Entering and Running a Tiny Program from the PDP-11 (PiDP-11) Front Panel | Hacker News Reader

Entering and Running a Tiny Program from the PDP-11 (PiDP-11) Front Panel

bigdanzblog.wordpress.com

50 pointszdw23 comments

There's certainly nostalgia this kind of thing evokes, but booting up SIMH and puttering around in RSTS/E or RT-11 is usually enough to put it to bed. No need to put hundreds of pounds of power hungry equipment in my garage.

Our high school had a PDP 11/34 locked away in a server room, we were only ever allowed to use the VT52s, Tektronix 4014s and later the much coveted VT-100 out in the main room. I wouldn't classify it as ASMR, exactly, but the somehow warm and tinny beep of the VT-100 when it booted, its quacking when you typed, or even the relay buzz for each key press on the VT52 are very evocative and fondly remembered sounds.

In another lab there was a PDP 11/05 which we could get our grubby hands on although it required some effort to get going - namely toggling in enough code to get the paper tape reader to read a boot tape that in turn read RT-11 from an RL02.

I learned a lot in those computer classes, not really about programming which I'd already had nailed. But I learned the sysop was impatient and would type his password faster than the LP36's ability to overstrike it because system was so busy that the ^H* came much later. I learned with access to the right manuals we could rewrite the login program itself and guarantee access past the semester's end. And I learned maintaining operational security is paramount because it's easy to get busted in a heartbeat. Lessons that still apply 40+ years later in cybersecurity.

> No need

"But I want muh blinking lights..."

I mean, this depends on whether you are a programmer or, say, more into electronics. Different emulators may emulate different representations of the same device, at different levels (such emulators exist at all).

I tend to agree though that emulation (and virtualization in general) can indeed be superior and more practical than dealing with the "real thing." (Sometimes I even think to myself, semi-jokingly, that the real thing emulates its design made, say, using a CAD software.)

>"But I want muh blinking lights..."

Totally. I wouldn't criticize anyone who does want der blinkenlights. More power to them! Was just saying at least for me booting a sim was (so far) enough to scratch the itch.

I have one on my desk, like the video in the linked post, but I used it more as an adult Lego set to make an elaborate toy on my desk than to spend time entering programs in octal via switches.

Toggling is the word we used. I only had the chance to do it once, since the PDPs were on their way out, but it did feel real: it's as close to the metal as I ever got.

Gotta admit, PDPs look really cool, and even modern. The toggle buttons, the blinking lights, it's all so fascinating to watch. Then, after 5 minutes, it occurs to you "If only there was a more efficient way of doing this".

I've never used a PDP, mind. I'm not quite that old.

I'm that old. I toggled the PDP-8 our high school had. I did it often enought that for over a decade after I stopped I had the RIM bootloader memorized. It's only a handful of instructions. You can see it on page 35 of http://bitsavers.trailing-edge.com/pdf/dec/pdp8/software/DEC...

I have a few old timey computer stories, including using a punch card to transport a chunk of nitrogen triiodide. Hijinks ensued.

I'd love to hear the stories.

DEC thought the same thing; so they added peripherals like Boot ROM cards. It's interesting to see how the PDP-11 series evolved from the beginning with teletypes and punch tapes, requiring the operator to toggle in programs on the switches into ferrite core memory, and ended up with multi-user UNIX systems with multiprocessing, Winchester drives, and networking.

I helped a friend with an IMSAI he was trying to get working, and did some test program entry using the toggle switches. It took a few tries to figure out the data entry and what it was telling me. Then I got it figured out. Then I was tired of it, and wished there was a boot EPROM.

Time elapsed, 15 minutes.

It was a damned shame when they took away the switch register, my first recollection of that was the 11/34, it had 7-segment LED displays and calculator style buttons. You could still load programs, but you did not have to convert octal to binary in your head to do that.

example: http://www.technology.niagarac.on.ca/people/mcsele/wp-conten...

Some later models simply had a boot button. No more blinkenlights. So sad.

I have always admired PDP-11, but the closest I have come to the “manual” programming was with an IBM/370, when I managed to type a small program, in hexadecimal, on a data entry device that used a magnetic tape, booted the system from the tape, and proudly showed to my peers a silly dialog with the typewriter console. An unforgettable experience.

For more on this kind of programming, check out the excellent deramp5113 channel on YouTube. He goes over how programs were entered on the front panel of the first desktop minicomputers in the 70s like the Altair 8800, and then how that transformed into paper tape and PROMs. Plus lots of other great early computing stuff like operating systems and teletype operation.

https://www.youtube.com/watch?v=suyiMfzmZKs&list=PLB3mwSROoJ...

(Also available at the Internet Archive for a no-ads but slightly clunkier experience: https://archive.org/details/movies?query=creator%3A%22mike+d... )

My dad used to work with a TI 980 minicomputer, it had similar panel switches and he’d have to toggle in instructions to start his program. I’m not sure that machine had an OS as such. I was too young to really be interested beyond watching the blinking lights

I'd really love to have one of these (real or PiDP), but you can have all the fun in the browser from here:

https://skn.noip.me/pdp11/pdp11.html

Heh, repeatedly toggling some of the switches too rapidly on there causes some interesting growth. My "EXAM" switch ended up dangling down past much of the documentation.

Has anyone come across any resources for instruction sets optimised for hand input? I assume there are as many pre RISC instruction sets optimised from programming in assembly directly rather than as a compiler target.

Guy I worked with said he used a mainframe that you had to clock in four 32bit words and then hit run and it would boot. And a newer one where you had to enter one 32 bit word. I think the instruction was like read X byes from the input device port into memory starting at address 1.

I don't know that it was done to optimize for hand input, but see the instruction list for the PDP-8:

https://gunkies.org/wiki/PDP-8_architecture

Yes:

- programmable calculators in 1970, HP something

- 'human computer', set of instructions how to calculate stuff by hand. Turing had chess engine, there was 'paper computer' in soviet tech hobby newspapers.

- mathematic equations

No I mean with toggle switches.

So it might be optimized for the binary codes to be easy to remember or have as few 1s for common operations.

Traditionally everyone entered "mov -(pc), -(pc)" at the top of memory at least once just because ..... (and then reloaded the bootstrap by hand)

As a high school student, I stopped our PDP-8 while it was running BASIC, then toggled in a little machine loop that changed all the occurrences of the 12-bit value that represented "Clear the Accumulator" with the 12-bit value that represented "Load the Switch Register into the Accumulator". I then loaded the soft-restart address into the program counter, set the switch register to 0, then used the continue switch to soft restart BASIC.

Surprisingly, everything appeared to work fine, until the following day, when someone soft restarted BASIC, but left one of the toggles up (because the soft restart address was octal 0200).

I wrote this up in more detail at https://ctm.github.io/docs/yld/programming/pdp8/first-hack.h...

Reply on news.ycombinator.com