[drfuchs]

The annotations say "... I don't have the exact instruction format for the either the LGP-30 or RPC-4000 ..." but the entire manual for the LGP-30 with this information, and way more, is available at http://www.bitsavers.org/pdf/royalPrecision/LGP-30/LGP-30_Op...

In 1973, Mr. Willoughby taught the Abington (Pennsylvania) High School computer programming class to code in LGP-30 assembly language. We didn't actually have an LGP-30; it just happened to be what he had been taught on when he was young. All assignments were graded by him simulating your code in his head. Of course, we then went on to learn the (slightly) higher-level NEAT/3 language, for which the school actually had an NCR Century-100 mainframe that would run programs that we submitted on punch cards. Mr. Willoughby's (nobody knew teachers' first names back in the day) theory was that it was important to learn the lowest-level machine language first, so you could understand what was really going on underneath. Worked fine for me; evidently it's not quite so universal anymore.

[WillAdams]

While not universal, I do recommend Charles Petzold's _Code_:

https://www.goodreads.com/book/show/44882.Code

which covers things down to a quite low-level and was recently updated:

https://www.amazon.com/Code-Language-Computer-Hardware-Softw...

[jpgvm]

I think going as far as assembly up front is probably not worth it, also it's hard to keep people engaged at that level as it's difficult to make the computer do exciting things with assembler.

However I strongly believe it's good to start with C. You can still rather quickly do interesting stuff, C is a small language so learning the language itself isn't a huge barrier.

A big benefit of the small language is that it leaves more time available to explore important concepts, not just the super low level ones (memory/pointers, etc) but really important parts of the stack that are infectious to everything else. Specifically things like syscalls and the libc interface that most other languages are essentially built on top of. Working with building blocks like pthreads directly is also very important IMO, both to get a handle on parallelism and concurrency primitives but also to learn why high level languages are so valuable. Similarly for stuff like important socket APIs ala select/epoll and implementing your own event loop.

I was lucky enough to learn all of this early in my career and luckier still to have been able to pass on this knowledge to many mentees over the years.

If there are any aspiring programmers here that want to build from a solid foundation then yeah, ignore the haters, write some C.

man pages and ironically ChatGPT are your friend, use them to explore the foundations of (most) modern code so when you start writing modern code in earnest/for money you will be substantially ahead of your peers in actual understanding.

[daemin]

No, go down to assembler, just pick a smaller machine like a microcontroller where simpler programs can do more things more easily.

When you know assembler you can always see what compiled programs run as no matter what language that program was written in, if you start with C you won't have that ability.

[pjmlp]

Assembler is much better approach, one gets to learn how the machines actually work (ignoring the microcoded part), and reach for something like a 8/16 bit computer emulator (ZX, C64, NES), or an Arduino like device with ARM/RISC-V.

Also the realization, C isn't that special, plenty of ways to play around with pointers.

[reverendsteveii]

If it makes you feel any better as of 2005 or so Pitt's computer science degree still includes learning assembly language and the particular Turtle-Headed Sadist that teaches it at the Johnstown branch campus won't let you use hex opcodes for the first few weeks of the course so you're literally coding in unadorned 1s and 0s. I hated him for it. I hate him for it. But nothing compares to the dopamine pop I had when figured out how to do long division using only addition, subtraction, conditionals and jumps. Not that that's a groundbreakingly difficult problem, it was just one of my favorite "aha!" moments in my entire career.

[elsjaako]

In the story it says:

> The new computer had a one-plus-one addressing scheme, in which each machine instruction, in addition to the operation code and the address of the needed operand, had a second address that indicated where, on the revolving drum, the next instruction was located

Is this correct? The manual you have seems to have a diagram with a +1 operation on the "counter register", which loops back down.

All instructions seem to use a memory address, but they use this for doing the instruction thing (adding, subtracting etc).

Maybe I'm just not understanding the format.

[drfuchs]

The RPC-4000 (the "big brother" of the LGP-30) had each instruction specify the next instruction address. I believe this was to allow for optimizing your program such that the next instruction would always be right under a read head on the drum when the processor was ready for it, because if you missed it, it took a whole revolution of the drum to get back to it (kind of like a cache miss).

In any case, it seems that while Mel wrote lots of code for the LGP-30, the actual hack in the story involved code that Mel was porting from LGP-30 to RPC-4000.

[TYPE_FASTER]

Sounds familiar. One of my first classes as an undergrad started with boolean logic and boolean gates, then IIRC (it's been a minute) assembly for a theoretical processor.

[riffraff]

I had a class (in the late 90s/00s) where we had to code to a theoretical RISCish CPU, but the teacher had built a simulator for it which allowed you to inspect registries, flags, execute instructions step by step etc

It was quite entertaining.

[marco_craveiro]

Same here. We used LMC [1] before we moved to a real architecture.

[1] https://en.wikipedia.org/wiki/Little_Man_Computer

[pjmlp]

On my case, on digital circuits lectures we did the whole thing, starting with boolean logic, gates implementation, using stuff like SPICE, eventually we designed our own toy CPU, the actual implementation on a breadboard was left as an optional exercise, which on my year I think no one did.