We call the PDP-8 a straight 8 because it doesn’t have any letters after the 8 like later variants like the 8S which was a serial variant of the 8 which originally was powerful.
I asked google gemini about subsequent PDP-8s, and got this list: PDP-8: 1965 original model PDP-8/S: 1966, 1-bit serial ALU which was slower but cheaper PDP-8/I: 1969 faster parallel ALU using TTL, making it cheaper than the straight-8 but faster than both previous models PDP-8/E: 1971 "A widely used variant known for its compatibility with a large range of input/output (I/O) devices." PDP-8/M: 1974: faster, more memory CMOS-8: 1979 "The last commercial PDP-8 model, based on CMOS microprocessors. These were not very successful commercially."
I was in the room that day and you didn't get to see that he was also wearing suspenders from the Wang Computer company under that. And sadly no pipe usually, but he always is dressed sharp.
Since I am not able to travel in my current season of life, I really appreciate you giving us virtual tours of amazing places like this one, that have preserved the old computer technology. Keep up the great work!
I personally love your enthusiasm! It's ridiculous for people to say you get too excited. Just the same, I love seeing vintage machines being kept alive! These are a unique opportunity to directly interact with the past, so I'm always happy to see it used!
The PDP-8e was the first computer I worked on in 1973. We started with core memory two 4k stacks of three boards each. These were the control unit for a computer output microfilm unit (COM). They controlled all of the devices that were in the system, a 1/2inch computer tape drive a camera, and a high precision display unit. I ended up with about 10 of these, we were still using them in the 1990's. I bought used ones for the parts, at one time they were almost free. They were great systems for device control due to the simple IOT instructions.
I've been lucky enough to go to the computer museum while they had the naval firing computer (forget model) running. The person running the paper tape had actually used the computer when it was installed, which was also cool. They also have a CRAY and a ton of old home computers. If anyone is going to VCF East, I'd highly recommend going to all of the museums at Infoage. The radio museum is great and the WWII truck museum they also have is really cool. IIRC they have an old "nuclear jeep" there. The trains and military museums/displays are really nice as well. Hoping to get back to VCF east next year. Great show!
The "Zero" bit in a PDP-8 instruction is if the remaining address bits reference the current memory page or the "zero page" in the current field. A field is 4096 words, broken up into pages of 200(8) words. Zero page is literally the addresses from 000-177. So without resorting to indirect addressing, any memory reference instruction can only get to 400(8) spots in memory, but half of them are depending on what address the instruction is at, and the other half are a common pool of locations.
Excellent seeing a PDP 8 in action. I spent hours in school programming the PDP 8E in PAL 8 assembly to do some simple math functions. Fortunately I was able to convince security that I wasn’t a threat as they normally booted you out of the lab at midnight. Unfortunately I did see the sun come up as I got my program to work properly. The boot loader to get the tape drive to spin up was about 20 steps with the front panel switches. I got to the point where I could boot the computer via the front panel from memory every time. Thanks for the memories!
If you have Memory Protect on when running other programs, the boot loader in the last page didn't get overwritten. Of course you have to turn off Memory Protect when using the loader, because it has to have write-access to the last page when running.
Haha I always thought it was so cool that they gave Scotty his own shuttle just to roam and explore the universe at the end of that episode. I'm still jealous of that -- I want my own shuttlecraft!
The site also has a radio/radar/wireless museum too since it was used during WW2 for those purposes. A really fascinating intersection of history and technology to explore.
Thanks so much for the fantastic PDP-8 demo! It sure brought back memories. The PDP-8e was absolutely the first computer I ever laid eyes (or hands) on. DEC was trying to sell one to our high school. They loaned one to the school for a couple of months (!) and we nerds - aka computer club members - were totally enthralled. One word about octal: groups of three switches make a very natural way to load a number into the front panel. You quickly develop a muscle memory for the digits 0 to 7. Since you have to key in the bootstrap loader every time you power the machine up, you get lots of practice! Another DEC product of the same period was the Computer Lab. It's a patch board system with a handful of J-K flipflops and logic gates. If you ever get a chance to acquire one and restore it you should. DEC loaned us one of those, too, and our computer club sponsor actually let me take it home over the Christmas holidays. I learned a ton about digital logic by playing with that thing.
I learned to program on machines that booted from paper tape after toggling in the boot loader , it's great that you captured this process on video so that everyone can see the wonder of this process. It was such an amazing thing to watch a program load from tape especially if it was something you wrote, wonder "is it going to work".
20:00 Watching you key in values with the switches and lights on the PDP8,... magical. It feels like I'm watching curiousMarc and Carl and the Computer History Museum working on the IBM1140! 👍
Oh wow! The Cosmac Elf single board computer! I personally just restored my original Cosmac Super Elf I bought 1977 to full working order. It was a fun job to do, although the original FND500 7-segment numeric LED displays are becoming a bit dim. I have searched for replacement (preferably more modern version), but so far not much luck (the newer ones need a different voltage and forward current). Anyway - It was fun to see all those different versions in this video. In the mean time I am restoring the second computer I bought many years ago as a kit (around 1979). This is the Acon Atom. This one has proven a more difficult little project. I got it halfway working so far. It proudly shows the "Acorn Atom" and prompt at start up, but does not respond to keystrokes. This is going to be a fun troubleshooting project in progression. It will be fun to see that one finally also fully working alongside my Cosmac Super Elf. I am not there yet, but it's going to happen...
Really nice video! I have worked on an SGI Indigo 2 - and it was pretty dated at the time (2001-2003). I still really like IRIX. An Octane or Onyx would be a nice addition to your collection. Fran Blanche just released a video at the same time as yours showing the Univac in operation and the volunteers pulling the cards out and loading a tape up. It was really noisy with everything running!
Thanks for showing the Data General Eclipse S/140. I had about a dozen of those painted in Unisys colors, don't ask... Also, Novas, SuperNovas, Nova II, MicroNova, S140, S250, C330, Lockheed LEC-16, Honeywell DDP-516, HP 1000, PDP 11/44, VAX 11/750, Unisys V77-800. Then came all of the microcontrollers...
Great tour! So good to see these all up and running. In high school we had a PDP-8 with a VT-52, two 8" floppy drives, a line printer, and a CR-11 punched card reader. The card reader could read pencil marks on the cards so that we could write our programs on paper, mark them onto the cards, then run them as a batch job. If you could find and show us a punched card reader, that would be outstanding. We also had eight to ten Commodor PET with that awful keyboard.
Thanks for the great recap of VCF East and the museum. I wanted to go, but was already taking time off earlier in the week for the eclipse and in a few weeks for a visit with my parents.
Come for interesting historic computers Stay for the vibes of the most excitable vintage computer lover Nice reporting and awesome museum! They totally should advertise in the VCF with photos of their interrior and by signs for the people who leave the VCF to also visit them on the way out. This way everyone gets the best of both worlds!
Lots of memories here. I noticed the Digicomo on the shelf, a “computer” made up of plastic, wire, and short plastic soda straws, which was really a mechanical simulation of a few flip-flops. The PDP-8, and its older almost twin brother the PDP-5 were my second computers after an IBM 1620 at UConn in the early 1970s. One of my buddies got a summer job at a company which was entering ALL of their programs with the front panel. He became a real superstar when he showed them the RIM loader.
I checked out the radio museum last year, was quite nice. Will have to check out the computer museum next it looks quite impressive. I saw some on youtube saying wang sold those computers as fully functional computers when in fact there were actually just the adding machine in a computer case.
That a mega machine was ahead of the time . I was worried there be no rabbit to see cause your out of town .. boy I was wrong .. A cotton you can pet ! Wow ! I raised a cotton tail but you could not pet it. She was litter trained and lived with us for year . I finally released her to the wild r and she did well !
That PDP8 was absolutely amazing to see. Loved the switches and the lights showing what was going on. Imo something like this would really help kids (and me) understand what the computer actually does.
Maynooth University's first computer was a Digital Equipment Corporation PDP-11 with a card reader, 2 x demountable hard disk drives, line printer, console and 2 x VT-52 terminals running MU-BASIC under RSTS/E
I am sorry you didn’t stop at our 8-Bit Desk Clones From South America exhibit. There were some wild clones like an Apple II in an Atari 1200 case. Maybe next year!
Really, really amazing! Great tour of the VCF 2024 exhibits and super tour of the vintage computer Museum run by Doug Crawford. His museum is like a "Disneyland of Vintage Computing." Lots of fun to watch and quite educational too. Great job at the restorations and an impressive collection of important machines. (How do you keep them all so clean?) Excellent tour and narration! Thanks for sharing this! ~ VK
I had an account on the PDP-8/I in the computer science department at the local university for a couple of years while I was in high school. It was behind a glass wall, so I could see it blink its lights and spin its DECTape searching for files, but I couldn't actually touch the front panel. It ran TSS/8, a time-share operating system that could support up to 16 users on model 33 Teletypes. One of the TTYs in the computer lab had a paper tape reader and punch. I would go there right after school, and stay until "late" -- after midnight sometimes. I taught myself BASIC and PDP-8 assembly language from photocopied hand-outs from vol. 2 of the manual, and the instruction set quick reference card. So of course at one point I wrote an assembly program to print out my name in block letters on the paper tape. The paper tape is 1 inch wide; if you don't have any of the oiled green or yellow paper tape, you can use 1 inch brown paper tape (the gummed type that you moisten and use to wrap parcels up with brown kraft paper). It had a speech synthesizer attached too, so it could sing a couple of songs and recite poetry. That was a research interest of one of the CS profs.
I remember programming the PDP-8 machine code was a pain in the butt because you can only conditional branch within the current memory page and to escape the page you have to jump. This was fixed in the PDP-11 with its plethora of memory addressing modes.
The PDP-8 doesn't have conditional branches, only conditonal skips - you branch by skipping over a jump. But it's true that you have to use indirect addressing to jump outside the current page. The 11 actually has a similar problem - the conditional branches use relative addressing with a limited range. But at least it doesn't chop the memory up into pages, so it's less annoying. Maybe you're thinking of the VAX? It does have long conditional branches if I remember correctly.
@@gcewing Fun Fact: The low numbered PIC microcontrollers also have the conditional as a skip and a jump instruction. It really is a very good design. Conditional branching takes a longer instruction and in some cases skipping one instruction was the needed action.
It has been a long time since my PAL programming days, but it was even more fun to get to another 4K of memory. We ended up having each of our PDP 8 with a full 32 m of memory. I can remember spending days figuring out how to patch the code to gain a few spaces to change the logic. Out assembly program ended up to be so large we had to compile it on another system, I think a PDP11 and all we had was the source listing.
Man, that Philbrick analog computer is a thing of beauty. If you really want to make an indelible mark on youtube vintage computing, do something with analog computers... anything really: there's a real lack of such content despite there being a resurgence of interest in them using modern techniques/components.
Thanks for the coverage on Core64 David. I'd love to talk about getting cores and tubes connected. After all, those tubes are what necessitated the blazing speeds core memory enabled...
Toggling instructions into memory evokes Spock's characterization - "stone knives and bear skins". The ancesters dealt with a primitive computing world for sure.
Epic video, man. More Bendix, including a cool souvenir, and the PDP-8 were highlights for me. That's a very interesting computer, and one I wish I could afford to house and feed. The Amiga was something else I always wanted but couldn't seem to afford. If you want another possible source for projects, look into the B2600 Workstation from HP. I supported a fleet of about 20 of these at my old job, and devised an imaging server using an old Dell Poweredge 2600 (yeah, the model number is a coincidence). It made possible the saving of whole disk images from these workstations. We were still using these workstations when I retired to operate production processes. I hope I can make a VCF if one comes to a location close enough. This is just too cool.
The "7777" trick is something that a LOT of older computers and microcontrollers used. Store the address of the routine you want the computer to start with at the highest location in memory (the boot/reset vector) and you can store the actual code anywhere you want. Microcontrollers usually execute the instruction at the top of program memory by default after a reset.
The last part of your comment is incorrect. The 8080 and a lot of others started at location zero after reset. The 6502 started at FFFC. 8086 started at FFFF0 IIRC. FFFE was what the 6800 did. CDP1802 started at zero also.
I enjoyed reading 'The Soul of a New Machine' by Tracy Kidder years ago, which is the story behind the development of the Data General Eclipse, so neat to see the museum has one.
The Eclipse S/140 is a 16bit machine. "The Soul of a New Machine" chronicles the development of DG's 32bit MV series machines. One of my coworkers left to go to DG's North Carolina facility and worked on development of the losing design. Amazingly, there are many of these original systems still in production(Rolm produced a mil-spec version of several DG machines). Others are running original DG system software and proprietary applications on modern hardware...See Wild Hare Systems. I have the Nova emulator running on my Dell laptop. It runs 40x faster than the original 800KHz CPU design.
There was a "tag the wumpus" game also. It seems the wumpas was being hunted to near extinction and were being protected and tagged for study. The game mechanics were the same.
@@paulstubbs7678 People need to understand that "BASIC" is not "basic". It is about the most advance language ever invented. Computer languages are a way to express the meaning of some software. More wumpuses are expressed in BASIC than any other language. By that measure, it is the most powerful of languages.
Your plans for the UE-2 are pretty hardcore, haha 9:00 well, it looks familiar, doesn't it? They have a bigger WANG than yours! Hellorld on a paper tape was so much fun, while I'm trying to get C64 BASIC to display big one on the screen using pokes and iteration. Still no cigar but I'm getting there... and will make a video when I get it. opefully it happens RealSoon.
If that PDP-8 program is not "hellord," i will be seriously disappointed. And i was not disappointed. ^-^ I really need to make one of these VCF events. I want to buy stuff, and i don't have the time to troll ebay or local thrift shops.
The Programmed Data Processor (PDP-1) is a high speed, solid state digital computer designed to operate with many types of input-output devices with no internal machine changes.
I would LOVE to see a demo of this analog computer at 10:29 That CRT looks like it has some long persistence phosphor on it with an orange filter over the front to help with contrast. It would probably work on it's own as a vector scope. If you remove the orange filter it will probably look like a P7 cascade phosphor they used on a lot of radar tubes.
I got to finish my own Nabu modem project some time... x.x it's pretty far along after all; but there was next to no concrete feedback on my last updates.
Id love to see what you can do with the NABU. I don't have a ton of time after my job so mine is sitting neatly in it's box still. It is a really cool piece of Canadian PC history.
25:40 We had a ‘visual header’ program on our paper tape based Elliott 903 at university in London in late 1970s. There was an epic piece of tape text on the computer room door that took me three years before I got the joke (I thought it was just profound before that): NOTHING SUCCEEDS LIKE A TOOTHLESS BUDGIE!
TBH I thought this year's VCF East was a little timid compared to last year's. I parked in the main lot(!) and there were a bunch of empty tables where there had been exhibitors on Saturday. I did get to see you working on the Hawk, which was cool.
That Elf exhibit looks awesome! Nice Cosmac VIP (the blue plastic case one). Wish I had one but they are $$$ these days. The VIP was one of 2 derivatives of the Elf that RCA themselves sold as complete machines, albeit half-heartedly as they were terribly mismanaged at the time. The other was a consolized version of the Elf, the much maligned and misunderstood RCA Studio II video game console of AVGN fame. You can port games from the elf family machines to the Studio II without too much trouble as they are mostly the same in terms of hardware but the Studio II is set up to run games off of ROM cartridges while the VIP and elf are meant to run stuff off RAM loaded from cassette or manually entered with the keypad. Most of the games for both platforms were written in Chip8 although there are some newer Studio II homebrews that I think are written in assembly as they run faster. The idea of having a game console and home computer line that could share game software, and therefore capture more market share and reduce dev costs, is kind of mind blowing forward thinking for 1976 when the VIP and Studio II were developed.
The PDP8 only had 8 instructions if you counted them the way DEC counted them. One instruction was effectively a NOP but external logic was used to make that the I/O instruction. The "operate" instruction did things like shifting and clearing so you could make a claim of more than 8 instructions if you counted all the combinations you could make.
Still, there were legitimately only 6 memory reference instructions and 4 addressing modes. It's impressive what they were able to do with that! Also, aren't the I/O instructions of any computer effectively NOPs if there is no external hardware attached?
@@gcewing On a PDP8, the external hardware had to decode that it was the 6000 (octal) instruction (IIRC) as well as decide what to do with the bits from the accumulator and/or whether to strobe new data into the accumulator. The CPU had no hardware related to the I/O operation. Micros generally do some sort of external access for an I/O instruction. Yes, it was amazing what people could do with the PDP8. Until the IBM PC era, it was the most common computer in the world. Things like computerized phone exchanges and numerically controlled milling machines used them. In there simplest form, they could only have 4096 12 bit words of memory. Later versions allowed for 8 "fields" of 4096 words. This still gave them less memory than an early PC.
What about the train rooms!? They are (or used to be?) in the same building as the computer museum, just down the hall. Playing with model train at MIT is essentially how the internet was born, dealing with half vs full duplex and switches.
The RCA 1802 CPU of the COSMAC ELF was a very strange CPU. Ultra low power consumption, but you paid for it in the unusual instructions available. No simple call a subroutine instruction that pushes the return address on a stack, the return address goes to a register, the called routine has the job of preserving the return value somewhere or on a stack, then restoring the PC for a return. The Voyager spacecrafts each have three 1802's that have been working mostly perfectly for decades now which is quite an accomplishment. The rather large transistors help making them quite radiation-resistant. Back in the 1980's they wrote some fresh code to implement better image compression. That must have been quite a challenge. I think they can still upload code to Voyager so there still is a need for 1802 programmers.
Right, no CALL/RETURN, but you can select different registers to be the program counter, then if you return to the previous register, it's like a "RETURN". This lets you build your own custom instructions and is well documented in the manual as SCRT (Standard Call/Return Technique). The 1802 is really a RISC processor.
s/unusual/gawd awful/ I did a significant project with one of those back in the day. You saved less energy than you would expect because it took so many instructions to do anything that the power had to stay on a long time. If you had a more able processor and could power off when idle the battery life would come out about the same. The 1802 didn't eat much power but the PROMs did.
@@kensmith5694 The 8080 took 100mA, the 1802 took 100uA (1,000 times less). If you used that with the low power CMOS RAM and ROM that RCA used, you would save 1,000 times more power... perhaps the speed was 1/2? ok, so you save just 500 times the power, definitely worth the investment in good engineering. Assuming you could go idle... some systems cannot go idle.
@@COSMACELF1802 I had to use PROMs because I didn't have the quantities to go with ROM. The drew more than the 1802. The CMOS RAMS took about 100uA when active and about 10uA when idle. There was a lot more digital stuff. Add to this about 5mA because that's what an LM7805 takes and you will see on time mattered. The 1802 takes 12 clocks to do a NOP. To add two numbers on a 1802 takes about 17 instructions as in Load ACC with high of address for Var1 Put High Into R1 Load ACC with low address for Var1 Put Low into R1 Load ACC with high of address for Var2 Put High Into R2 Load ACC with low address for Var2 Put Low into R2 Load ACC with high address for Result Put High Into R3 Load ACC with low address for Result Put Low Into R3 Load the number pointed to by R1 SEX R2 set the register used in the ADD ADD Store using R3 ******** On a Z80 Load the HL with the address of Var1 Load A with what HL points to Load the HL with the address of Var2 Add with memory ie: what HL points to Load the HL with the address of Result Store This was not the most efficient Z80 code but it is about 1/3rd the instructions
The guy with the pipe can be trusted. A good programmer should always be a pipe smoker. Have you notice that since people stopped smoking, computer sucks?
Is it really possible to make ferrite core memory now? You need very tiny ~1mm cores with especial hysteresis. There are needed vacuum tubes amplifiers for every row and column, to to make it effective you will need ~thousand of cores.
Well, by 1971 when I started my first job (using DEC LINC-8 and PDP12 computers) we still used core memory, but the associated electronics were solid state. Each 1k of storage cost about the same as my annual salary.
Oh cool, a framed cover of the 1974 first edition of Ted Nelson's "Computer Lib" on the wall behind the two of you in that last shot. A very formative book for people of a certain age. "You Can And Must Understand Computers NOW!". Nelson was very savvy at the time about the hopes of the AI transcendantalists. See p. DM 12 (of the "Dream Machines" upside-down second half of the book): "Time out for THREE COMPUTER DREAMS: "AI" (Artificial Intelligence); "IR" (Information Retrieval); "CAI" (Computer-Assisted Instruction). It's time for awe to be replaced with the critical eye." "The God-Builders! ARTIFICIAL INTELLIGENCE ...sort of" "COMPUTERS DON'T ACTUALLY THINK. You just think they think. (We think.)" "ARTIFICIAL INTELLIGENCE: the unrolling carpet. (But how far will it go?)" "Well-understood techniques, 'solved' problems (Impressive, startling and/or hokey demonstrations; Tic-Tac-Toe, Checkers, Qubic; Sentence generation) -> Current Research (Pattern Recognition, Neuron Nets, Heuristics, Simulation of minds & stuff) -> a long, long way -> Mysterious & Sexy Terra Incognita, Unknown & Unlikely ('Turing Test'; Chess: Morphy, Fischer; Self-organizing systems) -> Turing-Barrier (Can it somehow get past? Brute-force methods foiled.) -> El Beyondo. Gods! Supergods! Omniscience! Plus ca change. Nelson is now a very old man. I wonder what he makes of all the current AI ai ai ai.
BARF! I was forced to design a product around that thing and write the code to make it go. My fingers were a full inch shorter by time I was done. BTW: I did a full 32 bit math library and also did graphics on the thing.
Vintage computer enthusiasts are the only people who can say "do you want to come back to my place and see my Wang" and not be creeps. ^-^ And another vintage computer geek would be "Hell yeah! Show me that Wang!" ^-^ I am laughing SO hard right now...
We call the PDP-8 a straight 8 because it doesn’t have any letters after the 8 like later variants like the 8S which was a serial variant of the 8 which originally was powerful.
I asked google gemini about subsequent PDP-8s, and got this list:
PDP-8: 1965 original model
PDP-8/S: 1966, 1-bit serial ALU which was slower but cheaper
PDP-8/I: 1969 faster parallel ALU using TTL, making it cheaper than the straight-8 but faster than both previous models
PDP-8/E: 1971 "A widely used variant known for its compatibility with a large range of input/output (I/O) devices."
PDP-8/M: 1974: faster, more memory
CMOS-8: 1979 "The last commercial PDP-8 model, based on CMOS microprocessors. These were not very successful commercially."
“Which originally was powerful” what does that even mean?
@@pedrocx486 the straight 8 was a powerful machine but the 8S wasn’t
It could also be intended as a pun, because its big rectangular shape kind of resembles an engine block.
I sincerely hope that Thomas walks around those vintage computers with his pipe all the time. Adorbs.
He looked straight out of a movie lol
I felt like he was going to explain the amazing new advances being created in the ARPA laboratories
Not sure how often he has the pipe, but I have NEVER seen Thomas without a business suit on!
I was in the room that day and you didn't get to see that he was also wearing suspenders from the Wang Computer company under that. And sadly no pipe usually, but he always is dressed sharp.
I just got to where this guy shows up, and I love it. 😂 what a pro
Since I am not able to travel in my current season of life, I really appreciate you giving us virtual tours of amazing places like this one, that have preserved the old computer technology. Keep up the great work!
I personally love your enthusiasm! It's ridiculous for people to say you get too excited.
Just the same, I love seeing vintage machines being kept alive! These are a unique opportunity to directly interact with the past, so I'm always happy to see it used!
The PDP-8e was the first computer I worked on in 1973. We started with core memory two 4k stacks of three boards each. These were the control unit for a computer output microfilm unit (COM). They controlled all of the devices that were in the system, a 1/2inch computer tape drive a camera, and a high precision display unit. I ended up with about 10 of these, we were still using them in the 1990's. I bought used ones for the parts, at one time they were almost free. They were great systems for device control due to the simple IOT instructions.
I've been lucky enough to go to the computer museum while they had the naval firing computer (forget model) running. The person running the paper tape had actually used the computer when it was installed, which was also cool. They also have a CRAY and a ton of old home computers. If anyone is going to VCF East, I'd highly recommend going to all of the museums at Infoage. The radio museum is great and the WWII truck museum they also have is really cool. IIRC they have an old "nuclear jeep" there. The trains and military museums/displays are really nice as well. Hoping to get back to VCF east next year. Great show!
HELLORLD! on paper tape is awesome!
Yeah, now he *has* to get the G-15 to do that with its paper tape punch. (Even if it's just hard coded.)
The "Zero" bit in a PDP-8 instruction is if the remaining address bits reference the current memory page or the "zero page" in the current field. A field is 4096 words, broken up into pages of 200(8) words. Zero page is literally the addresses from 000-177. So without resorting to indirect addressing, any memory reference instruction can only get to 400(8) spots in memory, but half of them are depending on what address the instruction is at, and the other half are a common pool of locations.
Excellent seeing a PDP 8 in action. I spent hours in school programming the PDP 8E in PAL 8 assembly to do some simple math functions. Fortunately I was able to convince security that I wasn’t a threat as they normally booted you out of the lab at midnight. Unfortunately I did see the sun come up as I got my program to work properly.
The boot loader to get the tape drive to spin up was about 20 steps with the front panel switches. I got to the point where I could boot the computer via the front panel from memory every time.
Thanks for the memories!
I programmed in PAL-8 and PAL-11. Paper tape! Good Times!
you had a computer in school in the 1960s or early 70s my school didn't get them until the year 2000 you are probably older than my father.
If you have Memory Protect on when running other programs, the boot loader in the last page didn't get overwritten. Of course you have to turn off Memory Protect when using the loader, because it has to have write-access to the last page when running.
@@belstar1128 The year was 1976. I was studying electronics engineering. The PDP8 programming was part of the digital electronics course.
I assumed "Straight 8" was to distinguish from the later 8E, 8A, 8I etc variants.
Cue Scotty: "No bloody A, B, C, or D"
Haha I always thought it was so cool that they gave Scotty his own shuttle just to roam and explore the universe at the end of that episode. I'm still jealous of that -- I want my own shuttlecraft!
The site also has a radio/radar/wireless museum too since it was used during WW2 for those purposes. A really fascinating intersection of history and technology to explore.
i want go there so badly
Thanks so much for the fantastic PDP-8 demo! It sure brought back memories. The PDP-8e was absolutely the first computer I ever laid eyes (or hands) on. DEC was trying to sell one to our high school. They loaned one to the school for a couple of months (!) and we nerds - aka computer club members - were totally enthralled. One word about octal: groups of three switches make a very natural way to load a number into the front panel. You quickly develop a muscle memory for the digits 0 to 7. Since you have to key in the bootstrap loader every time you power the machine up, you get lots of practice!
Another DEC product of the same period was the Computer Lab. It's a patch board system with a handful of J-K flipflops and logic gates. If you ever get a chance to acquire one and restore it you should. DEC loaned us one of those, too, and our computer club sponsor actually let me take it home over the Christmas holidays. I learned a ton about digital logic by playing with that thing.
At one point DEC actually promoted just how "mini" the PDP-8 was by showing it in the back of a classic VW Beetle.
I learned to program on machines that booted from paper tape after toggling in the boot loader , it's great that you captured this process on video so that everyone can see the wonder of this process. It was such an amazing thing to watch a program load from tape especially if it was something you wrote, wonder "is it going to work".
Holy crap, the museum has a Hammond Novachord?!? As a tube guy, you need to look at that instrument. It's nuts.
That caught my attention as well. Glad I'm not alone!
20:00 Watching you key in values with the switches and lights on the PDP8,... magical. It feels like I'm watching curiousMarc and Carl and the Computer History Museum working on the IBM1140! 👍
Thanks so much for the tour. I'm a retired DEC FE and loved the trip down memory lane.
@15:10 The Screen on the Commadore PET is Displaying Ferris Bueller's Student record..........Nice touch form the museum folks.
Oh wow! The Cosmac Elf single board computer! I personally just restored my original Cosmac Super Elf I bought 1977 to full working order. It was a fun job to do, although the original FND500 7-segment numeric LED displays are becoming a bit dim. I have searched for replacement (preferably more modern version), but so far not much luck (the newer ones need a different voltage and forward current).
Anyway - It was fun to see all those different versions in this video.
In the mean time I am restoring the second computer I bought many years ago as a kit (around 1979). This is the Acon Atom. This one has proven a more difficult little project. I got it halfway working so far. It proudly shows the "Acorn Atom" and prompt at start up, but does not respond to keystrokes. This is going to be a fun troubleshooting project in progression. It will be fun to see that one finally also fully working alongside my Cosmac Super Elf. I am not there yet, but it's going to happen...
Nice little cameo you have there of the UE 555 on display in the museum at around 09:00
Really nice video! I have worked on an SGI Indigo 2 - and it was pretty dated at the time (2001-2003). I still really like IRIX. An Octane or Onyx would be a nice addition to your collection.
Fran Blanche just released a video at the same time as yours showing the Univac in operation and the volunteers pulling the cards out and loading a tape up. It was really noisy with everything running!
Thanks for showing the Data General Eclipse S/140. I had about a dozen of those painted in Unisys colors, don't ask... Also, Novas, SuperNovas, Nova II, MicroNova, S140, S250, C330, Lockheed LEC-16, Honeywell DDP-516, HP 1000, PDP 11/44, VAX 11/750, Unisys V77-800. Then came all of the microcontrollers...
Great tour! So good to see these all up and running. In high school we had a PDP-8 with a VT-52, two 8" floppy drives, a line printer, and a CR-11 punched card reader. The card reader could read pencil marks on the cards so that we could write our programs on paper, mark them onto the cards, then run them as a batch job. If you could find and show us a punched card reader, that would be outstanding. We also had eight to ten Commodor PET with that awful keyboard.
Thanks for the great recap of VCF East and the museum. I wanted to go, but was already taking time off earlier in the week for the eclipse and in a few weeks for a visit with my parents.
I love the museum because when it's NOT festival time you can try any of the turned on machines (not just behind glass)
VCF East XIX was awesome. Glad you could make it!
Come for interesting historic computers
Stay for the vibes of the most excitable vintage computer lover
Nice reporting and awesome museum!
They totally should advertise in the VCF with photos of their interrior and by signs for the people who leave the VCF to also visit them on the way out. This way everyone gets the best of both worlds!
That museum is awesome! 😍
I had the first compaq sold in new mexico. I remember carrying it on an airplane. I bought a 10 MB hard drive that I thought would never fill up.
Lots of memories here. I noticed the Digicomo on the shelf, a “computer” made up of plastic, wire, and short plastic soda straws, which was really a mechanical simulation of a few flip-flops. The PDP-8, and its older almost twin brother the PDP-5 were my second computers after an IBM 1620 at UConn in the early 1970s. One of my buddies got a summer job at a company which was entering ALL of their programs with the front panel. He became a real superstar when he showed them the RIM loader.
I checked out the radio museum last year, was quite nice. Will have to check out the computer museum next it looks quite impressive. I saw some on youtube saying wang sold those computers as fully functional computers when in fact there were actually just the adding machine in a computer case.
I'm fascinated by old weather channel tech!
Very interesting, thanks, that diy paper tape reader writer looks very nice
Wow, this was all great, nice highlighting of the museum. & Thomas has it going ON
BUNNY!
My daughters walked in just a the right time LOL
Thomas really leaning in to that retro style, I approve!
That a mega machine was ahead of the time . I was worried there be no rabbit to see cause your out of town .. boy I was wrong .. A cotton you can pet ! Wow ! I raised a cotton tail but you could not pet it. She was litter trained and lived with us for year . I finally released her to the wild r and she did well !
That PDP8 was absolutely amazing to see. Loved the switches and the lights showing what was going on. Imo something like this would really help kids (and me) understand what the computer actually does.
AAhh...this explains why you weren't at the Meet And Greet in Grand Prairie this weekend! 😉 Good to see the stuff at VCF East! Thanks for the tours!
Maynooth University's first computer was a Digital Equipment Corporation PDP-11 with a card reader, 2 x demountable hard disk drives, line printer, console and 2 x VT-52 terminals running MU-BASIC under RSTS/E
Thanks for a wonderful video.
Your exitement is one of the reasons I love your channel - nihl contentae barstidium! 😋
thank you it was educative and awesome .
26:00
- Hello are you scared?
- HELLIAM!
I am sorry you didn’t stop at our 8-Bit Desk Clones From South America exhibit. There were some wild clones like an Apple II in an Atari 1200 case. Maybe next year!
Really, really amazing! Great tour of the VCF 2024 exhibits and super tour of the vintage computer Museum run by Doug Crawford. His museum is like a "Disneyland of Vintage Computing." Lots of fun to watch and quite educational too. Great job at the restorations and an impressive collection of important machines. (How do you keep them all so clean?) Excellent tour and narration! Thanks for sharing this! ~ VK
I had an account on the PDP-8/I in the computer science department at the local university for a couple of years while I was in high school. It was behind a glass wall, so I could see it blink its lights and spin its DECTape searching for files, but I couldn't actually touch the front panel. It ran TSS/8, a time-share operating system that could support up to 16 users on model 33 Teletypes. One of the TTYs in the computer lab had a paper tape reader and punch. I would go there right after school, and stay until "late" -- after midnight sometimes. I taught myself BASIC and PDP-8 assembly language from photocopied hand-outs from vol. 2 of the manual, and the instruction set quick reference card. So of course at one point I wrote an assembly program to print out my name in block letters on the paper tape. The paper tape is 1 inch wide; if you don't have any of the oiled green or yellow paper tape, you can use 1 inch brown paper tape (the gummed type that you moisten and use to wrap parcels up with brown kraft paper).
It had a speech synthesizer attached too, so it could sing a couple of songs and recite poetry. That was a research interest of one of the CS profs.
I remember programming the PDP-8 machine code was a pain in the butt because you can only conditional branch within the current memory page and to escape the page you have to jump. This was fixed in the PDP-11 with its plethora of memory addressing modes.
The PDP-8 doesn't have conditional branches, only conditonal skips - you branch by skipping over a jump. But it's true that you have to use indirect addressing to jump outside the current page.
The 11 actually has a similar problem - the conditional branches use relative addressing with a limited range. But at least it doesn't chop the memory up into pages, so it's less annoying.
Maybe you're thinking of the VAX? It does have long conditional branches if I remember correctly.
@@gcewing Fun Fact: The low numbered PIC microcontrollers also have the conditional as a skip and a jump instruction. It really is a very good design. Conditional branching takes a longer instruction and in some cases skipping one instruction was the needed action.
It has been a long time since my PAL programming days, but it was even more fun to get to another 4K of memory. We ended up having each of our PDP 8 with a full 32 m of memory.
I can remember spending days figuring out how to patch the code to gain a few spaces to change the logic. Out assembly program ended up to be so large we had to compile it on another system, I think a PDP11 and all we had was the source listing.
Man, that Philbrick analog computer is a thing of beauty. If you really want to make an indelible mark on youtube vintage computing, do something with analog computers... anything really: there's a real lack of such content despite there being a resurgence of interest in them using modern techniques/components.
That paper tape reader takes me back to my first job programming Ferranti FM1600B computers
Thanks for the coverage on Core64 David. I'd love to talk about getting cores and tubes connected. After all, those tubes are what necessitated the blazing speeds core memory enabled...
Toggling instructions into memory evokes Spock's characterization - "stone knives and bear skins". The ancesters dealt with a primitive computing world for sure.
Epic video, man. More Bendix, including a cool souvenir, and the PDP-8 were highlights for me. That's a very interesting computer, and one I wish I could afford to house and feed.
The Amiga was something else I always wanted but couldn't seem to afford. If you want another possible source for projects, look into the B2600 Workstation from HP. I supported a fleet of about 20 of these at my old job, and devised an imaging server using an old Dell Poweredge 2600 (yeah, the model number is a coincidence). It made possible the saving of whole disk images from these workstations. We were still using these workstations when I retired to operate production processes.
I hope I can make a VCF if one comes to a location close enough. This is just too cool.
The "7777" trick is something that a LOT of older computers and microcontrollers used. Store the address of the routine you want the computer to start with at the highest location in memory (the boot/reset vector) and you can store the actual code anywhere you want. Microcontrollers usually execute the instruction at the top of program memory by default after a reset.
The last part of your comment is incorrect. The 8080 and a lot of others started at location zero after reset. The 6502 started at FFFC. 8086 started at FFFF0 IIRC. FFFE was what the 6800 did. CDP1802 started at zero also.
That looks fantastic :)
I enjoyed reading 'The Soul of a New Machine' by Tracy Kidder years ago, which is the story behind the development of the Data General Eclipse, so neat to see the museum has one.
The Eclipse S/140 is a 16bit machine. "The Soul of a New Machine" chronicles the development of DG's 32bit MV series machines. One of my coworkers left to go to DG's North Carolina facility and worked on development of the losing design. Amazingly, there are many of these original systems still in production(Rolm produced a mil-spec version of several DG machines). Others are running original DG system software and proprietary applications on modern hardware...See Wild Hare Systems. I have the Nova emulator running on my Dell laptop. It runs 40x faster than the original 800KHz CPU design.
12:28 'Hunt the Wumpus' now there is something I have to get back to and get that running on my SWTPC6800 clone I made 'quite a few years ago'.
There was a "tag the wumpus" game also. It seems the wumpas was being hunted to near extinction and were being protected and tagged for study. The game mechanics were the same.
@@kensmith5694 The one I have is written in basic, so super easy to make variants.
@@paulstubbs7678 People need to understand that "BASIC" is not "basic". It is about the most advance language ever invented. Computer languages are a way to express the meaning of some software. More wumpuses are expressed in BASIC than any other language. By that measure, it is the most powerful of languages.
I wonder if there are any working examples of the Texas Instrument TI-980B mini-computer still working? They were stunning looking machines!
Your plans for the UE-2 are pretty hardcore, haha
9:00 well, it looks familiar, doesn't it?
They have a bigger WANG than yours!
Hellorld on a paper tape was so much fun, while I'm trying to get C64 BASIC to display big one on the screen using pokes and iteration. Still no cigar but I'm getting there... and will make a video when I get it. opefully it happens RealSoon.
And it's a proper Wang with logarithmic logic...
I live locally...but unfortunately wasnt able to make it this year, I am looking forward to attending in 2025
I was there too, great fun
Very much liked the PDP stuff, I'm currently building a PiDP 8/I so really interesting stuff.
If that PDP-8 program is not "hellord," i will be seriously disappointed.
And i was not disappointed. ^-^
I really need to make one of these VCF events. I want to buy stuff, and i don't have the time to troll ebay or local thrift shops.
Thank you for carrying me back to my Youth
i stuck my head into the musesum while i was there but it was too many people in there to even realize you were in there.
I have used a DEC PDP11 at Trostan Avenue Ballymena Northern Ireland.
The Programmed Data Processor (PDP-1) is a high speed, solid state digital computer designed to operate with many types of input-output devices with no internal machine changes.
I would LOVE to see a demo of this analog computer at 10:29
That CRT looks like it has some long persistence phosphor on it with an orange filter over the front to help with contrast. It would probably work on it's own as a vector scope. If you remove the orange filter it will probably look like a P7 cascade phosphor they used on a lot of radar tubes.
I got to finish my own Nabu modem project some time... x.x it's pretty far along after all; but there was next to no concrete feedback on my last updates.
Amazing !...as usual !
Id love to see what you can do with the NABU. I don't have a ton of time after my job so mine is sitting neatly in it's box still. It is a really cool piece of Canadian PC history.
25:40 We had a ‘visual header’ program on our paper tape based Elliott 903 at university in London in late 1970s. There was an epic piece of tape text on the computer room door that took me three years before I got the joke (I thought it was just profound before that): NOTHING SUCCEEDS LIKE A TOOTHLESS BUDGIE!
Would love to see you at vcfmw this year.
Didn't make it this time but next time I hope to show up and see everything
My guinea pig says hi to your cornucopia bunny :)
TBH I thought this year's VCF East was a little timid compared to last year's. I parked in the main lot(!) and there were a bunch of empty tables where there had been exhibitors on Saturday. I did get to see you working on the Hawk, which was cool.
That Elf exhibit looks awesome! Nice Cosmac VIP (the blue plastic case one). Wish I had one but they are $$$ these days. The VIP was one of 2 derivatives of the Elf that RCA themselves sold as complete machines, albeit half-heartedly as they were terribly mismanaged at the time. The other was a consolized version of the Elf, the much maligned and misunderstood RCA Studio II video game console of AVGN fame. You can port games from the elf family machines to the Studio II without too much trouble as they are mostly the same in terms of hardware but the Studio II is set up to run games off of ROM cartridges while the VIP and elf are meant to run stuff off RAM loaded from cassette or manually entered with the keypad. Most of the games for both platforms were written in Chip8 although there are some newer Studio II homebrews that I think are written in assembly as they run faster. The idea of having a game console and home computer line that could share game software, and therefore capture more market share and reduce dev costs, is kind of mind blowing forward thinking for 1976 when the VIP and Studio II were developed.
The ELF II Reproduction Boards are available on the COSMAC ELF forum. You can add personality modules to it to run VIP and STUDIO II Programs.
@@COSMACELF1802 Awesome! I am probably the only person in the world who has voluntarily played Studio II games for over an hour at a time😄
"I'm not an electrical engineer"
I've been an electrical engineer for 10 years, you're more electrical engineer than I am.
We used to upgrade those Compac luggables through 386, 486 and then Pentium. All had custom data acquisition cards.... Those where the days.....
The PDP8 only had 8 instructions if you counted them the way DEC counted them. One instruction was effectively a NOP but external logic was used to make that the I/O instruction. The "operate" instruction did things like shifting and clearing so you could make a claim of more than 8 instructions if you counted all the combinations you could make.
Still, there were legitimately only 6 memory reference instructions and 4 addressing modes. It's impressive what they were able to do with that!
Also, aren't the I/O instructions of any computer effectively NOPs if there is no external hardware attached?
@@gcewing On a PDP8, the external hardware had to decode that it was the 6000 (octal) instruction (IIRC) as well as decide what to do with the bits from the accumulator and/or whether to strobe new data into the accumulator. The CPU had no hardware related to the I/O operation. Micros generally do some sort of external access for an I/O instruction.
Yes, it was amazing what people could do with the PDP8. Until the IBM PC era, it was the most common computer in the world. Things like computerized phone exchanges and numerically controlled milling machines used them. In there simplest form, they could only have 4096 12 bit words of memory. Later versions allowed for 8 "fields" of 4096 words. This still gave them less memory than an early PC.
What about the train rooms!? They are (or used to be?) in the same building as the computer museum, just down the hall. Playing with model train at MIT is essentially how the internet was born, dealing with half vs full duplex and switches.
The TRIAD System Mini computer, Was that related to later triad systems for running hardware stores?
Props++ for the mention & great video! Chat soon :)
My middle school had some of those PETs. I think they would run out of memory playing Space Invaders.
It must be amazing to play with such an old wang that still works.
The RCA 1802 CPU of the COSMAC ELF was a very strange CPU. Ultra low power consumption, but you paid for it in the unusual instructions available. No simple call a subroutine instruction that pushes the return address on a stack, the return address goes to a register, the called routine has the job of preserving the return value somewhere or on a stack, then restoring the PC for a return. The Voyager spacecrafts each have three 1802's that have been working mostly perfectly for decades now which is quite an accomplishment. The rather large transistors help making them quite radiation-resistant. Back in the 1980's they wrote some fresh code to implement better image compression. That must have been quite a challenge. I think they can still upload code to Voyager so there still is a need for 1802 programmers.
Right, no CALL/RETURN, but you can select different registers to be the program counter, then if you return to the previous register, it's like a "RETURN". This lets you build your own custom instructions and is well documented in the manual as SCRT (Standard Call/Return Technique). The 1802 is really a RISC processor.
s/unusual/gawd awful/
I did a significant project with one of those back in the day. You saved less energy than you would expect because it took so many instructions to do anything that the power had to stay on a long time.
If you had a more able processor and could power off when idle the battery life would come out about the same. The 1802 didn't eat much power but the PROMs did.
@@kensmith5694 The 8080 took 100mA, the 1802 took 100uA (1,000 times less). If you used that with the low power CMOS RAM and ROM that RCA used, you would save 1,000 times more power... perhaps the speed was 1/2? ok, so you save just 500 times the power, definitely worth the investment in good engineering. Assuming you could go idle... some systems cannot go idle.
@@COSMACELF1802 I had to use PROMs because I didn't have the quantities to go with ROM. The drew more than the 1802. The CMOS RAMS took about 100uA when active and about 10uA when idle. There was a lot more digital stuff. Add to this about 5mA because that's what an LM7805 takes and you will see on time mattered.
The 1802 takes 12 clocks to do a NOP.
To add two numbers on a 1802 takes about 17 instructions as in
Load ACC with high of address for Var1
Put High Into R1
Load ACC with low address for Var1
Put Low into R1
Load ACC with high of address for Var2
Put High Into R2
Load ACC with low address for Var2
Put Low into R2
Load ACC with high address for Result
Put High Into R3
Load ACC with low address for Result
Put Low Into R3
Load the number pointed to by R1
SEX R2 set the register used in the ADD
ADD
Store using R3
********
On a Z80
Load the HL with the address of Var1
Load A with what HL points to
Load the HL with the address of Var2
Add with memory ie: what HL points to
Load the HL with the address of Result
Store
This was not the most efficient Z80 code but it is about 1/3rd the instructions
The guy with the pipe can be trusted.
A good programmer should always be a pipe smoker.
Have you notice that since people stopped smoking, computer sucks?
"Streight 8" designates it as a base unit, 8S, 8M, 8I, or 8 Lab.
Adrian Black gets very excited too when he gets a computer working again
Is it really possible to make ferrite core memory now? You need very tiny ~1mm cores with especial hysteresis. There are needed vacuum tubes amplifiers for every row and column, to to make it effective you will need ~thousand of cores.
Well, by 1971 when I started my first job (using DEC LINC-8 and PDP12 computers) we still used core memory, but the associated electronics were solid state. Each 1k of storage cost about the same as my annual salary.
How about VCF Midwest, September 7-8, 2024 in Schaumburg, Illinois (suburb of Chicago)?
Those diodes are silicon. There's no germanium in those flip chips. There's a bunch in the older "system modules", though.
Bunnies! 😍
The buttons on the Wang computer, Wang calculator and PET look like they wouldn't be out of place on a certain vintage of cash register...
Have yet to see a Honeywell or Cincinnati Milacron. We have seen some Data Generals recently.
@4:32 I think I built one of those in the past 😉
Oh cool, a framed cover of the 1974 first edition of Ted Nelson's "Computer Lib" on the wall behind the two of you in that last shot. A very formative book for people of a certain age. "You Can And Must Understand Computers NOW!". Nelson was very savvy at the time about the hopes of the AI transcendantalists. See p. DM 12 (of the "Dream Machines" upside-down second half of the book): "Time out for THREE COMPUTER DREAMS: "AI" (Artificial Intelligence); "IR" (Information Retrieval); "CAI" (Computer-Assisted Instruction). It's time for awe to be replaced with the critical eye." "The God-Builders! ARTIFICIAL INTELLIGENCE ...sort of" "COMPUTERS DON'T ACTUALLY THINK. You just think they think. (We think.)" "ARTIFICIAL INTELLIGENCE: the unrolling carpet. (But how far will it go?)" "Well-understood techniques, 'solved' problems (Impressive, startling and/or hokey demonstrations; Tic-Tac-Toe, Checkers, Qubic; Sentence generation) -> Current Research (Pattern Recognition, Neuron Nets, Heuristics, Simulation of minds & stuff) -> a long, long way -> Mysterious & Sexy Terra Incognita, Unknown & Unlikely ('Turing Test'; Chess: Morphy, Fischer; Self-organizing systems) -> Turing-Barrier (Can it somehow get past? Brute-force methods foiled.) -> El Beyondo. Gods! Supergods! Omniscience!
Plus ca change. Nelson is now a very old man. I wonder what he makes of all the current AI ai ai ai.
Hooray for the CDP-1802!!!
BARF! I was forced to design a product around that thing and write the code to make it go. My fingers were a full inch shorter by time I was done.
BTW: I did a full 32 bit math library and also did graphics on the thing.
Vintage computer enthusiasts are the only people who can say "do you want to come back to my place and see my Wang" and not be creeps. ^-^
And another vintage computer geek would be "Hell yeah! Show me that Wang!" ^-^
I am laughing SO hard right now...
The Commodore Amiga is great