Disobeying jet engines - why?

Given that "design time" is a never ending story for most "programmers", this, however obvious and right, may be not that easily achievable... :-) :-) :-).

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Rich Grise wrote:

As regards life-critical gigabuck systems, if the *programmers* spend more time debugging their code than writing it, that's even worse. That sort of code should be correct the first time it's run.

John

You haven't met many competent programmers, I take it? Even a good software writer such as myself get a bad taste in my mouth when I see somebody entering their program before it's even written. ;-)

Thanks, Rich

I read somewhere that after the first crude C compiler was up and running the next version was written in C.

Software bugs are like herpes. Everybody's got it, some people just don't know yet.

On Feb 1, 10:33 am, Richard Henry wrote: [...]

GCC is written in C.

The FreePascal compiler is written in Pascal.

The same was true of the first assemblers, circa 1955--once they had _anything_ working, it was better than using machine code. I've been reading an interesting book called "IBM's Early Computers" that goes into all of that stuff, including the Fortran II compiler having been originally written in Fortran I, whose compiler was written in assembly.

Cheers,

Phil Hobbs

I'd like to see one written in QBasic. Heh!

Tim

-- Deep Fryer: A very philosophical monk. Website @

Hey, I have that book. It's amazing what kluges they did, when it's sort of obvious in hindsight what a computer should be like. I'd like to some day design a modern, microcoded, byte-oriented computer using a small number of tubes.

The Burroughs machines were all coded in Algol, including the os and the Algol compiler. There was never an assembler. They wrote the first Algol compiler in Algol, and two guys worked side-by-side to hand-compile it into machine code.

Autocoder on the 1401 was interesting. You typed character opcodes directly into core, from the Selectric. Beast of a machine.

There's a newish book, about Watson 1, that's good too.

John

Back in the '80s subsets of LOGO (particularly the turtle graphics parts) seemed to be popularly implemented in BASIC...

I think the new PowerBasic compiler is written in PowerBasic.

John

At least one was written for the Motorola 6800. I used to work alongside the bloke who wrote it.

I wrote a couple of Quickpoint (numerical control) compilers in Basic-Plus. The language facilities were primitive, but in a few months the machinists had evolved some very sophisticated programming constructs. I was impressed.

One compiler drive a huge Whitney nc punch. Every hit shook the building. Whenever I tested new code, I had a few machinists standing around with fingers on big red buttons. But after that, if ever I needed a hole drilled or a bracket bent, I got it.

John

IN one of my college courses (Systems Programming?) I had to write an assembler for a fictional machine in any language available on the campus time-sharing system. I chose PL-1, written on punch cards and debugged in 4-inch-thick listings.

Heck, young'un, punched cards were a huge stepup from paper tape!

John

Dammit, I remember both punched cards and paper tape!

Very early in the 1980's, I had a university computer programming course using punched cards! Yes, I had to keep them in order!

Paper tape is something I saved a small number of programs on before that time, in the late 1970's, and that surely appeared to me to be something that "occaisionally works".

- Don Klipstein ( snipped-for-privacy@misty.com)

om...

Young'un? Perhaps they were a "step up", but it had nothing to do=20 with age. Maybe money. I didn't use paper tape (and most of that=20 mylar) until I was several years out of college, and had long passed=20 the card stage. Both suck equally.

--=20 Keith