, > >> so 32 individual interfaces aren't possible. I thought the maximum > >> number supported was 19. > > Arthur> Not to argue, but somehow DCA got at least 128 terminals in a PDP-8/i > Arthur> (serial # 45 - probably DCA's serial #) I understand that the UART's > Arthur> or whatever equivalent were custom on this box. > > But those were multiple line boards, right? Each DCA board would have used > a few device numbers. Yes, exactly, the DCA serial ports were muxed somehow. The setup was: A High School on Long Island would "pay" a certain amount of money per terminal per year. The school could get state money (NY BOCES) to "pay" for this, so in the end, the school really never wrote a check to anybody. Anyway, on average, there would be at least 4 terminals muxed over a point-to-point line which was usually 4800 baud from the high school to BOCES III in Dix Hills. If you were lucky they would give you an LA120 that could run at 1200 baud and it would be the quickest thing you ever saw. And, I do remember that the LA120 was available when the KA10 was still "the" machine. Originally BOCES had all these high schools (I am unsure of the number, but at least 30 w/4 terminals each) connected through the PDP-8i my friend has in his father's basement, into a KA10, apparently bought used and had 256Kw of core. At the height of it's usage, there were more than 120 jobs logged in on any given day. I learned MACRO that year ('80?). At that time, it seems that they had already been using KS10's for a little while, and eventually moved the schools over to any of four KS10's with a fifth being used for the BOCES Business Office. And, the 8i was used to boot the KA10, so it really was a front-end. aak Article: 15448 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!cambridge1-snf1.gtei.net!news.gtei.net!bos-service1.ext.raytheon.com!cyclone.swbell.net!cyclone-sf.pbi.net!209.10.34.151!newsfeed.sjc.globix.net!hub1.meganetnews.com!hub1.nntpserver.com!xmission!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!Tomobiki-Cho.CAC.Washington.EDU!mrc From: Mark Crispin Newsgroups: alt.sys.pdp10,alt.sys.pdp8 Subject: Re: booting a KA10 from a PDP-8/i? Date: Thu, 24 Jan 2002 12:59:30 -0800 Organization: Pandamonium Reigns Lines: 20 Message-ID: References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> <3C4DFC9E.1BF0C88E@bartek.dontspamme.net> <3C4F5576.475D89BD@bartek.dontspamme.net> <3C4F7247.FDE5224F@Empire.Net> <3C4F7869.6451C4F8@bartek.dontspamme.net> <3C500360.A00B8C00@bartek.dontspamme.net> <3C502D04.73678C85@bartek.dontspamme.net> <51f05u01lldtscr5v4pfgv5j58c18oe8cu@4ax.com> <3C503F7C.CF01F739@bartek.dontspamme.net> NNTP-Posting-Host: tomobiki-cho.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1011905972 20422 (None) 140.142.17.39 X-Complaints-To: help@cac.washington.edu NNTP-Posting-User: btik In-Reply-To: <3C503F7C.CF01F739@bartek.dontspamme.net> Content-Length: 314159 (believe this at your own risk) Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15448 alt.sys.pdp8:5258 On Thu, 24 Jan 2002, Arthur Krewat wrote: > From what I've heard, it was a DMA device. This may be how the 8i got the > bootstrap into the KA10. Remember, this was a DCA device - the mentality > was one of "do everything from the 8i". I once knew the people at DCA fairly well, but I have never heard of them doing such a thing. The issue at hand is that it's pointless. The thing that is bootstraped into the KA10 is not the operating system, it's the disk bootstrap that will load the operating system. Since every KA10 had a paper tape reader and bootstrap built in, it's pointless to do it any other way. I also don't think that even a DCA device was able to write into KA10 memory unilaterally. The KA10 set the DMA addresses, not the 8/i. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. Article: 15445 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!washdc3-snh1.gtei.net!news.gtei.net!feeder.qis.net!news.maxwell.syr.edu!cyclone2.usenetserver.com!usenetserver.com!news01.optonline.net!news02.optonline.net.POSTED!not-for-mail From: Arthur Krewat Newsgroups: alt.sys.pdp10,alt.sys.pdp8 Subject: Re: booting a KA10 from a PDP-8/i? Organization: Kilonet.net Lines: 15 Message-ID: <3C50682A.FE561B12@bartek.dontspamme.net> References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> <3C4DFC9E.1BF0C88E@bartek.dontspamme.net> <3C4F5576.475D89BD@bartek.dontspamme.net> <3C4F7247.FDE5224F@Empire.Net> <3C4F7869.6451C4F8@bartek.dontspamme.net> <3C500360.A00B8C00@bartek.dontspamme.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.78 [en] (X11; U; SunOS 5.8 i86pc) X-Accept-Language: en Date: Thu, 24 Jan 2002 20:07:38 GMT NNTP-Posting-Host: 24.186.100.134 X-Trace: news02.optonline.net 1011902858 24.186.100.134 (Thu, 24 Jan 2002 15:07:38 EST) NNTP-Posting-Date: Thu, 24 Jan 2002 15:07:38 EST Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15445 alt.sys.pdp8:5256 Christopher Stacy wrote: > > > I agree with "Ms 4.88's" that a normal stock KA10 certainly did not > have a PDP-8 front-end, but that doesn't mean yours didn't! > You boot a KA10 by pressing a console switch that reads from the > directly attached paper tape reader. But your site could have hacked > up their system to use a PDP-8 to replace the bootload function; > I don't see anything technically difficult about that at all. I keep mentioning it, but I have to again - this PDP-8/I was a piece of DCA equipment. I have no idea what the original intent of DCA was in designing this thing - it's possible they meant it to boot the KA10. aak Article: 16621 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen!iad-peer.news.verio.net!news.verio.net!iad-read.news.verio.net.POSTED!not-for-mail Newsgroups: alt.sys.pdp10,alt.sys.pdp8 Subject: Re: booting a KA10 from a PDP-8/i? References: <3C4F5576.475D89BD@bartek.dontspamme.net> <3C4F7247.FDE5224F@Empire.Net> Organization: Chez Inwap X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: inwap@inwap.com (Joe Smith) Originator: inwap@inwap.com (Joe Smith) Lines: 44 Message-ID: <6tCn8.4749$T_.95368@iad-read.news.verio.net> Date: Mon, 25 Mar 2002 09:54:10 GMT NNTP-Posting-Host: 130.94.177.175 X-Complaints-To: abuse@verio.net X-Trace: iad-read.news.verio.net 1017050050 130.94.177.175 (Mon, 25 Mar 2002 09:54:10 GMT) NNTP-Posting-Date: Mon, 25 Mar 2002 09:54:10 GMT Xref: dfw-artgen alt.sys.pdp10:16621 alt.sys.pdp8:5345 In article <3C4F7247.FDE5224F@Empire.Net>, John Sauter wrote: >Arthur Krewat wrote (excerpted): > >And, the 8i was used to boot the KA10, so it really was >a front-end. > >Huh? At Sanders we had a KA10 with a PDP-8/i terminal >multiplexor on the I/O bus, but it had nothing to do >with booting the KA10. The KA10 had a hard console >with lots of lights and switches, and booting was done >from its paper tape reader. > John Sauter (J_Sauter@Empire.Net) The 8i was not a front-end. It was merely a device that could respond to RIM (Read-In Mode) which allowed the disk bootstrap program to be loaded from PDP-8 memory instead of from paper tape. Traditional boot: Set RIM device switches to the code for PTR. Put bootstrap tape into the paper tape reader. The first segment on the tape is the RIM10B loader. The second segment can be either a full-fledged BOOTS (a rather long paper tape with code that can accept a file name from the TTY and load the specified file from disk) or a short pre-boot that simply reads raw bootstrap blocks off of a reserved area of disk unit zero. Press the Read-In-Mode switch. Non-paper boot: Set RIM device switches to the code for PDP-8 interface. The disk pre-boot has already been put into the PDP-8's core memory. Press the Read-In-Mode switch. 1) Short read-the-boot-blocks routine gets sent to the KA with the PDP-8 acting like a virtual paper-tape reader. 2) Said routine locates logical disk 0, reads in the first part of the disk bootstrap, executes it (which reads in the rest of the full disk bootstrap code). This method simply saved wear-and-tear on the KA's paper tape reader. -Joe -- See http://www.inwap.com/ for PDP-10 and "ReBoot" pages. Article: 15392 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen.news.verio.net!dfw-peer.news.verio.net!news.verio.net!crtntx1-snh1.gtei.net!cpk-news-hub1.bbnplanet.com!news.gtei.net!nntp.abs.net!feeder.qis.net!sn-xit-02!sn-post-01!supernews.com!corp.supernews.com!not-for-mail From: Bob Supnik Newsgroups: alt.sys.pdp10,alt.sys.pdp.11,alt.sys.pdp8 Subject: Re: SIMH 2.9 beta release Date: Wed, 23 Jan 2002 10:03:28 -0500 Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <60rr4ugmk9ubv7mfnk4tuiu85li24u4j0k@4ax.com> X-Newsreader: Forte Agent 1.8/32.548 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Complaints-To: newsabuse@supernews.com Lines: 42 Xref: dfw-artgen.news.verio.net alt.sys.pdp10:15392 alt.sys.pdp8:5232 My bad, I didn't realize Peter had used Telnet for the console. Re PDP-8 terminal counts - the TSS/8 manuals (for the 8/I) offer different maximum terminal counts. The system operators guide says up to 4 PT08s (dual interfaces), and the rest on DL8/Is, to a maximum of 32 users. The monitor manuals says a mixture of PT08s and 680/Is, to a maximum of 32 users. A later version of the source (EDUSYSTEM-50) says maximum 20 users, with one on the console, and 19 on either a terminal multiplexor, or PT08/KL8's, or a mixture. In any case, PT08s/KL8s consume device numbers at a rate of 2 per line, so the system would be limited by device numbers. Terminal mux's like the DC08 would use only a few device numbers for multiple lines. The 680/I, which was a front end on early -10s, used the DL8/I terminal multiplexor to achieve high line counts. /Bob On 23 Jan 2002 00:28:41 -0800, spedraja@ono.com (Sergio Pedraja) wrote: >Hi again: > >> >> 7. Altair Z80. Compilation problems on Windoze fixed (thanks to Peter >> >> Schorn). >> > >> >I had some problems with this compiling below MINGW. In appeareance >> >it does reference to some wrong/changed library or so. > >I've dicovered the problem under Mingw. Is needed to modify the compilation >line of the AltairZ80 smulator in this way: > >gcc -o ..\bin\altairz80 -I.. ..\scp*.c ..\sim*.c altair*.c -lm -lwsock32 > >It appears to work. > >Greetings > >Sergio