4 ms·

If I recall correctly, Steven Levy's book "Hackers: Heroes of the Computer Revolution" also covered some of this story. Woz was intimidated by disk controllers

by Malic 8y ago

If I recall correctly, Steven Levy's book "Hackers: Heroes of the Computer Revolution" also covered some of this story. Woz was intimidated by disk controllers because he didn't know much about them, other than they were big complicated things. So starting with his own ignorance as a fresh slate, he ended up with something better than his former employer HP had in their products. Sometimes (Sometimes! Not always!) not knowing better can be helpful.

czechdeveloper 8y ago
Being genius helps with that.
- digi_owl 8y ago
  And limitations. Woz did his thing under much the same constraints as Tramiel held the Commodore engineers to, keep the chip count, and thus the price, down. Every time Woz added a chip, he carefully looked for very way he could use that chip for different tasks.
- archagon 8y ago
  I feel like "genius" is almost always an ex post facto label that's doled out to make us feel better about our own lack of ambition. That person did something really wild and interesting? Clearly they're a genius, so why even bother striving for anything quite so good? Same with the recent Steve Jobs threads. Or any thread featuring prominent programmers or founders. I think immediately reaching for the "genius defense" (whether deserved or not) does far more harm than good.
  - newnewpdro 8y ago
    You're right. Exceptional labels are often used to abdicate responsibility of the mediocre for their own mediocrity. It's easier to label someone as special than it is to admit you didn't put in the time they did, or that your fear of failure prevents you from trying. This particular Woz story doesn't scream of "genius" to me. It screams of curiosity and confidence, a fearlessness, of a young and foolhardy geek swimming in opportunity and it enabled him to discover some interesting, innovative things. The part where his (probably senior) hardware dept. colleague trolled him into completely ripping apart his work and redoing it all for cosmetic reasons with a single sentence saying "Steve, you mean you’re going to go with less than perfect?" had me laughing heartily, and caused me to reflect on the times similar things had been done to me early in my career.
- ec109685 8y ago
  He had spent a good chunk of his life taking designs and simplifying them and was a genius in that aspect, seeing things that other people couldn’t do when it came to circuit design.
pvg 8y ago
I think Wozniak's own modesty obfuscates the story a little. Yes, he didn't know the ins and outs of contemporary disk controllers but he was an experienced designer and programmer. One of the reasons he could ditch a lot of the complexity and expense of the Shugart (not HP) controller he was looking at is because he could leverage the flexibility and extensibility of the host, which he also designed.
- hinkley 8y ago
  He mentions that toward the end.
  - pvg 8y ago
    Definitely, it pops up throughout but a casual read (and popular retellings of this story) might leave you with the impression that he almost accidentally stumbled his way, savant-like, into a superior and cheaper design. That unintentionally sells the work a little short. Not directly related but fun to think about: the word 'interrupt' does not appear at all.
    - apricot 8y ago
      Wozniak apparently didn't like interrupts that much. They are almost unused on the Apple II -- for instance, the keyboard doesn't use them --, and grokking interrupts was the single hardest thing for me when I migrated from the Apple II to the IBM PC.
      - pvg 8y ago
        https://www.youtube.com/watch?v=xECUrlnXCqk https://www.youtube.com/watch?v=xECUrlnXCqk
AlphaWeaver 8y ago
Steven Levy's book is a great read, it covers fastening topics from the very early history of computing forward closer to modern day. Really really interesting to learn.
ajross 8y ago
> he ended up with something better than his former employer HP had in their products. That's actually severely understating the accomplishment. The Disk ][ controller was implemented with like 8 chips, all of them off the shelf TTL things you could get at Radio Shack (OK, two were small ROMs). It talked directly to the drive which needed almost no electronics on it at all. And it was both faster and higher density than all the custom ASICs being sold at the time. It could read bits off the drive at full rotation speed, filling all of RAM within seconds. For comparison, the Atari 810 drive was designed contemporaneously (though it arrived later to market). It paired a 6507 (6502 with some address lines clipped off), 4k of RAM and a WDC controller chip inside the drive, which then spoke to the main computer over a semicustom UART (or UART-like) chip on the motherboard. And with all that hardware, it managed to deliver the data at 19.2kbps. This thing was an absolute masterpiece of the era.
- Zardoz84 8y ago
  We can say that there was 3 kinds of floppy drives : - Drive by a floppy controller chip : Controlled by a WDC, Intel or custom ASIC. For example Amstrad, Sinclair +3 or Amiga floppy drives. - Full intelligent : Like Commodore or Atari 8 bit floppy, that carries a computer inside. - Dumb : Like Apple II and I think that IBM PCs. Full intelligent not was a bad idea when the home computers had these low speeds, as could (on theory), read/write the floppy and at same time do other stuff on the computer. Also, allowed to move the file system implementation and part of the I/O firmware to these embed computer to the floppy drive. But to work nice, need a fast and good connection to the computer. Having a serial communication via bit-bagging as did Commodore, killed it.
  - ajross 8y ago
    > Full intelligent not was a bad idea when the home computers had these low speeds That's the excuse, but no. In practice the actual software run on those machines was synchronous. You'd load the game/document, work/play, then save. There wasn't enough space in a 48k world for software complex enough to have background tasks. Honestly the only example I can think of where anyone did anything like this was the animated splash screens during load that some of the later Lucasfilm games on Atari had. And they never would have bothered if the disk was as fast as Apple's. Also, FWIW: the PC had a "dumb" drive enclosure, but the controller was an ASIC on the motherboard. And it was drivable via a separate DMA chip, allowing true background operation. With 10x the memory of the 8 bit systems, it was not uncommon to find DOS tricks to do background I/O (e.g. RAMdisks were very popular).

by Malic 8y ago

czechdeveloper 8y ago

Being genius helps with that.

digi_owl 8y ago
And limitations. Woz did his thing under much the same constraints as Tramiel held the Commodore engineers to, keep the chip count, and thus the price, down. Every time Woz added a chip, he carefully looked for very way he could use that chip for different tasks.
archagon 8y ago
I feel like "genius" is almost always an ex post facto label that's doled out to make us feel better about our own lack of ambition. That person did something really wild and interesting? Clearly they're a genius, so why even bother striving for anything quite so good? Same with the recent Steve Jobs threads. Or any thread featuring prominent programmers or founders. I think immediately reaching for the "genius defense" (whether deserved or not) does far more harm than good.
- newnewpdro 8y ago
  You're right. Exceptional labels are often used to abdicate responsibility of the mediocre for their own mediocrity. It's easier to label someone as special than it is to admit you didn't put in the time they did, or that your fear of failure prevents you from trying. This particular Woz story doesn't scream of "genius" to me. It screams of curiosity and confidence, a fearlessness, of a young and foolhardy geek swimming in opportunity and it enabled him to discover some interesting, innovative things. The part where his (probably senior) hardware dept. colleague trolled him into completely ripping apart his work and redoing it all for cosmetic reasons with a single sentence saying "Steve, you mean you’re going to go with less than perfect?" had me laughing heartily, and caused me to reflect on the times similar things had been done to me early in my career.
ec109685 8y ago
He had spent a good chunk of his life taking designs and simplifying them and was a genius in that aspect, seeing things that other people couldn’t do when it came to circuit design.

pvg 8y ago

I think Wozniak's own modesty obfuscates the story a little. Yes, he didn't know the ins and outs of contemporary disk controllers but he was an experienced designer and programmer. One of the reasons he could ditch a lot of the complexity and expense of the Shugart (not HP) controller he was looking at is because he could leverage the flexibility and extensibility of the host, which he also designed.

hinkley 8y ago
He mentions that toward the end.
- pvg 8y ago
  Definitely, it pops up throughout but a casual read (and popular retellings of this story) might leave you with the impression that he almost accidentally stumbled his way, savant-like, into a superior and cheaper design. That unintentionally sells the work a little short. Not directly related but fun to think about: the word 'interrupt' does not appear at all.
  - apricot 8y ago
    Wozniak apparently didn't like interrupts that much. They are almost unused on the Apple II -- for instance, the keyboard doesn't use them --, and grokking interrupts was the single hardest thing for me when I migrated from the Apple II to the IBM PC.
    - pvg 8y ago
      https://www.youtube.com/watch?v=xECUrlnXCqk https://www.youtube.com/watch?v=xECUrlnXCqk

AlphaWeaver 8y ago

Steven Levy's book is a great read, it covers fastening topics from the very early history of computing forward closer to modern day. Really really interesting to learn.

ajross 8y ago

> he ended up with something better than his former employer HP had in their products. That's actually severely understating the accomplishment. The Disk ][ controller was implemented with like 8 chips, all of them off the shelf TTL things you could get at Radio Shack (OK, two were small ROMs). It talked directly to the drive which needed almost no electronics on it at all. And it was both faster and higher density than all the custom ASICs being sold at the time. It could read bits off the drive at full rotation speed, filling all of RAM within seconds. For comparison, the Atari 810 drive was designed contemporaneously (though it arrived later to market). It paired a 6507 (6502 with some address lines clipped off), 4k of RAM and a WDC controller chip inside the drive, which then spoke to the main computer over a semicustom UART (or UART-like) chip on the motherboard. And with all that hardware, it managed to deliver the data at 19.2kbps. This thing was an absolute masterpiece of the era.

Zardoz84 8y ago
We can say that there was 3 kinds of floppy drives : - Drive by a floppy controller chip : Controlled by a WDC, Intel or custom ASIC. For example Amstrad, Sinclair +3 or Amiga floppy drives. - Full intelligent : Like Commodore or Atari 8 bit floppy, that carries a computer inside. - Dumb : Like Apple II and I think that IBM PCs. Full intelligent not was a bad idea when the home computers had these low speeds, as could (on theory), read/write the floppy and at same time do other stuff on the computer. Also, allowed to move the file system implementation and part of the I/O firmware to these embed computer to the floppy drive. But to work nice, need a fast and good connection to the computer. Having a serial communication via bit-bagging as did Commodore, killed it.
- ajross 8y ago
  > Full intelligent not was a bad idea when the home computers had these low speeds That's the excuse, but no. In practice the actual software run on those machines was synchronous. You'd load the game/document, work/play, then save. There wasn't enough space in a 48k world for software complex enough to have background tasks. Honestly the only example I can think of where anyone did anything like this was the animated splash screens during load that some of the later Lucasfilm games on Atari had. And they never would have bothered if the disk was as fast as Apple's. Also, FWIW: the PC had a "dumb" drive enclosure, but the controller was an ASIC on the motherboard. And it was drivable via a separate DMA chip, allowing true background operation. With 10x the memory of the 8 bit systems, it was not uncommon to find DOS tricks to do background I/O (e.g. RAMdisks were very popular).