back
157 comments
Brings back some memories because this was the first "large" program available to me for study when I was first learning to program, around 1978.

Having talked my way into using the local university EE dept's 11/45, I found the source tree was mounted (they had a 90MB CDC "washing machine" drive) and decided to print it out on the 132-col Tally dot-matrix printer.

Some minutes into this, the sysadmin bursts into the terminal room, angrily asking "who's running this big print job".

I timidly raise my hand. He asks what I'm printing. I tell him the C compiler source code because I want to figure out how it works. He responds "Oh, that's ok then, no problem, let me know if you need more paper loaded or a new ribbon".

I have to say, the way indentation and brackets were done here looks like it's just inviting subtle bugs. Take this for example:

    if (peekc) {
      c = peekc;
      peekc = 0;
     } else
      if (eof)
       return(0); else
       c = getchar();
If you judge the coding style of the early 1970s with modern standards, it isn't going to look great. 50 years is a loooong time.

dmr was one of, if not the first C programmer(s).

And he was almost certainly using ed(1) as his editor and a mechanical teletype at 7.5 or 10.0 characters per second as his terminal...

The C language (and all of Unix) was designed to be very terse as a consequence.

If you can only read it line by line:

    return(0); else
makes a bit of sense.
I dont think it would look that out of place if he was using the ternary operator which is the same thing after all.

E.g.:

  if (peekc) {
    c = peekc;
    peekc = 0;
  } else
    eof ?
      return(0) :
      c = getchar();
The first else clause sill looks weird, but the final part isn't nearly as out of place (well i guess assigning in a ternary would be weird, but in terms of indentation) and its not like we actually changed anything.
Can't return in a ternary.
It would have made code review a nightmarish activity for the team.
Why nightmarish? A reviewer may explain that it is prone for other human to overlook the end of conditional and go sleep as usual. I never understood the emotional component of blaming someone’s personal code styles, as if it were a religion with sacrifices and blasphemy instead of just a practical way of coding in a heterogeneous team.

This triggers me because many people jump on “bad c0de” in the forums, but then you read some practical code on github, and it is (a) not as perfectly beautiful as they imagine at all and (b) is still readable without nightmares they promised and (c) the algorithms and structure itself requires programmer’s perception and understanding levels far beyond the “it’s monday morning so i feel like missing an end of statement in a hand-written scanner” anyway.

That team had Bell Labs researchers, Ken Thompson and Doug McIlroy being among them. Their brains could handle much harder things!
The difference between Dennis Ritchie and the average programmer of today is, Dennis Ritchie did not write bugs in the first place.
Well, some say (I disagree) that C is a bug.
And what does it mean there? (automatic storage duration)
It means it lives on the stack, so it "exists" until the function returns.
Do you mean automatic type deduction?

Edit : I know it’s also storage specifier but it also does type deduction here, hope I’m not confused with the terminology

The original keywords - https://github.com/mortdeus/legacy-cc/blob/master/last1120c/...

Interestingly, long was commented

It makes sense long would have needed a comment. It needs a comment because “long” and “double” are terrible names for data types. Long what? Double length what? Those type names could easily have opposite meanings and meant long floating point / double length integer. WORD/DWORD are nearly as bad - calling something a "word" incorrectly implies the data type has something to do with strings.

If you don't believe me, ask a non programmer friend what kind of thing an "integer" is in a computer program. Then ask them to guess what kind of thing a "long" is.

The only saving grace of these terms is they’re relatively easy to memorise. int_16/int_32/int_64 and float_32/float_64 (or i32/i64/f32/f64/...) are much better names, and I'm relieved that’s the direction most modern languages are taking.

(Edit: Oops I thought Microsoft came up with the names WORD / DWORD. Thanks for the correction!)

> ask a non programmer

Why should a non programmer understand programming terms? Words have different meanings in different contexts. That's how words work. There is no need to make these terms understandable to anyone. The layman does not need to understand the meaning of long or word in C source code.

Ask a non-golf player what is an eagle or ask a physicist, a mathematician and a politic the meaning of power.

Word and long may have been poor word choices, but asking a non-programmer is not a good way to test it.

> Microsoft’s WORD/DWORD are nearly as bad. Don’t call something a “word” if it doesn’t store characters.

"Word" as a term has been in the wide use since at least 50s-60s, you can't really blame MS for that

https://en.wikipedia.org/wiki/Word_(computer_architecture)

Yes, int_16/int_32 or something like that makes a lot more sense. Today, not when this compiler was written.

The PDP-9, PDP-10, and PDP-18 have 18 bits registers. The world had not settled on 16/32/64 bits at all.

Even the intel 80286 far/fat pointers are 24 bits.

It could not be clearer that you didn't read the page referenced by the comment you replied to.

"long" was commented out.

No, the names are fine and self evident after glancing through K&R for 15 min.

The real mistake in retrospect is that int and long are platform dependent. This is an amazing time sink when writing portable programs.

For some reason C programmers looked down on the exact width integer types for a long time.

The base types should have been exact width from the start, and the cool sounding names like int and long should have been typedefs.

In practice, I consider this a larger problem than the often cited NULL.

int is neither 32 nor 64. Its width corresponded to a platform’s register width, which is now even more blurred because today’s 64 bit is often too big for practical use and CPUs have separate instructions for 16/32/64 bit operations, and we agreed that int should be likely 32 and long 64, but the latter depends on a platform ABI. So ints with modifiers may be 16, 32, 64 or even 128 on some future platform. intN_t are different fixed-width types (see also int_leastN_t, int_mostN_t, etc in stdint.h; see also limits.h).

Also, don’t forget about short, it feels so sad and alone!

> It makes sense long would have needed a comment.

I think you misunderstood. There's no explanatory comment. The "long" keyword is commented out, meaning that it was planned but not yet implemented.

    ...
            init("int", 0);
            init("char", 1);
            init("float", 2);
            init("double", 3);
    /*      init("long", 4);  */
            init("auto", 5);
            init("extern", 6);
            init("static", 7);
    ...
There is no short or unsigned either. Maybe int modifiers were pending work?

for is missing too.

Very interesting. Does anyone have any idea how the C compiler was bootstrapped. Was it first written in B, or was it written in PDP-11 assembler?
It was written in B. And B was written in B. To trace the bootstrap up into the high level, you have to go further back than the PDP-11. To bootstrap B on the PDP-7 and GE-635 machines he had access too, Ritchie wrote the minimum subset required to compile a compiler in B, in a language known as TMG (in the vein of, and a big influence on, Yacc and similar). This minimal bootstrap was then used to compile a B compiler written in B, and further development was self-hosted.

Later, the language would be retargeted to the PDP-11 while on the PDP-7. Various changes, like byte-addressed rather than word-addressed memory, led to it morphing into C after it was moved. There was no clear line between B and C -- the language was self-hosting the whole time as it changed from B into C.

Mr. Ritchie wrote a history from his perspective published in 1993. I've mostly just summarized it above. It's available here: https://web.archive.org/web/20150611114355/https://www.bell-...

The basic sequence was:

- Rewrite B compiler in B (generating threaded code)

- Extend B to a language Ritchie called NB (new B). This compiler generated PDP assembly There is no version of the NB compiler known to exist.

- Continue extending NB until it became the very early versions of C.

You can read the longer version of this history here:

https://www.bell-labs.com/usr/dmr/www/chist.html

I think it was written in B as Ritchie extended B for the PDP-11. By that point the B compiler was written in B

https://sci-hub.do/10.1145/155360.155580

Flint knives and bear skins.
So do I get this right: this is not the first C compiler (since that one would be written in B) and the first C compiler written in C?
Thank you for pointing this out. It obviously can't be his first C compiler because it's written in C! :)

I'm not even sure it's the first C compiler written in C, though - it just says in the github description "the very first c compiler known to exist in the wild."

Regardless, if it's from 1972 it's a very early version.

It's not uncommon for a compiler to be partially written in its own language, so therefore this may as well be the first C compiler.
For the hobbyists and the curious using 2.11 BSD, its C compiler is the end of the line for development of the Ritchie C compiler, as far as I know.
I tried to find the last version that Dennis wrote: my findings are here:

https://github.com/dibyendumajumdar/C

It is surreal feeling when looking at first code, not sure how to explain why it is the case.
I felt the same upon seeing an actual Dead Sea Scroll at the Jordan Museum in Amman.

The feeling dissipated somewhat when our guide explained that it’s a treasure map

This is gold. I like how the files are named 00 01 10 11 because long filenames are scary.

    hshsiz 100;
    hshlen 800; /* 8*hshsiz */
    hshtab[800];
These were at file scope. I assume they default to int, but when was the demand for = added?
For those who wonder, the extract comes from https://github.com/mortdeus/legacy-cc/blob/ccbe90a2803e2cc7e...

IMO, it is close to Assembly:

* you reserve space and possibly set an original value with "hshsiz DB 100" or "paraml DB ?"

* assignment is a different business which involves a runtime instruction (MOV)

Hence the same difference (no = sign for the first, passive, compile-time operation; an = sign for the second, active, runtime operation) in this proto-C.

(Of course, this doesn't answer your question of "when" did the syntax of those 2 operation fuse :-) )

Long filenames are scary when your system has a 2.5MB disk.
It is very interesting how many hardcoded magic numbers it has. No headers and defines because they did not existed yet?.

The full compiler is super compressed. Hundreds of lines per file. It would be great if there was an explanation of the general idea of the design somewhere.

I'm only seeing a few thousand lines of code, so I think it would fit in someone's head easily enough. Might be the numbers were on paper.
So how do you "bootstrap" compile this? I mean without cheating, say on a rebuilt computer after all computers where destroyed?
Hand-compile it for the initial bootstrap would be one way to go. That, or write a minimal C in assembler, then make it self-hosting, then expand it into the full implementation.

Edit: Elsewhere in-thread, retrac posted a detailed summary of how C developed from B.