4 ms·

266 pages of well-written, edited, fact-checked electronics and computer programming. Monthly. Let that sink in. There is so much good info packed into this on

by sbf501 4y ago

266 pages of well-written, edited, fact-checked electronics and computer programming. Monthly. Let that sink in. There is so much good info packed into this one issue, it is worth terabytes of crap on stack overflow. Clearly there was less information back then, but it was far higher quality. I'd like to understand how C came to dominate the world when LISP could replace all of the scripting languages we used today. Was it tooling on cheaper machines that made C so popular? Or that it was closer to ASM than LISP? I never really formed a good opinion on this, I'm missing lots of history.

mianos 4y ago
C is much closer to the machine and very fast. When I learnt C I went to the book store. There were two C books (K&R and Byte Books C) and one Lisp book. I learnt both. Lisp was fun but, at the time, I never appreciated the introspection. It was only when, a long long time later, I learnt Python and remembered the ability to iterate the elements of a class and thought how handy that was for some programs. C could more easily access the lower level resources as well. I still wonder why one of these, like forth, lisp later Python did not become the standard command line interpreters whereas it would never have been C, and I have written very featured C interpreters.
- pjmlp 4y ago
  Lisp Machines, Lisp just as C, had low level primitives.
hota_mazi 4y ago
To be fair, out of these 256 pages, 200 are really just ads. Your point still stands, but the actual substance is a lot smaller than you describe.
- zozbot234 4y ago
  The advertisements are among the most interesting things about old magazines like BYTE, though. They preserve a snapshot of what the computer industry was like back then. Even the pricing information alone is very valuable.
  - hota_mazi 4y ago
    Oh absolutely! Browsing through these pages today, I find myself reading the ads a lot more than the actual articles, which haven't aged well. But back then... man, these ads were cancer and they made up 80% of the weight and price of the magazine.
    - rbanffy 4y ago
      Those ads made the magazine possible and accessible. Of course, McGraw Hill made a profit, but BYTE would be a lot more expensive if it were not for the ads. One interesting point about the ads is that the time it took for a new technology to appear in the ads was usually much shorter than for it to appear in articles. In that sense, the ads were opportunities to peek into the near future (and, in the case of vaporware, alternate timelines). I like the way magazines like BYTE are preserved and I hope we preserve our current tech as well.
- sbf501 4y ago
  Counterpoint taken.
kazinator 4y ago
C came to dominate the world because of Unix. But not directly, because Unix didn't really dominate the world other than some of the institutional and corporate world. GNU and Linux picked up the domination torch and are still working on it. I strongly suspect Microsoft was the catalyst which propelled C, absent of Unix. Microsoft had been infatuated with Unix from early on. They had their own version called Xenix, based on AT&T licensed code. Part of the MS-DOS API consists of "Xenix functions", imitating some Unix-like things. MS-DOS, and consequently Windows feature Unixy conventions like .. for the parent directory and . for current, even though the underlying link concept isn't there. Microsoft produced a C compiler + IDE which I believe highlighted and legitimized C as a viable language in the mass market world of IBM PC compatibles. According to the Wikipedia page on Microsoft Visual C++, Microsoft already had a C compiler out in 1983. "Microsoft C 1.0, based on Lattice C, was Microsoft's first C product in 1983. It was not K&R C compliant." Microsoft picking up Lattice C and running with it is probably what caused the C explosion. Borland joined in the fray in 1987 with Turbo C. But wait, Microsoft had a Lisp product too: why didn't that help? [plain HTTP!] http://www.edm2.com/index.php/Microsoft_LISP http://www.edm2.com/index.php/Microsoft_LISP Probably they just didn't peddle it hard enough. Microsoft also used C for developing Windows, and supported C development for Windows: https://en.wikipedia.org/wiki/Microsoft_Windows#Early_versions https://en.wikipedia.org/wiki/Microsoft_Windows#Early_versio... "Microsoft sold as included Windows Development libraries with the C development environment, which included numerous windows samples." The rise of MS Windows dragged C with it, with the API's and code samples being expressed in C. Plus, it was probably not lost on the PC programming population that if they learn and work with C on that platform, their skills will be applicable in Unix and elsewhere.
- zozbot234 4y ago
  The main competitor to C would have been PASCAL, and the drawbacks of standard PASCAL for low-level programming (a necessity on the computers of the day, where even ASM programming was commonly resorted to in order to eke out the best performance) were very well known. C was not perfect by any means (the segmented memory models on 16-bit x86 required non-standard extensions to support, for example) but it was a lot better than the alternative. LISP would've been a total non-starter for serious use given its requirement for GC and being an interpreted language. Quite comparable to the dog-slow Java applets of the mid-1990s, and in fact even worse on early hardware.
lbruder 4y ago
Worse is better. C and Unix were faster to implement new stuff as they were only 'good enough'. Lisp machines wanted to do everything 'right', right from the start, and were slower to implement new features. Unix was already winning by then... https://en.wikipedia.org/wiki/Worse_is_better https://en.wikipedia.org/wiki/Worse_is_better
- pjmlp 4y ago
  And now the industry is going after C Machines, as the ultimate solution to fix C memory corruption issues, the irony.
- westoncb 4y ago
  Irony is the "worse is better" guy was a lisp fanatic IIRC :)

by sbf501 4y ago

mianos 4y ago

C is much closer to the machine and very fast. When I learnt C I went to the book store. There were two C books (K&R and Byte Books C) and one Lisp book. I learnt both. Lisp was fun but, at the time, I never appreciated the introspection. It was only when, a long long time later, I learnt Python and remembered the ability to iterate the elements of a class and thought how handy that was for some programs. C could more easily access the lower level resources as well. I still wonder why one of these, like forth, lisp later Python did not become the standard command line interpreters whereas it would never have been C, and I have written very featured C interpreters.

pjmlp 4y ago
Lisp Machines, Lisp just as C, had low level primitives.

hota_mazi 4y ago

To be fair, out of these 256 pages, 200 are really just ads. Your point still stands, but the actual substance is a lot smaller than you describe.

zozbot234 4y ago
The advertisements are among the most interesting things about old magazines like BYTE, though. They preserve a snapshot of what the computer industry was like back then. Even the pricing information alone is very valuable.
- hota_mazi 4y ago
  Oh absolutely! Browsing through these pages today, I find myself reading the ads a lot more than the actual articles, which haven't aged well. But back then... man, these ads were cancer and they made up 80% of the weight and price of the magazine.
  - rbanffy 4y ago
    Those ads made the magazine possible and accessible. Of course, McGraw Hill made a profit, but BYTE would be a lot more expensive if it were not for the ads. One interesting point about the ads is that the time it took for a new technology to appear in the ads was usually much shorter than for it to appear in articles. In that sense, the ads were opportunities to peek into the near future (and, in the case of vaporware, alternate timelines). I like the way magazines like BYTE are preserved and I hope we preserve our current tech as well.
sbf501 4y ago
Counterpoint taken.

kazinator 4y ago

C came to dominate the world because of Unix. But not directly, because Unix didn't really dominate the world other than some of the institutional and corporate world. GNU and Linux picked up the domination torch and are still working on it. I strongly suspect Microsoft was the catalyst which propelled C, absent of Unix. Microsoft had been infatuated with Unix from early on. They had their own version called Xenix, based on AT&T licensed code. Part of the MS-DOS API consists of "Xenix functions", imitating some Unix-like things. MS-DOS, and consequently Windows feature Unixy conventions like .. for the parent directory and . for current, even though the underlying link concept isn't there. Microsoft produced a C compiler + IDE which I believe highlighted and legitimized C as a viable language in the mass market world of IBM PC compatibles. According to the Wikipedia page on Microsoft Visual C++, Microsoft already had a C compiler out in 1983. "Microsoft C 1.0, based on Lattice C, was Microsoft's first C product in 1983. It was not K&R C compliant." Microsoft picking up Lattice C and running with it is probably what caused the C explosion. Borland joined in the fray in 1987 with Turbo C. But wait, Microsoft had a Lisp product too: why didn't that help? [plain HTTP!] http://www.edm2.com/index.php/Microsoft_LISP http://www.edm2.com/index.php/Microsoft_LISP Probably they just didn't peddle it hard enough. Microsoft also used C for developing Windows, and supported C development for Windows: https://en.wikipedia.org/wiki/Microsoft_Windows#Early_versions https://en.wikipedia.org/wiki/Microsoft_Windows#Early_versio... "Microsoft sold as included Windows Development libraries with the C development environment, which included numerous windows samples." The rise of MS Windows dragged C with it, with the API's and code samples being expressed in C. Plus, it was probably not lost on the PC programming population that if they learn and work with C on that platform, their skills will be applicable in Unix and elsewhere.

zozbot234 4y ago
The main competitor to C would have been PASCAL, and the drawbacks of standard PASCAL for low-level programming (a necessity on the computers of the day, where even ASM programming was commonly resorted to in order to eke out the best performance) were very well known. C was not perfect by any means (the segmented memory models on 16-bit x86 required non-standard extensions to support, for example) but it was a lot better than the alternative. LISP would've been a total non-starter for serious use given its requirement for GC and being an interpreted language. Quite comparable to the dog-slow Java applets of the mid-1990s, and in fact even worse on early hardware.

lbruder 4y ago

Worse is better. C and Unix were faster to implement new stuff as they were only 'good enough'. Lisp machines wanted to do everything 'right', right from the start, and were slower to implement new features. Unix was already winning by then... https://en.wikipedia.org/wiki/Worse_is_better https://en.wikipedia.org/wiki/Worse_is_better

pjmlp 4y ago
And now the industry is going after C Machines, as the ultimate solution to fix C memory corruption issues, the irony.
westoncb 4y ago
Irony is the "worse is better" guy was a lisp fanatic IIRC :)