4 ms·

> The Commodore 64 is a fascinating machine. It is the single best selling computer model in human history. The fact that the C64 still holds that title in 2021

> The Commodore 64 is a fascinating machine. It is the single best selling computer model in human history. The fact that the C64 still holds that title in 2021 — close to three decades after being discontinued — is nothing short of amazing. I'm still somewhat amazed at how long computer models stuck around for back then (which is the reason for the c64 being the best selling computer model - it was being sold for many years). On one hand, it was probably a mindset thing that you built things to last, and the concept of iterative development wasn't really there. How long does a computer model stick around these days? I type this on an 10 year old Windows laptop, which works perfectly well for most things (battery life is pretty bad, but otherwise works perfectly well). The 10 year old iPad 2 I have is next to useless (very laggy, software updates are non existent, can't upgrade to even a modern browser).

timr 5y ago
> I'm still somewhat amazed at how long computer models stuck around for back then > On one hand, it was probably a mindset thing that you built things to last, and the concept of iterative development wasn't really there I don't think so. There were some really cheap bits in the C64 -- the power supplies were prone to burning up, and the floppy drives were so fragile that if you looked at them funny, the head would mis-align. One of the first electronics projects I remember as a kid was building a jury-rigged cooling fan for the 1541, because otherwise it would get too hot and stop working on long reads. I think it's just a function of the primitive state of the industry at that time. It's a bit like any other startup: they had to build demand for the product before they could afford to version-cycle it. But also, the manufacturing, sales, and marketing infrastructure didn't exist. Consider that literally none of the modern electronics industry existed back then -- Steve Jobs was going to Japan to convince companies to sell him disk drives! Today, if you want to get hardware made, you can pretty much dial up custom fab in China. Back then, if you wanted to build a new computer, it meant building a factory in Texas or California and sitting in conference rooms in Tokyo to convince people to sell you the latest components. It wasn't really until the first generic PCs that the rate of change began to skyrocket.
- rbanffy 5y ago
  > they had to build demand for the product before they could afford to version-cycle it. That mindset didn’t really exist back then. Instead of releasing a better C64, Commodore made the C16, the Plus 4 (both nice, twice as fast as the 64). The first backwards compatible was the C128, but it isn’t really an upgrade over the 64, but a different computer with a 64 built in - when running 64 software, it’s not even faster. In that sense, it’s a bit funny they didn’t learn the IBM 360 lesson - that a family of compatible computers can outlast the individual models essentially forever.
  - rob74 5y ago
    The problem with the C64 (and the Amiga later) was that it had games which depended on the exact timing and other details of the hardware, so it was hard to create a downward compatible improved system. Arguably Commodore also wasn't really interested in that - the Amiga certainly had a window of opportunity to survive as a platform, but when Commodore decided to invest into new models, it was too little too late...
    - rbanffy 5y ago
      > games which depended on the exact timing PC's solved that with the then-ubiquitous turbo button. The C64 had scan-line timed interrupts, and games could tie to that to work properly on faster machines, had they been available.
      - grujicd 5y ago
        Same for Amiga, games could (and most probably did) use scan line hardware interrupts to change colors, sprites, even screen modes as scan line progressed. Faster CPU or blitter wouldn't mess up anything for most games which properly used available hw resources. Of course, you can always make code dependent on CPU timing, but it was not necessary on these machines.
        rbanffy 5y ago
SavantIdiot 5y ago
I'm still using a 2013 MacBook Pro as my primary dev machine. Works as well as it did when I first got it, even switching to 100% remote work. I'm astonished. I've never said, "Gee I wish this was faster." All I had to do was replace the battery three years ago. Compare this to the PC craze from 1990 to ~2005 where you literally had to buy a new machine every 1~2 years to keep up. (My company's IT dept updated our PCs every 18 months.) Those days are long gone. Good riddance.
- incanus77 5y ago
  Same here (late 2013) until a few weeks ago when I got an M1 mini. Half because my battery started swelling (again) and I need to send it in, and half because I wanted to start experimenting on ARM Mac. I mean, it’s incredibly fast and quiet, but my MBP was really keeping up still.

timr 5y ago

> I'm still somewhat amazed at how long computer models stuck around for back then > On one hand, it was probably a mindset thing that you built things to last, and the concept of iterative development wasn't really there I don't think so. There were some really cheap bits in the C64 -- the power supplies were prone to burning up, and the floppy drives were so fragile that if you looked at them funny, the head would mis-align. One of the first electronics projects I remember as a kid was building a jury-rigged cooling fan for the 1541, because otherwise it would get too hot and stop working on long reads. I think it's just a function of the primitive state of the industry at that time. It's a bit like any other startup: they had to build demand for the product before they could afford to version-cycle it. But also, the manufacturing, sales, and marketing infrastructure didn't exist. Consider that literally none of the modern electronics industry existed back then -- Steve Jobs was going to Japan to convince companies to sell him disk drives! Today, if you want to get hardware made, you can pretty much dial up custom fab in China. Back then, if you wanted to build a new computer, it meant building a factory in Texas or California and sitting in conference rooms in Tokyo to convince people to sell you the latest components. It wasn't really until the first generic PCs that the rate of change began to skyrocket.

rbanffy 5y ago
> they had to build demand for the product before they could afford to version-cycle it. That mindset didn’t really exist back then. Instead of releasing a better C64, Commodore made the C16, the Plus 4 (both nice, twice as fast as the 64). The first backwards compatible was the C128, but it isn’t really an upgrade over the 64, but a different computer with a 64 built in - when running 64 software, it’s not even faster. In that sense, it’s a bit funny they didn’t learn the IBM 360 lesson - that a family of compatible computers can outlast the individual models essentially forever.
- rob74 5y ago
  The problem with the C64 (and the Amiga later) was that it had games which depended on the exact timing and other details of the hardware, so it was hard to create a downward compatible improved system. Arguably Commodore also wasn't really interested in that - the Amiga certainly had a window of opportunity to survive as a platform, but when Commodore decided to invest into new models, it was too little too late...
  - rbanffy 5y ago
    > games which depended on the exact timing PC's solved that with the then-ubiquitous turbo button. The C64 had scan-line timed interrupts, and games could tie to that to work properly on faster machines, had they been available.
    - grujicd 5y ago
      Same for Amiga, games could (and most probably did) use scan line hardware interrupts to change colors, sprites, even screen modes as scan line progressed. Faster CPU or blitter wouldn't mess up anything for most games which properly used available hw resources. Of course, you can always make code dependent on CPU timing, but it was not necessary on these machines.
      - rbanffy 5y ago

SavantIdiot 5y ago

I'm still using a 2013 MacBook Pro as my primary dev machine. Works as well as it did when I first got it, even switching to 100% remote work. I'm astonished. I've never said, "Gee I wish this was faster." All I had to do was replace the battery three years ago. Compare this to the PC craze from 1990 to ~2005 where you literally had to buy a new machine every 1~2 years to keep up. (My company's IT dept updated our PCs every 18 months.) Those days are long gone. Good riddance.

incanus77 5y ago
Same here (late 2013) until a few weeks ago when I got an M1 mini. Half because my battery started swelling (again) and I need to send it in, and half because I wanted to start experimenting on ARM Mac. I mean, it’s incredibly fast and quiet, but my MBP was really keeping up still.