On 6/17/13 7:03 PM, Jonathan Katz wrote: Even closer for me, I live in Valpo, but alas I have little interest having given away a complete PDP-8 some 30 years ago. I could potentially do some investigation for others... --tom

--- On Mon, 6/17/13, steve shumaker <shumaker at att.net> wrote: ? Indeed. They aren't very expandable, but even a fully-loaded PDP-8/E is pretty "useless" these days. But that's not the point. I've often described the 8/L as "the most classic computing goodness crammed into the smallest package". Only the Nova can compete with it. Of course, these are aesthetic judgements with much room for disagreement, thus I may have touched off another discussion of what makes a classic computer "interesting" or "classic"... --Bill

Why do you say that? A PDP8/E will do everything it could do when it was made. And if you happen to do want to do such a task, it's as usegful as ever. I am curiouse as to why you claim that. I don't have an 8/L, only an 8/E, but while I will agree that the PDP8 is an elegannt machine I don't think it's unique in that respect. There are plenty of other machines of the same period with similar elegance to me. FWIW. the programamble machine that I've worked on that I consider to be the most elegant is the HP9100B calculator (never worked on a -A). That machine is glorious inside. -tony

Sure. This may techncially be off-topic (in that the machine was sold as a calcualtor), but since said machines can run user programs, have s storage device, can interface to peripherals, etc, it's hard to think of a sane defintion of 'computer' that excludes them and that doesn't also exclude things that are classed as computers. Anyway, HP calcualtors... There are a couple of websties to look at for more inforamtion : http://www.hpmuseum.org/ is mostly for the handheld machines http://www.hpmuseum.net/ is for the larger machines and has a lot of manuals to download. The HP9100 was HP's first electronic calculator (I specify it like that becuase there were some cpecial slide rules shipped with HP RF instruments to calcualtor reflection coeffficients, etc). It is a classic example of a point I was making in another tread. The component count is quitree small (under 1000 transistors). There are no digital ICs (there are 8 op-amps o nthe magnetic card reader PCB, but nothing in the processor/mmemory sections). The compoonets are themselves simple. But the way they are interconnected is very complex. BAscially, the machine is a special-purpose processor. It's microcoded (64*29 bit core-on-a-rope microcode). The main program ROM is an inductively coupled PCB ROM, with 512 locatiosn of 64 bits each. There's a total of 44 flip-flops in the machine (again very small), 43 JKs (40 for the processor registers, including the program counters for the microcode and main ROM, 2 in the clock divider and oen to select the user program memory page. And a single SR flip-flop to drive the error indicator. The flip-flops are interconencted by diode AND and OR gates (no NOTs, although both the Q and Q/ (E and F in HP's terminology) outputs of the flip-flops are used. Tehre isn;'t even a a ful ladder. To add 2 registers, you incremeent one and decrememtn the other until the second has got to 0 (seriously). User memory is a little core plane, 6 bits wide (I forget how many lcoations). A location can store amongst other thigns, a digit (4 bit BCD + sign bit + blanking bit) or a user program keystroke (there happen to be 64 differnt keycodes...) Amazingly this machine is a full prgramamble ccientific calculator. The dispaly is a little CRT displaying 7 segemt digit patterns. There's a magneitc card reader you cna use to store your programs on and reload them later (although. of course the core memeory is non-volatile) adn a periperhal conenctor that will give you serious headaches if you try to use it. The machine is often claimed ot he RPN. I feel it isn't. There are 3 registers, X, Y,Z which appear to eb a stack, but there is no automatic lift and drop. Keyboard intput goes into X, the 2-numer operators like '+' act as Y:= Y + X, X being unchanged. OF corse there are keys to flit, drop and roll the stack, but that doens't make it RPN to me . OK ,that's the first one (more info on request). The next series are knwon as 98x0 or 9800 machines. They are all based roudn the same 16 bit bit-serial processor built from abotu 80 TLT ICs. The HP9810 is essentially the same ot the user as the 9100. 3 level stack, etc. The memeroy is now volatile 1103 DRAMs, so the magnetic card reader becomes very imporntant. There's slso the option of a built-in thermal strip pritner. Amaxingly, although it uses ICs, the 9810 is larger than the 9100, and doesn't do as mauch in its basic form (there are no transcendental functions, you have to add a ROM module for that). The dispaly is now 3 rows of LED digits (a total of 45 digit positions) for the 3 stack registers. The HP9820 is essentially the same machine but usign an infix notation langaage (so you cna acutally do 2+4 and get the answer 6. Many of the PCBs are the same as those in the 9180, and theothers have closely-related circuitry. It has twice ans much ROM firmware and a 16 character alphnumeric display. The big brother of this family is the HP9830. It has the same processor boards, but everyhtign else is differnt, although somewhat simuilar in design. It has a full QWERTY keyboard, a 32 character alphanumeric display and runs BASIC. It's a fairly simple ABSIC, with no string operations, but it's BASIC. And you could get add-on ROM modules to add strings, matrices, advanced I/O, etc. There's a built in digital cassette drive to store programs and data in numbered files, there was even a hard disk unit as a peripheral. The HP98x5 machines are something of a hotch-pptch. The 9805 is a non-programamble machine that took plug-in modules (the statistcis module is the only remotely common one). I have never seen this machine, I beleive it's closyl related ot e HP46 and HP81 printign calcualtors, and uses the same processor (custom 'Arithmetic and Registers' and 'Control and TIming' ICs). The 9815 is the only old HP calculator to use a stnadard CPU -- a Motorola 6800. This machine is a bit like an overgrown HP97. It's a 4 level stack RPN machine (now with automatic lift and drop), it's prgramamble, it has a built-in printer and cartridge tpae drive. The HP9825 used an HP custom BPC processor module (a ceramic hybrid circuit under a massive heatsink) and runs a language called HPL which is essentaily an extended version of the HP9820 programming language. The HP9845 is a twin-procesosr (user program an I/O) machine running BASIC and with a CRT display. It's pusing it to call that oen a calculator. What more can I tell you? Well that depends on what your are interested in. I could bore you to death by talking aobtu the fetch-execute cycle of the 9800 procesosr, for example. -tony

These old HPs were in general very solidly made and have interesting designs. I certianly like them. Firstly, I ahve to say that the HP80 series is probably my least favourite deskop from HP. Farr too mach custom silicon for my taste :-). That said, I hae a couple of 85s, a couple of 86Bs, an 87 and an 87XM . Oh and a 9915. So perhaps I do like them after all... There are manuals for them on http://www.hpmuseum.net/ if you are lacking aything. In terms of their relationship to earlier machiens, there is very little in common in the hardware (the only machine closely related to the HP85 series is the HP75 handheld computer [1]). But spriritually, yes... It appears that one major use that HP envisaged for their machines was as data loggers/control machines. It makes sense, HP sold measuring instruemnts. One of the first interfaces for the hP9800 series was the 11203 BCD interrace which could read in 10 (I think) digits of BCD from the display drivers of a sigital instruemnrt and present them to the computer as if they'd bneen typed o nthe keyboard. And of course therwas an HPIB itnerface for the hP9830. [1] THe somments in the ROM listings for that machine are hilarious... And the HP85 carried o nthis tradition. Ther wrre BCD and HPIB itneraces for it too. I beleive some large HP instruemtns (gas chromatographs, etc) came with an HP85 as the cotnroller. -tony

--- On Tue, 6/18/13, Tony Duell <ard at p850ug1.demon.co.uk> wrote: Indeed. In a commercial setting, though, the cost of keeping such an old machine operational must be weighed against the cost of replacement. It only makese sense when the machine is embedded in a long-lived system that has not suffered so badly operate and maintain on the whole. You certainly wouldn't incorporate a PDP-8 into a new design, so if you happen to have a PDP-8 lying around that you want to put to use, the best use would be recreational or educational, save for the needs of a very few "deep-legacy" users. I happen to think PDP-8s are wonderful things, and quite useful for the right purposes, though not principally the ones for which they were used at their time of manufacture. That's why I put "useless" in scare quotes. They are quite useful to me. :) My aesthetic favors discreteness and transparency, thus the "four virtues" of classic computing as my crude rule-of-thumb for assessing how "interesting" an old machine might be: 1) discrete logic (transistors best, SSI better, MSI ok, LSI meh) 2) core memory (or delay lines, drum, something exotic) 3) hardwired control 4) hardwired front panel (direct connection to registers, preferably with all registers displayed) It's just a rule of thumb. Machines in which the ISA is strongly influenced by implementation concerns (e.g., "microcoded" instructions in early DEC architectures) or in which the CPU architecture is strongly influenced by the memory technology (serial drum architectures) are interesting simply because the represent a different way of doing things, in a time when very basic things in computing were hard. Today, the basics are standardized and thus easy, and designers pour the sweat into further layers of refinement instead. About hardwired control: Microcoding is a brilliant idea, and has an elegance of its own. But it also decoupled the ISA from the implementation in a way that allowed for standardization of higher-level logic blocks (e.g., 2900 bit-slice) and less individual character in the hardware. Thus, there is a tendency, in my view, for the hardwired machines to be more interesting, at least with respect to the CPU. The 8/L wins because it is so very small, and yet scores highly against all the criteria above. It also helps that such a CPU minimally configured with an ASR-33 is a very satisfying and compact yet period-accurate computing experience, well-supported by the available software. While this is true of many of its contemporaries, later machines tended to require larger configurations in order to deliver the most authentic period experience. Indeed. A serious contender, and I agree it is a splendidly elegant device. I'd also propose the LGP-21 and even the LGP-30 as having an unusually high "classicness" to size/weight ratio. --Bill

On Fri, Jun 21, 2013 at 04:03:57PM +0200, Philipp Hachtmann wrote: Actually you can get up to 12k with the BM08 box. I have one but haven't found any documentation for it yet. The BM08 has one more set ot switches and a light on the front for the extra IF/DF bit you need.

On 06/21/2013 07:19 PM, David Gesswein wrote: According to Vince, you can go all the way up to 32kW with the BM8L: http://so-much-stuff.com/pdp8/8L/8L.php I don't know who is right :) /P

On 22.06.2013 15:49, Vincent Slyngstad wrote: Sounds reasonable...! BTW Vince, did you get my email? --

On Sat, Jun 22, 2013 at 9:49 AM, Vincent Slyngstad <vrs at msn.com> wrote: I know of the BA08 and have looked at whatever docs are out there for it. I have one and I remember reading that comment about inhibiting CDF for fields >2. I have only seen your pictures of this - it sounds very interesting but I agree they can't have made many of them. -ethan

On 06/22/2013 11:38 PM, Vincent Slyngstad wrote: Not that it is important, but I also have a BM8L . I send you paper copies of the documentation, but that was already some years before. ( ~2004) I could use a pic of the innards, for when I actually get around to doing something with it ! Further rare PDP8 items in my home : a fully maxed up 8A. (128K memory, RL01 drive, and FPP-8A ) Furthermore some omnibus boards for hardware acceleration of an OCR application. Jos Dreesen

On 23.06.2013 17:05, Vincent Slyngstad wrote: Oh that's a good point. I want to extend your statement to most of the vintage computers I came in touch with. I think that it's important to keep and get running strange peripheral stuff that was used with the old computers. I have several unknown Omnibus cards which did something I don't know. They seem to be close to some military stuff. Or "better" scientific stuff. I also have a Camac controller for the Unibus. It's made by "DEC special systems" and has about four or five rows of wirewrap sockets (dual flip chip modules). I have no software for it. But I have a set of HUGE drawings. But no idea how to scan them. With my Honeywell H316 I got an RTP rack and an RTP bus debugging front panel. But no interface to the H316. This stuff could be attached to any computer to control nuclear powerplants and other sensitive machinery (rtp still makes the stuff and has a website). One problem with the second source stuff is that it's more difficult to get documentation for it. So archiving documentation for very rare or probably non-existent hardware should be given some priority over the stuff that everybody has (speaking of redundant piles of pdp11 and pdp8 documentation). Kind regards Philipp :-) --

Perhaps someone recognizes either of the two following Omnibus cards from Tennecomp Systems? Have any documentation perhaps? - TP 5/11 Display Control: two DAC & 7x5 character generator - TP 33 ADC & Control Interface: Rotochem ADC & control interface; dual Berg Thank you for any leads; all that I've gleaned thus far is what I've listed above. On Sun, Jun 23, 2013 at 12:43 PM, P Gebhardt <p.gebhardt at ymail.com> wrote:

On 24 June 2013 15:42, Christian Liendo <christian_liendo at yahoo.com> wrote: The article skipped from 1964 directly to 1976.. it's kind of like skipping from 1964 to 1976 in the space program. -Tor

On 24 June 2013 18:15, Josh Dersch <derschjo at gmail.com> wrote: Another crappy "article" from IO9. I guess Data General and DEC simply didn't exist. Nor Commodore or Radio Shack or ... or ... -- Yoyodyne Propulsion Systems: "The Future Begins Tomorrow" Visit us at: http://www.yoyodyne-propulsion.net -------- "Gregor Samsa awoke one morning to discover that he had been transformed into a giant cockroach." Nah, it's too good. --Max Bialystock

On 6/24/2013 12:57 PM, Josh Dersch wrote: The 8/L arrived on Friday and I've finally had some time to take a few pictures; I have a brief synopsis here: http://rottedbits.blogspot.com/2013/07/new-plaything-pdp-8l.html But the gist is:Aside from the rebranding, itappears to be a completely standard (if minimal) 8/L setup -- 4KW of core, teletype interface and that's it.It looks to be complete and in very good condition, though the rack is ugly as sin(and weighs a ton). There's nothing else in the rack aside from the 8/Litself. The teletype it came with is going to need a lot of cleaning, but is solid and complete. Also -- Thanks to all for the maskingtape removal recommendations; I'm going to start with something minimal like WD-40 and move on to more powerful options if need be. Thanks! Josh

On Sunday, 14 July 2013, Josh Dersch <derschjo at gmail.com> wrote: standard (if minimal) 8/L setup -- 4KW of core, teletype interface and that's it.It looks to be complete and in very good condition, though the rack is ugly as sin(and weighs a ton). There's nothing else in the rack aside from the 8/Litself. The teletype it came with is going to need a lot of cleaning, but is solid and complete. What kind of Teletype did it come with? Being a minimal 8/L it has to be an ASR model. But did it come with a Model 33 (rinky dink plastic one everyone knows about), or a Model 35 (metal beast of Model 28 lineage, which is built like a medium tank... While weighing as much)? going to start with something minimal like WD-40 and move on to more powerful options if need be. Have you ever restored a Teletype? If not, one piece of advice whixh will save you many, many tears: Keep that can of WD-40 far, far away from the Teletype. The stuff will gunk up in quite a "lovely" manner, making your life several new forms of hell. Cheers, Christian G. P.S. Excuse any and all typos, as I'm writing this on my PlayBook, and... well to say I yearn for my Model M is an understatement.

On 14 July 2013 23:02, Josh Dersch <derschjo at gmail.com> wrote: That's too bad, a 35ASR would look far nicer with that rack and that 8/L. Then again, moving a 35ASR is an entirely different story. BitSavers has the relevant technical bulletins (bulletin 273B). And Dave Gesswein has the DEC LT33 modification documents on his site. Any further Teletype help can be acquired on the Greenkeys mailing list. With a vision impairment, a real keyboard of any form is significantly better then a touchscreen. Also, touch typing on a touchscreen isn't excatly possible. Cheers, Christian G.

you keyboard. Then manuals, I time days. Better still, join the Greenkeys mailing list. I did, and it was well worth it. I could never have got my 33 working again otherwise, despite having all the manuals. One thing to be really careful of, don't ever take the cover off the keyboard unless the keyboard is still installed in the sub-base. I would also suggest that a complete strip down would be really challenging and only the really confident, knowledgeable and patient should attempt such a thing. Even the experts tell me some bits are quite tricky and not to dismantle certain parts. Regards Rob

Because I was told parts would fly everywhere. I got this advice from multiple people. Perhaps it is an example of "received wisdom" that has no real basis, who knows, but I bowed to the greater knowledge. early, do to work and I Have to say I don't remember. But I also know that, for me at least, with all those parts, remembering how to put them back together would be a challenge, no matter how good my notes were.

I've checked the approprraite section of the manual. Yo do dismantle the keyboard with ti removed fro mthe base pan. YEs, yuo ahve to take a bit of care removign the RH side as it's spring loaded and parts may driop off (not fly) if you don't hold it in place when removing the cover. But the idea the keyboard will fly apart is total nonsense Hmm... I think it's impossible to remvoe soem parts of the keybaord with it still fixed to the base pan, actually... These days you are lucky. The parts catalogue is available, and that has expoded diagrams of everything. I, on the other hand didn't have this but did have a mis-spent childhood of taking eveything apart. So the only notes I made were the order of apparaently similar parts (like the selector rails) .Everything else I trusted to memory. And had no problems. -tony

As ever, I read the physical paper manual... However, the Model 33 manuals are available on the web, try here, for example : http://www.hpmuseum.net/exhibit.php?hwdoc=540 Download Volume 1 of the Teletype technical manual, the section on dismantling the keyboard starts on page 111 of that .pdf. -tony