[RetroBytes] takes us on a whirlwind tour of the history of the Digital Equipment Corporation (DEC), its founder Ken Olsen, and during intermission builds up a working replica of the PDP-11 from a kit. DEC was a major player in the early computer industry, cranking out a number of models that were both industrial workhorses and used in computer laboratories to develop many of the operating systems and tools whose descendants we still use today. On top of that, DEC’s innovative, employee-friendly, and lightweight company structure was generally well-liked by its employees and a welcomed departure from the typical behemoths of the day.
This video takes us from the beginnings of DEC and its roots in MIT up to the PIP-11 era, highlighting major architectures and events along the way such as the PDP-1, PDP-8, and PDP-11. [RetroBytes] says he has a DEC Alpha sitting on the sidelines, so there may be a few follow-up videos in the future — perhaps one on the VAX as well.
We’ve covered this particular PDP-11 replica last year, and if these replica kits are your cup of tea, check out our coverage of kit designer [Oscar Vermeulen]’s presentation. Have you ever used real PDP or VAX computers? Let us know your war stories in the comments below.
My first experience with a DEC computer was with a DEC-10 (soon followed by a DEC-20). Every memory is positive. These were nice machines with superlative software for the times. Being able to work on a terminal running the SOS editor and using the DDT debugger was a dream come true compared to keypunches and card decks.
I use to also use a dec-10 and dec-20.. remember DECWAR on decsystem-10s?
I don’t remember that, but it’s been more than 40 years so…
Back in the early 80’s I was using a PDP 11/70 to process some data collected from a NASA Venus satellite at GSFC. When my program would start to run it was heavy on the disk i/o and after about 5 or 10 minutes of running it would cause the 11/70 to crash. Which would make the other people using the time-sharing setup really unhappy to say the least. It got so that when the other programmers would see me entering the room they would be careful to only run stuff that they wouldn’t feel bad about if the 11/70 crashed – which it usually would thanks to me. Other than me being there running my code the 11/70 had been really reliable. A couple of DEC technicians came in there for a couple of days and after a lot of effort and head scratching on their part they tracked it down to a loose screw on one of the arms in one of the disk drives. These were the big washing machine size things with the removable disk packs. The techs came to the conclusion that when I started all the disk i/o, the drive would heat up, the metal on the arm would expand and the head would go out of alignment making for bad reads and crashing the 11/70.
Probably an RL02 see https://gunkies.org/wiki/RL01/02_disk_drive
Probab ly not an RL02, as those didn’t need alignment. I bet something like an RP06. (/bqt)
RM05?
I was tech support on the CDC version of the RM05, which actually was an Magnetic Peripherals/Control Data drive. If the head screws were torqued to spec you couldn’t beat them out of alignment.
The PiDP-11 kit (and also the PiDP-8 kit) is just great! You can get a relatively modern 2.11BSD OS running on it in no time and then have your yearly Python-Detox with it: writing code in C or FORTRAN if you feel like it, in a 16 bit address space with overlays etc using at most 4MB in total for the entire system. It is amazing what CAN actually be done with a system like that. The blinkenlights are nice to look at, but actually working with an emulated PDP-11 is the actual fun part.
I have fond, as well as pained, memories of several DEC systems. PDP-8 in the lab (first lab I worked in) way back in the day (also a SOL in the same lab, though that wasn’t really the labs. It was acquired by one of the postdocs via some funding or other and just lived there), running, IIRC, OS/8. Had a medusa-head worth of cables interfacing it to whatever was going on there at the time. There was also one in my high school with an ASR-33, all hand-me-down from a local college.
A variety of 11’s and Vaxen, running a variety of systems, as, well, the least pleasant running VMS under the iron control of an independent IT department (large university- we were beholden to them. I think that they had IBM370-envy), with Unix running on another VAX, but mightily pleasant to work on, despite being admined by the same IT department. I think the senior people in IT didn’t get unix, so left that system to the one or two guys (ALL guys, back then) that did, and they were definitively “on the bus”. The coming of the desktop PC clipped the IT czar wings as there mission shifted to network management. The unix guys ended up pretty much running that end.
In the late 1980’s, I picked up a lot of Unibus cards second have at Eli Heffrons. Some were quite interesting, if cryptic. Fun times
Yeah, the first big machine I ever used was a PDP-11/70 (had already been using micros, KIM-1 and Pet). Fortran running batch under RT-11 and (woo hoo) BASIC under interactive RSTS/E. Back in the late ’70s there was a computer hardware reseller in Ann Arbor, up on N Main, in one of those warehouses pinched between Main and the railroad tracks. They always had some PDP-8s and I kept thinking I should buy one, but never did. Oh, well, it would have an honored place in my rack if I had (right next to my two KIM boards).
I got a job offer from DEC in Vermont in 1979 (with three increasing salary offers, no less), and like an idiot turned them down to work at a very unsatisfying electronics company that I didn’t stay with for long. Even though DEC died before I would have retired, I still very much regret missing out on the (presumed) fun of working for them. It was a PDP-8/I that I learned programming on, in college. This video has inspired me to buy a PiDP-11 kit to use as a front end for my home security system, ensuring that the system will be completely opaque to anyone other than me.
That’s the spirit! I’ve thought about that myself as my PiDP 11/70’s PI4 is basically ‘idle’ as the services in the background are not CPU/Memory intensive. I only use SSH to get into mine, so all people see is switches and lights. To give SSH a bit more ‘old school’ look I like to use cool-retro-term as the terminal software :) . Hook up say a RP2030 PICO board to bring in the IO off of a USB port or use Wifi to interface to external devices… Lots of options there. I extended the back of mine with a 3D printed ‘PiDP-11 case extension’ (from thingiverse) to give mine more room in the back to hide more electronics and cables. Then used freeCad to design a custom laser cut back plate which attached to nice wood base (not what was in the kit) for a nice ‘solid’ feeling case. Worked out well. I can reach over and flip switches without worrying about tipping the front panel over!
That’s RP2040…. Can’t edit post….
I found it very frustrating that the old photos were obscured by overlayed spinning videos and falling money, and that Ken Olsen was talked over. This tells the same story (and moves onto the Vax) but in Ken Oldsen’s own words and with old photos intact: https://www.youtube.com/watch?v=VEf2Xf7Urn8
In 1979 I used a PDP8 with a Focal program to test Langmuir probes for the Dynamic Explorer satellites at the Space Physics Research lab, Ann Arbor. Later that year built some processing cards for 5 pipe lined PDP11 (LSI11) machines used for seismic signal processing.
All the graybeards are coming out. My introduction to DEC equipment was the PDP 11/34 and for which I had to modify IEEE488 (HPIB) driver to make to promiscuous. A perusal of the PDP hardware handbook revealed there was a FUBAR register. The register name is not the well-known acronym for “f*d up beyond all recognition” but rather a “failed unibus address register”. Funny to me at the time.
My favourite bit of DEC trivia is that the basic unit of time on a VAX system was named the micro-fortnight.
It’s a unit used in one specific part of VMS. It’s not the basic unit of time, though, and it’s not generic VAX. But I also like the unit…
Under the hood, most people are still using DEC: the WindowsNT kernel is heavily based on VMS.
I’m not. But the system I’m using is heavily based on Unix.
Heavily influenced by VMS, given Dave Cutler and the group of DEC engineers that went on to develop Windows NT at Microsoft, but not “based on VMS”. There are many parallels between NT and VMS, but the Windows NT kernel is more like a Mach microkernel. It also goes the other way–OpenVMS picked up a lot of features from Windows NT.
And I forgot to add, the file system automatic versioning of files was sometimes an annoyance but well appreciated.
The filesystem really didn’t version files. It didn’t know foo.txt;1 was related to foo.txt;2. It was the Record Management Service (RMS) which was layered over the filesystem that did that stuff. But I guess for most users the effect was the same. Long live sys$qio() !
I’m not really qualified to rank operating systems, and I certainly wasn’t as a young 20s college kid working on a VAX. But I’ve always felt that VMS, or more specifically, it’s CLI, was a step above everything else I’d been exposed to. There was a certain consistency in options across commands, and the help system was almost as good as the real set of manuals on the bookshelf.
I’d agree. I only worked on VMS for a few years, and spent a similar amount of time on CDC Cyber NOS/VE. I spent a bit more time on IBM maiframes (TSO and VM/CMS), and I’ve spent lots more time on various flavors of Unix/Linux, as well as MS-DOS, Windows, and MacOS.
VMS certainly had consistency. The EVE text editor, with its smooth scrolling windows was both intuitive to use and a thing of beauty. I loved the fact that VMS kept multiple old versions of your files around until you explicitly purged them. Many times, that saved me from fat-fingered mistakes.
Unix is nice, but its design was much more open, and so various individuals or small teams developed different programs independently, and it shows. VMS clearly had some central planning bureau enforcing standards.
IBM also had central planning standards for their MVS/TSO system, but their central planning bureau appeared to be on a mission to keep the art of programming as obscure as possible. Why else would they have though IEFBR14 is the right name for a program which is nothing but a return statement? And why would they have accepted as logical the idea that you needed to run IEFBR14 with appropriate DD statements any time you wanted to create or delete a file? (er, I mean dataset)
Of all the operating systems I’ve used, I think VMS was the easiest to learn, and the least frustrating to program in. Unix/Linux won the market, and deservedly so, but it lacks the consistency of VMS.
“Why else would they have though IEFBR14 is the right name for a program which is nothing but a return statement?”
IEFBR14 is the assembler code for branch return – you obviously never studied 370 Assembler.
Originally it was a one op-code program, but at some point code bloat reared it’s ugly head and the register was formally set to zero before the branch instruction – doubling the lines of code in the module!
https://en.m.wikipedia.org/wiki/IEFBR14
I started using a PDP-5 in college, (OK, I am dating myself). It had 12 bit words, 4K of memory (yes K, that is not a typo). Assembly language was the only programming tool, paper tape (the stuff with holes in it) was your I/O. It did have a CRT graphics display. You would position the beam, then illuminate a dot.
I moved on the the big time, a PDP-8i and a LINC with two LINC tape drives, wow, tape without holes. This was sort of a DEC machine. Out in the real world I met the DEC 10 and a CRT terminal and Fortran II. 10 years later at a new job there were VAX780s and PDP-11s, we used for laboratory data collection. 11/23s in the lab, VAX 730s as data concentrators, VAX780s for data crunching. Connected with big yellow. 10 mbit Ethernet. Used micropower Pascal to program the 11s.
Can you tell us more? How was the “environment”? (I presume that a PDP-5 wasn’t a shared-time system…)
The PDP-5 I used in 1967 at Washington University (St. Louis) was in a plain 19″ rack, not the nice one with doors that DEC sold. The usual front panel switches to toggle in bootstrap, an ASR33 teletype. It was the precursor to the PDP-8. Strictly single user.
My experience with DEC was just as a ‘user’ at High School and College in early 80s. Both were VAX machines though. High school was intro to programming on paper fed terminals in BASIC. In College it was Pascal, Fortran, and Assembly for the most part. Also bought a DEC Rainbow for home use at the time. Dial in to VAX at 300 baud. Finally I think 2400 baud before I graduated.
Enjoy my little PDP 11/70 front panel. Key in a boot sequence from the front panel and go … Or just use the menu in simh. Also just sitting there blinking lights is soothing too :) . The RPI-4 behind the scenes though is doing actually work. It is running a range of useful services in the background like Pi-Hole, NTP server, redis server, etc. So this system is running 24×7.
In 1981 I worked at an electrical utility where we had a VAX 11/780. It was in its own big air-conditioned room. I was developing an application using FORTRAN to calculate and optimize the buying and selling of electricity between neighbouring power grids. It used a Monte Carlo simulation technique plus linear programming.
My program really pushed the poor 11/780 to the limit. I wasn’t the only one using it of course, and other engineers ran their transient stability programs and loadflows on it. At that time most of these big programs were all run in batch mode overnight. Initially, I was just doing testing so I only ran one month. My program was a barn burner and that used 20 minutes of CPU time. But one thing that I didn’t know at the time is that for some reason, the VMS operating system wouldn’t run other programs in memory very much if one program was doing lots of heavy duty computations. It would switch when that program did some I/O. Monte Carlo and linear programming is very CPU intensive. So at first holding up the computer for 20 minutes wasn’t much of a problem and it didn’t affect the other users.
The first time I ran a full blown test simulation of 12 months, when I arrived at work there was a posse waiting for me at my desk. My run had tied up the computer almost continuously for half the night!! Many of their runs didn’t even get started! So I very quickly learned that I had to limit my usage by reducing the number of Monte Carlo iterations and only doing the full run when I had to prepare a forecast (I had it do some I/O to debug files while running to give the others a chance to get some CPU cycles!)
Some 25 years later, I was no longer using that program and it was the PC age. I was curious to see how long it would take to run on my home computer. I compiled my program on my PC which used an AMD Athlon CPU running at 2.8 GHz. So instead of 4 hours it took . . . . . . . . . . . . .35 seconds!!!!!!!! Makes sense as the 11/780 only had a clock speed of 7.52 MHz. How far we’ve come!
I should have said above that my program used 20 minutes of CPU time per month simulated, hence 4 hours for 1 year.
I too was working on a 11/780 in 1981 – and yes, you could treat it like a pc (modern pc) and use 100% of it on a task.. I did that regularly, as my view was there was no point having it ideal overnight :-)
But you could drop it’s priority so that your job ran effectively the idle task, and then let it run for many days..
Work wise I did that with data, spare time wise I was writing ray tracing algorithms which where a little bit to much for the hardware.. I had a (at the time) fantastic techtronics 4027 graphics terminal.
It’s not that VMS prioritize compute heavy programs. It actually prioritize interactive programs. However, anything running from the batch queues is deprioritized, so your program would starve any batch jobs. :)
My high school (circa 1970) had two ASR-33 terminals connected to a 32K PDP-8 (with a 64K disk!) running TSS/8 that was shared by a number of high schools in the Boston area. I learned how to program BASIC, FORTRAN, and MACRO-8 assembly language on that thing. When I went off to college, I intentionally picked a school (Worcester Polytechnic Institute) that had a PDP-10 that was freely available for use by students. I learned far too much about that 10 and ended up getting expelled near the end of freshman year after an unfortunate hacking incident.
Now needing a job, I stumbled across a small office above the Woolworth store in Central Square in Cambridge. A few PDP-8s, 11s, and early DG Novas in racks, several ASR-33s and 35s with dedicated lines to one of the PDP-10s at company headquarters, plus some desks. I was in heaven. Our primary task was writing embedded scientific and automation systems on minicomputers, all in assembly language. Development used in-house proprietary MIMIC cross-assemblers, linkers, and CPU/peripheral simulators running on the PDP-10. Code was punched to paper tape on one of the ASR-35s (30 characters per second, woohoo!) and loaded into one of the target machines for final testing. I eventually wrote some simulated peripherals and the entire simulator (IIRC) for the long forgotten GRi-99/909 minicomputers.
My manager back then, Bob Supnik, wrote most of the MIMIC code, and started translating it to C as a retirement project about 20 years ago. The SimH simulators running on the PiDP-8 and PiDP-11 are basically the MIMIC code I was using in 1973.
The University of Saskatchewan had a PDP-11 in the very early 1980’s. They got an educational Unix license from Bell Labs, which arrived on one RK05 disk pack. The OS source code had the best comment ever
/* You are not expected to understand this */
I still remember the crazy flashing lights off the address and data bus along with the a boot address of 773164. When the system was near idle you could see the keystroke interrupts being processed. It was also very easy to tell when someone was running Ingress
Darwin Peachy (later of Pixar fame) was a grad student at the time. Fastest typist ever
At University of Regina in ‘74 there was a PDP 11/20 we could get stand alone time on. Also some early video terminals in labs and around campus offices with a Gandalf terminal multiplexor. I was there maintaining a Xerox Sigma 9 mainframe (formerly Scientific Data Systems) and jumped ship when Honeywell bought that Xerox division. I got on with DEC then in ‘76.
The screen saver for my Onyx Boox Max 3 e-reader is the front page of the “Introduction to Programming | PDP-8 Family Computers” circa 1969. My old college sweetheart. Bytes were 6 bits and proud of it!
We had a vax 11/70 and one of two PDP11/44’s in a machine room at work (Marconi, 1985).
Mostly running UNIX with VT220’s and we had mutiple termcap ascii windows on them too.
There were windows so you could see in.
One day the computer manager (Steve?) was in there fiddling with something and the Halon went off.
The room filled with vapour but you could just about make out a figure standing up sharpish and running around between the machines.
Of course the whole office was watching by now, with some amusement, but he did make it out within a few seconds completely unharmed, if a little dishevelled.
The Heathkit H-11
link … https://en.wikipedia.org/wiki/Heathkit_H11
I spent the late 80’s deploying 11/70s, 11/84s in newspaper production. A rock-solid platform that did so much with so very little by today’s standards. 173000g! I miss those beasts.
I started work at DEC in Ann Arbor MI as a sales Engineer in February 1969. To be in sales at DEC you had to have previous computer design experience, which I had from 3 years at Burroughs. I handled the Univ of Michigan and several other interesting customers- including Heathkit. In 1971 I moved to Maynard and worked in the medical Systems groups, did Product management for Mumps, later DSM-11. I moved through Sales Training, the Education group, the Rainboaw group and the VAXstation 8000 graphics system. I moved back to the field as a Unix consultant and supported several big database systems. I retired from HP 2 weeks after they merged with Compaq. It was a great ride, right in the center of some of the most exciting work in the computer industry. I still like to tinker with things like Arduino, and compare their capability to the first PDP-8i I sold to Owens Illinois to run a glass plant- a 12K machine with a 32k disk that cost over $60K. But it did the complete control of a mixing system to make glass bottles. In my years in technical marketing I traveled over the world, from Japan to Rio to Sydney. I saw the world, met some of the greats of the industry and had a blast doing it.
The FUBAR register is actually a register of the DW780 Unibus adaptor for the VAX-11/780. Still a great name for a register, though.
Quite a trip down memory lane. My first experience of DEC machines was writing software on a PDP11/23 under RT11 to run an $0.5M radio receiver in a mixture of FORTRAN and assembler at GTE in the early 80s. I learned quite a lot – who knew that accidentally leaving ,, in your link map file would cause the linker to call an anonymous block of blank memory into the run time image – and debugging assembler code manipulating binary integer numbers greater than 18E9 is a real PITFA.
After that I left and went to work for DEC at the Maynard Mill. One of the interesting aspects of DEC was the fluidity of the development process. We developed a lot of products that never saw the outside world. In particular the comms group at one time had multiple products aimed at the newly developing telecomms and networking area with surprisingly overlapping features. They seemed to be working on the theory of survival of the fittest.
I used to run into Ken Olsen from time to time, which was not really that surprising as the EMC group occupied the office space next to his. Usually this would happen when my tech and I were wheeling a cart load of new product out the front door to go test it at one of the open field test sites. I very soon discovered that I’d better be very familiar with the product features and project status as Ken would immediately stop to chat and ask about the product and any problems we’d run into and how we’d solved them and how soon I thought we’d get the FCC and VDE approvals. I gathered that way back there had been a long saga of reliability issues when getting VDE approval for the VT100 terminal so EMC was a subject of very keen interest to Ken. Ken was much respected in the company for many reasons. He was certainly no lightweight unlike many other executives I’ve run across.
DEC was for that time quite an enlightened company. One of the things they’d done was to open a manufacturing plant in one of the poorest districts in Boston with the successful goal of employing local people to manufacture the LK201 keyboard that was used on pretty well every DEC product at that time. The LK201 was a tidal change in keyboard design – most modern keyboards owe their layout and construction in some way to the LK201 but the original design had an unsustainable BOM cost and the manufacturing plant was struggling to stay in business. The circuit designer fortunately got me involved in some Eco changes that he needed to get approved and I noticed that the grounding and related physical design of the product were actually a bit bass ackwards. A couple of simple grounding changes and we were able to strip over $5 off the BOM parts cost while still reliably meeting Class B emission requirements and so keeping the manufacturing plant in business for a few more years.. One of the very few times in my career when I felt like my expertise made a direct contribution to society.
And there were many other good and important things about engineering life in DEC that were ahead of their times but are seemingly completely missing online and which should not be lost to history. TECO ( the cryptic scripting language that masqueraded as a lowly text editor). Department meetings at the Boylston Conference Centre ( food you would absolutely die for) . The most excellent Department Leaving Lunches (especially the final one when I split to go and join Data General) . And hopping on a chopper anytime you needed to get to an offsite meeting or Logan airport in a hurry – note , I’m still waiting for those autonomous flying taxis to appear.
I have done work for a number of companies – British Aerospace, GTE , DEC, Data General, Axil, Stratus, Intel, IBM and a few others.
However DEC was by far the most interesting, challenging and progressive engineering organization I ever worked in.
Back in the 1970’s I designed some peripherals for use with a PDP-11 (-40, IIRC). Without a doubt, its bus had the WORST architecture I ever encountered. Generating an interrupt was a very complex process, and it was absolutely impossible to make interrupts work using synchronous logic. And I was working for a guy who would not permit any crash logic whatsoever (a position with which I wholeheartedly agreed). Even the Navy’s NTDS bus and crazy architecture Univacs were easier to handle.
When I started college, my school had a DEC System 10, which if I remember all these years later was a PDP-10 running TOPS-10. I used Fortran and BASIC on it. I first experienced email and Usenet (before the great renaming) on that system.
PDP-8, PDP-11, VAX, DEC-10, DEC-20, worked for DEC (corporate IT, not development), used VAXes as a client
But back in college, I was popular as if we wanted to run the system outside of established hours, I was the only one who could toggle in the bootstrap code to start it running… lol
I remember first learning about interrupts on a PDP-8. Instructor got fed up with us not understanding it, so he went to the rack, pulled the panel, pulled a few boards, then said “There. That wire there. That’s the interrupt!”
PDP-8 with large floppy. Like 8″ floppy. Guys pulled the leg of the Lab Tech one morning at breakfast, telling him that the ‘record player’ wasn’t working and he couldn’t get his Beatles 45 out of it. Only time I ever saw him run.
VMS was a great system to administer, and work on. Except for the standard service account, with full privileges, being installed with a password of “p*****” everywhere, so the field service techs wouldn’t be delayed when they came on site…
Amazing how the main backboard/backplane (?) for the VAX-11/750 were made. Scare one off of wirewrap for life. Factory floor claimed there was one woman in the whole company that would trouble shoot those. Claimed that she’d sit there looking at it for up to a minute, then reach out and point ‘That’s the one that’s wrong.’
At school in the early 80s, one of my best mates’ dad had a pdp-11/34 for metallurgy at British Rail Engineering Limited in Derby. Compared with the ZX Spectrums we had it sounded amazing*! By the time I went to Uni (UEA, East Anglia, where the Climate Research Unit of the fake ClimateGate controversy); we had just phased out the pdp-11s leaving a VAX-11/780 and MicroVax I for undergraduate use. The VAX, which we programmed in Pascal did seem enormously powerful at the time and the University’s allocation of 1 minute of run-time per day per computer science student was usually easily enough. We also tried to hack directly into the memory of the VAX by creating variant records containing an integer and a pointer (but that didn’t work, mostly because we didn’t understand the memory map well enough).
In the early noughties I managed to pick up a MicroVax II and micropdp-11/73 from, of all places, a pet food warehouse in Sheffield that was decommissioning lots of old DEC hardware. We shoved both of them and a couple of monitors and keyboards and spare expansion boards into my Mk 2 VW Golf and then managed to drag it back home. We didn’t manage to use them much, because they were already password protected and we didn’t have the resources or time to get past that.
In the end I blew up the micropdp-11/73’s PSU when trying to figure out a compatible baud rate to talk to my Mac a few year later (it’s partially fixed now). Both the Microvax II and pdp-11/73 are sitting at my Dad’s house now!
https://oneweekwonder.blogspot.com/2008/04/rip-11.html
*Later I figured out the pdp-11/34 isn’t a whole lot slower than a ZX Spectrum for most integer programs, it was just that compilers were so much better on the pdp-11.
https://oneweekwonder.blogspot.com/2013/12/z80-dhrystones.html
I started working for DEC in 1969, and retired from DEC when Compaq Computer bought them in 1998. I’ve built PDP-10’s, the PDP8-E’s, and worked directly with the engineers developing the PDP-11 in the “New Products” group. My years with DEC as an engineering tech is probably the most fulfilling years of my life.
The college I went to from 1977-1981 had a pdp-11/40 with 112KW of RAM, an RP-11C disk controller, 2 RPR-02 and one RP03 disk drives (a total of 80MB of on-line storage), a 9-track tape drive, a pair of TU11 DECtape drives, a bunch of DZ and DH serial multplexors and a pair of parallel line printer drivers. Ran RSTS/E with a 30 user max. This was the only multi user system we had. I learned programming and computer organization on that machine.
On graduation, I moved to New Hampshire, then MA, to work for DEC. Did graphics firmware for the Professional-300 line, designed video terminals, and learned a lot.
I started working on a PDP-11/60, running RSX-11M, in 1978, programming in FORTRAN. My task was to analyse and present results of sound measurements in a military environment. The results were presented graphically on a VT-55 terminal, together with the ‘bog roll’ thermal printer on the side. Disks were RK-06. Memory was a serious limitation and I quickly had to learn about overlays.
From that beginning, I carried on working in the DEC environment, through VAX-VMS and alpha-VMS, up until the 2000s.
Well I was in a Navy spook squadron in the early ’80s and we had an 1173 in our maintenance spaces and another one in our scif it would take several hours after each flight to write up the mission in the scif One of our technical support people in Virginia said hey you guys need 1173 for each one of your planes so he shipped out two complete systems and took us about 2 weeks to install it but we ended up with six terminals and two printers in each plane we had a 5 mb hard drive that we would load up the mission parameters in the SCIF take it out to the airplane boot up the planes 1173 upload all the stuff fly the mission download it to the briefcase and upload it to the scaff we save 3 hours for each mission it was great I was told that the time is the first airborne installation of 1173 I don’t know maybe
Oh we also had dentures and dragons running on the airplanes and the maintenance 1173 it proved to be so popular they only let us run it during lunch time or after work You’re in a room there’s a door and a window…
In ’92 I went on an industrial placement as part of my MSc. Most people were using PCs at the time, but I was using a VAX to program Basic. I used one of those orange terminals. The machine itself was in a portacabin. It was hot,and it had fans in there. Whenever someone turned the fans on, he turned them off again. I kept running out of disk space. The admin asked me how much I wanted. Tentatively, I said 1M, which he granted. I was thinking back to my ZX81 days when I came to the conclusion that I would never be able to fully use 56k of RAM.
Shortly after that I started a PhD. The group I was working for had a bit of dough, and we bought a Dec Alpha and some PCs. The Dec cost £20k, whilst the PCs were under a tenth of that at the time. Most of the university was using Windows, with stuff like LaTeX. The open source ecosystem was much parser in those days. Nowadays I don’t think there’s any good reason for universities to use commercial software.
In hindsight, I think the Dec was a waste of money. Sure, they were more powerful than the PCs, but in terms of bang for buck, PCs were a better bet. Looking back, it seems obvious that the days of minicomputers were numbered.
My 1st DEC machine was an underused PDP-8; the revelation for me was the complete doc set — descriptions and diagrams of exactly what all the bits did. The PDP-11 was a great learning machine; somewhere, I still have my 11/34 manuals. Used several PDP-11 and VAXes in grad school at Stanford in the 1980s before SUN workstations and x86 boxes took over.
I have fond memories of the various PDP11’s, VAX’s and their myriad tape drives, disk drives, monitors, etc. that I worked on in support of various spacecraft at NASA/GSFC … today, I found this website in search of RZ40 disk drives for VAX4000-xxx’s still in use. I am happy to report that the Microvax II is still healthy and that I still have spare RD54’s.
Our freshman engineering class voted to use the department’s PDP-8. The computer science department had a DEC PDP-11/20 and a PDP-11/45, later upgraded to an 11/70 running RSTS/E.
I have a DEC VAXStation II/GPX that I want to fire up and see if it works and install BSD on it.