Dennis Ritchie has passed away

[porter]

Wow, I didn't C this happening so soon.

Will he be buried at C?

[R-ten-K]

I C what you guys did there. . .

[kramlq]

That is a bit of a chicken an egg problem, no? C was fundamental to the development of Unix. You could not have one without the other.

From what I have read recently. UNIX was implemented first in B or BCPL or some language like that around 1968/1969 and then rewritten in C and assembler and released in 1971.

I can't find the cartoon with the egg and chicken in bed smoking ... and one them says "well that answers that question".

R.

Actually portability wasnt an original design goal, so it was written in assembly in the late 60s. I think in the early 70s, it was rewritten in C (and some asm) to make general changes and enhancements easier to do (including changes for new PDP configurations). Then in the late 70s, someone realised it might be possible to port the kernel to Interdata, and once that was done, the shell and utilities would be a fairly straightforward recompile.

Portability was really just an afterthought, but UNIX was a successful proof of concept for portable OSes (along with OS6 and SOLO). So from then on, people started trying to make kernels portable from the start. BCPL, B and Z were used in OS6, TRIPOS, Thoth and probably others, but C obviously became the language of choice due to the spread of UNIX.

[Oskar45]

That is a bit of a chicken an egg problem, no? C was fundamental to the development of Unix. You could not have one without the other.

From what I have read recently. UNIX was implemented first in B or BCPL or some language like that around 1968/1969 and then rewritten in C and assembler and released in 1971.

I can't find the cartoon with the egg and chicken in bed smoking ... and one them says "well that answers that question".

R.

You are very right. The *original* assembler version of UNIX was ported to another machine in only 1971 [from the PDP-7 to the PDP-11/20; merely 40 years ago]. The "Second Edition" [1972] was written in B [which derived from BCPL]. Due to B's limitations, Ritchie went on to develop C [probably a failure by todays standard]. Afterwards [in 1973], the UNIX system was rewritten in C - since then we are stuck with a rather unflexible environment. I'm not sure I'd consider that a blessing today.

[jan-jaap]

Ritchie went on to develop C [probably a failure by todays standard].

That's like saying that a Volkswagen Beetle is unsafe by todays standards. Or, actually (since C is some 40 years old): a Ford Model T.

[Oskar45]

Ritchie went on to develop C [probably a failure by todays standard].

That's like saying that a Volkswagen Beetle is unsafe by todays standards. Or, actually (since C is some 40 years old): a Ford Model T.

No. Sorry, you are really quite wrong- you apparently didn't get what I meant...

[R-ten-K]

Sigh, here's the story directly from the creator's mouth:


http://cm.bell-labs.com/cm/cs/who/dmr/chist.html

(Before the nitpicking gets out of hand.)

[Oskar45]

Sigh, here's the story directly from the creator's mouth:


http://cm.bell-labs.com/cm/cs/who/dmr/chist.html

(Before the nitpicking gets out of hand.)

... pray, where was I wrong?

[R-ten-K]

where did I say you were wrong?


I just wanted to get the tangential nitpicking pissing context over so this thread could get back to its original intent.

Cheers

[PymbleSoftware]

[ BCPL, B and Z were used in OS6, TRIPOS, Thoth and probably others, but C obviously became the language of choice due to the spread of UNIX.

From what I read recently, one derivation of TRIPOS had the last few lines of BCPL removed from it a handful of years ago, its all C and PowerPC assembler now.

You are very right.

No. I was just regurgitating something I had read a few days prior. I've never been to New Jersey and I wasn't involved so I don't know. I was just merely attacking the "chicken and egg" thing which bothers me for some reason.



R.

[porter]

If C is a failure by today's standards then I would happily accept failure. Complaining about C's inability to do something like complaining that a Model T-Ford was not suitable for use as a lunar lander module.

[Oskar45]

where did I say you were wrong?


I just wanted to get the tangential nitpicking pissing context over so this thread could get back to its original intent.

Cheers

I deeply apologize. Back to the original thread. Just re-read the "Unix Programmer's Manual, Volume 2". Of course, without Dennis we wouldn't be there where we are now...As I said before, I bow before him.

[Oskar45]

If C is a failure by today's standards then I would happily accept failure. Complaining about C's inability to do something like complaining that a Model T-Ford was not suitable for use as a lunar lander module.

Programming languages only teach you not to want what they cannot provide. C is no exception.

[porter]

Programming languages only teach you not to want what they cannot provide. C is no exception.

C was defined before multithreading, object orientation and exception handling, yet all of these can be done in C. The C language was deliberately kept small with the view that everything else gets added as libraries etc. Given that 99% of language compilers, intepreters and operating system kernels are built using C, then C isn't doing a bad job.

[PymbleSoftware]

If C is a failure by today's standards then I would happily accept failure. Complaining about C's inability to do something like complaining that a Model T-Ford was not suitable for use as a lunar lander module.

Programming languages only teach you not to want what they cannot provide. C is no exception.

I am not entirely sure. Years ago I looked at an old version of povray source code. They had object-orientated code in C by use of structs and function pointers. I have seen functional programming in C++ with template meta-programming.

I think this may have drifted from the topic of one of the creators of UNIX and the C programming languages passing.

There is a long list of thing you can do in C that you can not do in C# or Java. I still have a lot of respect for the designers of the language.

Exception handling in C:

Code: Select all


#define TryCatch( cond, label )   if( cond ) goto label


void func()
{

int32_t a;
char *X = (char *) malloc( 0xff );
TryCatch( ( sizeof( a ) != 4 ), CATCH );

/*** code **/

free( X );

return;

CATCH:
            free( X );
             /** rest of clean up **/
}


Something to be said about workmen and the tools they have at their disposal.

R.