Jack Tramiel Got A Good Deal, And Ruined Everything

A sideshow in the playground wars of the early 1980s over who had the best home computer lay in the quality of their onboard BASIC interpreters. Where this is being written the cream of the crop was Acorn’s BBC Basic, while Sinclair owners could hold their own, and the Commodore 64 was regarded as powerful, but not easy to program. It’s a teenage memory brought to mind by [Liam Proven], who argues in a blog post that Commodore’s BASIC left a problematic legacy that can still be felt today.

It’s an interesting proposition, and one with its roots in Commodore founder Jack Tramiel’s 1977 deal with Bill Gates to acquire a version of Microsoft BASIC for his machines, in which he paid a one-off fee for unlimited uses of the language rather than a per-sale levy. The argument in the post is that this led to later Commodore machines being hamstrung by an outdated BASIC interpreter as a cost saving measure. It fits well with those 1980s memories from school computer labs, because by comparison its competitors six years after the deal had the benefit of language extensions missing in Commodore’s 64.

Where [Liam]’s analysis becomes interesting is in how he perceives the effect of this long-in-the-tooth BASIC; he postulates that the sheer number of Commodore 8-bit machines sold ensured it had a dominant position in the market place and thus coloured the perception of BASIC as a programming language in the years that followed. We’re not so sure about his view that this led eventually to some of the shortcomings in computing today, but we agree wholeheartedly with him that Commodore were less than competent in marketing their hardware.

We look forward to hearing your take on the matter in the comments, and meanwhile for some perspectives on the Commodore of the day who better to relate them than somebody who had a ringside seat. Our colleague [Bil Herd] has shared with us some of his Commodore recollections over the years, including the Commodore 128 story, an account of the 1985 CES show, and a two-parter on the TED chip and its speech capabilities.

Header image: Commodore BASIC / Public Domain, and Evan-Amos / Public domain.

89 thoughts on “Jack Tramiel Got A Good Deal, And Ruined Everything

  1. Well, I remember the C64 BASIC being “good enough” for the things to do on this machine; more considerable stuff, like the Mandelbrot set display, went anyway a more efficient way of programming using all the PEEKs and POKEs the Nerd differed from the usual school kid. And there was Simons’ Basic of course, that also made life easier – and an EPROM programmer “someone” also had at hand. So, definitely no howling about an outdated BASIC.

    1. You’ve touched on the nub of the problem, everything required PEEK and POKE.

      At the time I had Sinclair, used Acorn, Commodore, and Tandy(Radio Shack). Without re-enacting playground arguments of yesteryear, the later machines didn’t need all those PEEKs and POKEs. The Acorn machines in particular took feature complete to a new level.

      1. The guy defeats his own argument and pretends he didn’t…

        Everything required PEEK and POKE – yes, but the 8 bit machines didn’t have the resources for higher level languages anyhow, so to do anything really useful involving graphics and sound, you had to drop down to assembly anyhow.

        It was the same thing on the PC later, where you did have better BASIC and other languages, yet to actually get it to do anything interesting you had to start peeking and poking into registers, and running inline assembly to bypass the standard slow library routines.

        1. 8 bit didn’t have the resources for higher level languages? Suggest you look at the range offered by Acornsoft on the Beeb on the 6502 and the range available on CP/M.

          Anyway, in my experience those who didn’t use the C64 for playing games either got bored or ended up using 6502 assembler. Apart from the lucky BBC owners we all picked up bad habits from Basic largely because it was self taught or passed on in conversations between friends and magazine articles.

          1. The BBC had some stuff like the BCPL bytecode interpreter. CP/M had some fairly crude C compilers and most had some kind of small pascal. But nobody used them to write anything serious because they were too slow, and their output was too bloated and slow.
            Once you got to 68000 at 8MHz and to PC the size bloat was less of a problem and you’d mix asm and C (our game engine was in C our graphics engine in asm for example).

    2. Why should they care about the BASIC? It was enough to do your homework, it was enough to write simple apps. The C64 was fundamentally a games console disguised as a home computer and Commodore new exactly what they were selling and what mattered. It was enough of a home computer that the kids (and Dad) could get one for ‘learning’ and to convince people it wasn’t just another Atari VCS but most of them wanted it for games really.

      There is also a second consideration people are ignoring: compatibility. Commodore couldn’t change the C64 because the fact it was a C64 was more important than anything else. They did the C64 redesign to get the chip count down, and they did the C128 which was somewhat successful and C64 compatible. Every attempt they made to build a non C64 compatible 8bit system tanked. Nobody wanted a C16, a plus 4 or a CBM-II system because they were not C64 compatible.

      Some of the comparisons I’d also question. The BBC had a superb BASIC, but the OS and BASIC took 32K of ROM so by the time you’d actually put the computer into a graphics mode you didn’t have room to write anything useful. It was also incredibly expensive.

      Sinclair basic had some support for graphics and stuff but it was minimal and slow, so you could plot graphs and do minor demos and make a beep. BASIC supported the Spectrum hardware “well” because there wasn’t anything to support but a 1 bit beeper and a simple unaccelerated frame buffer.

      MSX maybe the counter-example. MSX did go to some trouble to expose real hardware in BASIC usefully. but almost everyone who bought an MSX still bought it for the games 8)

      BASIC convinced your parents to buy you a “computer”, Games convinced most people to actually ask for one.

      Compatibility determined games volume, games volume determined sales. Game industry priorities were simply platform by volume, plus new platforms where the manufacturer was bankrolling gmes to get going (Amstrad CPC for example). Even then you usually wrote for the top seller (Spectrum, C64) and did a hack port to the rest.


      1. “Compatibility determined games volume, games volume determined sales.”

        Just as gamers are still the ones pushing the envelope for faster processors, larger RAM, better graphics, better cooling.

      2. I had a Chemistry professor who decided to give everything to 10 significant digits so no one could use (8 digit) calculators for their homework. This pissed me off because it had nothing to do with learning Chemistry, which is what I was paying the university to teach me. It took about 2 hours to write a program on the C-64 that would multiply two 128 digit numbers –and show the work for numbers <20 digits long.

      3. “The BBC had a superb BASIC, but the OS and BASIC took 32K of ROM so by the time you’d actually put the computer into a graphics mode you didn’t have room to write anything useful.”

        The amount of ROM taken is irrelevant. And while it had 32K of RAM (only 16K on the short-lived Model A!), the graphics bitmap was the main consumer (smallest being mode 7 at 1k, largest being mode 0 at 640x256x1bpp = 20k).

        1. It was highly relevant. It had a flat 64K address space. If 32K of it was wired to ROM then that left 32K for RAM without a lot of extra paging logic. Had the BASIC and MOS fitted in 16K the sideways space would have trivially let everyone have 48K. As it was it took until Watford Electronics figured out how to do two banks of RAM (The WE32) and hide the video, and Acorn to copy them in the BBC B+ and the 128 before that mess got sorted. In the games world the BBC was a real problem child because of the lack of RAM.

          We had 40K usable on a spectrum, similar on a CPC, even more on a C64. Lots of stuff just didn’t fit or had to run in cut down form. (and for those outside the UK – only posh kids had a floppy drive and sufficiently close to nobody had an Apple II so UK game ports never bothered with Apple II). By the time the BBC B+ and 128 appeared it was too late, except for Acornsoft compatibility and volume said BBC B or don’t bother.

          From a gaming house point of view the BBC basic was irrelevant, the lack of memory was a killer. Of course the BBC was aimed at education so the BASIC actually was a good thing there.

          1. Sorry, I missed that you were specifically concerned with games. However, in the BBC’s favour, I’ll wheel out a couple of superstars – Elite (obviously), Revs (which indirectly led to F1GP etc on the PC) and Exile (a huge game, especially given the memory limitations). Uncoincidentally, all of these pulled mid-screen mode-change tricks to save memory, and Revs even used the screen memory as data memory, just setting the physical colour for all the logical colours to the same (iirc, it’s been a looooooong time…).

            But I guess that could just go to reinforce your point that it was either too hard or not worth the hassle to port things across – I certainly don’t recall any BBC port of a cross-platform game being better than the Spectrum one (excepting Elite).

          2. Commodore +4 offered 60K free RAM for BASIC due to RAM/ROM switching, despite it had no MMU. But it had pretty smart memory management (on the other hand, it made the machine slower, but because of that the CPU run at the border at 1.76MHz to slightly compansate this).
            All together, with more advanced programming tools, gfx and sound commands and some nice features (TRAP, Renumber and so) +4 could still offer 60K free RAM for BASIC programmers.
            And it had a monitor for ppl wanted to start some machine code programming… It was a nice machine, it was punched to death because of the initial (really irrealistic) price tag. At $150 starting price (half than the C64) it could be a killer, even 3+1 could be left out.
            If anyone think BASIC is not good enough for a game, worth to check Magic Blocks on plus4world, it is fully made in BASIC, and not compiled. Fast, funny and nice (certainly I could not do the same, so it is a masterpiece).

            C128’s BASIC is even more advanced, supports loading picture or from/to memory addresses, can play proper music, and have full sprite support. As it has MMU, it could support RAM BANKs.

            (certainly there were other nice home computers with nice features, like Enterprise 128 and the Spectrum +3 was OK too (too bad they did not develop a more advanced GFX mode, like multicolor on Commodore, they already had the memory for that in +3 (128K)) )

      4. IBM failed to make a dent in the home market with OS/2 Warp because there were no games, and IBM never had any graphics performance enhancement like DirectX. Microsoft started their own games division specifically to improve sales to the home market.

      5. The limited memory of the BBC Micro, after allocating 32K to ROM and up to 20K for graphics, was certainly an issue. But ‘sideways RAM’ addons (paged into the address space used by the language ROM) soon became available, and this was implemented as standard on the later BBC Master models.

    3. Commodore BASIC V2 on the Commodore 64 was the same old Commodore BASIC ported from the Commodore PET/2000/4000 computers which had no commands for colour, graphics, or sound! Commodore, as well as shops and magazines usually gave buyers no warning about this. It certainly came as a very nasty surprise to me after a few months or several months studying the official Commodore “An Introduction to BASIC” course, as well as the Programmers’ Reference Guide!! Simon’s BASIC, as well as other extended BASICs didn’t have the ability to produce stand alone programs that would run on another C64 without the relevant Copyright BASIC being loaded. Even Simon’s BASIC didn’t have many third party books or magayine artcles written about it, so whenever you read program listings they were almost always in Commodore BASIC V2, peppered with PEEK and POKE commands. It was like trying to remember the contents of a phone book with all the mainly 5 digit memory locations. According to the number of commands, I think Amstrad CPC Locomotive BASIC and MSX BASIC were the best, although they didn’t have procedures like BBC BASIC and Sinclair QL SuperBASIC. Machine Code wasn’t for beginners. You can read more about it on my blog https://commodore64crap.wordpress.com/

  2. The problem with the 6502 is that it cannot be sped up much, was 40 columns and didn’t have a good “Print Using” command that other manufacturers like Apple had.

    The Commodore 64 was limited to 38K and programs with large arrays would have an out of memory error.

    Jack was making computers that cost less and dumbed down on their capabilities by making it as cheap as he could.

    The C-16 had less memory and an improved BASIC and 3rd party sellers were price gouging peripherals for the Amiga. That’s Commodore’s legacy.

    Outsourcing your operating system of BASIC to another company just shows they weren’t willing to innovate or reinvent themselves.

    They basically fired MOS because their leaders saw the handwriting on the wall and left. Why couldn’t they have made 80 columns, different chips, more products? If I answer that question, its because the company killed it and didn’t want to be a powerhouse. If you want to overcharge your customer, you take all the money and don’t grow. Some of them should have been fired but you can’t fire the boss.

    Apple was in the schools.

    All I see is disaster. You’re beating a dead horse.

    1. I remember reading about the Sid chip team. They worked on the chip for about a month and that was it. That means they left.

      The point is they made something because they had ability and left because Commodore wasn’t willing to pay them beyond that.

      If you don’t have more products in the pipeline and if you are gouging for what you are selling then its a small minded company charging a premium. Some things are expensive but if you don’t take the risk, you don’t make anything. Who are you paying that was left over and what did they make? Exactly.

    2. To be fair, there was an 80 column video card available and you really could use all of the 64K if you could work with the weird bank switching.

      Any actual programming on this platform was done in pure assembly language, so the quality of the basic interpreter is irrelevant from that standpoint.

      Golly it’s SO much fun to play monday morning quarterback, esp. when you are just guessing about all of it! Please do go start your own company and ship a million units and then get back to us.

      1. Most televisions could not display 80 columns. I typed in programs that offered it and it was disappointing. Anything using the cartridge port required the user to lose memory severely limiting what you could do with it.

        The Commodore 64 did not have a built in Machine Language editor or monitor.

        Commodore was an economic cult and they told everyone their customers were loyal when there was no loyalty.

        If you want upgrades, you have to pay people for them and Commodore did not want to spend the money.

        1. “Commodore was an economic cult and they told everyone their customers were loyal when there was no loyalty.”

          I hate to be the one to burst your bubble but this is the standard operating model for the US corporation and it has been for hundreds of years. You can read any number of history books about US industry and this is always the case, you can be talking about cars or canned meat or furniture or prescription drugs, so maybe you can cut the fake indignation.

          1. The first generation of Amiga users would not use a BBS run on IBM BBS software because they hated IBM. That does not exist today with Amiga users.

            Amigaworld got rid of the Andy Warhol style art which launched the magazine and went all business.

            Commodore was a cult because even when Commodore was in bankruptcy, Amiga owners wouldn’t believe the company would go under and when Commodore went bankrupt, we were in amazement of the ones who wouldn’t believe it. The Commodore computer mags were selling the Amiga and their motto was that Commodore users were more loyal and were more loyal to buy Commodore products again which means they were in on the bias to sell more magazines. When you have the computer wars, it is more about their computers becoming religion.

            It is disgusting when you go to an Amiga club, watch demonstrations and the operator of the computer goes through every font to show everyone every font and you are stepping on peanut shells because no one sweeps the floor.

            The Amiga scene became full of cult personalities identifying with their machines and some of the engineers became outspoken on the Amiga and to my knowledge have never done anything for the Amiga since Commodore went bankrupt while their followers are still faithful and waiting for the day that some company will take interest and bring their computers back to them.

            The Amiga has become entertainment for some but I have to cut my comments short but one of the models which sold tech became entertainment.

        2. You want to criticize the C64? Go ahead. But I have a challenge for you ;)

          The design of the C64 arguably took a year, but almost all of that (Jan through Nov, 1981) was spent developing the VIC-II and SID chips. The rest of the computer was apparently laid out and made to work in roughly one single month — it had to be ready for CES 1982, per Tramiel’s orders, and that was the first weekend in Jan of that year — and per timeanddate[dot]com, the year started on a Friday! (The rest of that info is from Wikipedia).

          So here’s your challenge. Lay out the circuitry for a period system, using period chips that Commodore would’ve had in 1982. You can use the VIC-II and SID if you want, but no other custom chips. Extra credit would include writing the ROM image… I really don’t expect you to actually build & test, but it should be fairly realistic. 6502[dot]org has a lot of resources that will help with that ;)

          Since you’re not doing the software side (theoretically) — you have two weeks. Good luck ;) I can lay out circuitry for something about the sophistication of a KIM-1 in an afternoon or three, but one-upping the C64, even in circuitry alone, is a task that is well beyond me, and I know it. There are enthusiasts even now, today, in 2020, who are finding new things to do and new ways to do them with that computer.

          1. @Alphatek — the BBC Micro was based on a system that Acorn had already prototyped, and I’d in no way consider it superior to the C64, having just now looked at the specs. Sure, it was government-sanctioned (but that can go both ways, look at what us idiots in the USA are getting for our troubles at the voting booth last go-round!) and it made quite an impression on many young blokes over there — but it was by no means a superior machine.

            Nice try tho!

            Also, I was mostly targeting Chuckz, whom it appears has decided to ignore me. I’m not surprised, I highly doubted they’d take up the challenge even *before* I wrote it — the point of course being made by exactly that sort of action. He *can’t* do better, he doesn’t have the knowledge… which also doesn’t surprise me. Those who’ve used a C64 (I spent my last semester in college, mind you this was 2008-9, with one I’d dug out of a cabinet in the science building, dead serious… it was a C64c, mind you, but it’s not like that doesn’t count!) understand just how flexible, capable, and powerful a system it is, versus the other machines of the day.

            It’s not for nothing that the Guiness Book has a line in it for that machine, as the most popular single model of computer sold ever. It’s the folks who *only* read the Wiki page and sneer because to them “old” starts at WinXP (ha! I have a Compaq III sitting behind me that I’m in the process of restoring… 12MHz 286, MS-DOS 3.3… surprisingly ginormous hard drive, it’s 42MB — no, not a typo, forty-two MEGABYTES whoooooo — but back in that era, that was “YOOOOOGE!” LOL. Also: vat eez diss “mouse” yoo speek ov…? ;) ) and they can’t imagine how you could do anything, even at an “Arduino” level, with a CPU — not MCU, so no on-chip anything extra, LOL — that only sort of hit ONE MEGAHERTZ.


            The C64 was an *amazing* machine. Don’t knock it if you’ve not tried it — which you kind of owe it to yourself to do, for several good reasons. One-time AOL user? AOL got its start as “Quantum Leap”, as a BBS-type network for Commodore systems, primarily the C64 and C128. (True story: I was on AOL in the mid-1990s!) Ever used a thing called GeoWorks, an OS in the Win3.x days (and very early Win95 days… before MS got a bit big for its pants and wiped ’em out) designed for computers that weren’t good enough for Windows…? Yeah, well GeoWorks started its life as GEOS, the Graphical Environment Operating System for C64 (and later C128 and supposedly Plus/4 although I don’t know why they bothered, that one *was* cr#p) computers… ;) I could go on for quite a while… a lot of modern stuff got its start in that era, around users of those systems… for good reasons.

            Actually, while I’m here — hey, shoutout to the Apple II. While I *highly* disagree with the modern approach Apple takes to its products, by and large (“hey, we’re gonna make nearly-impossible-to-service machines and tell you to just throw them out when they die, so we can make profits obscene enough that Bernie Madoff would turn redder than a summer tomato, and we’re gonna charge you what we feel like, regardless of what it actually costs to build, because we know you’ll pay it and we don’t freaking care if you have to mortgage your home to do so” etc) — although, their interfaces are well-thought out, and the charger hardware is thorough enough that I can’t *not* respect it — the really old pre-“Classic Mac”-era hardware? Deep respect. I’ve used it. It doesn’t outdo a C64, mind you, but it’s not half bad for what it is, and The Woz put a *ton* of thought and brilliance both into those machines. Good, honest engineering is a thing that I can appreciate, and *those* machines have it. (OTOH, something like an Intel iMac, where the experts (eg Louis Rossman, iFixit, etc) generally say, you’re gonna bust the first dozen screens you try to put in there, just plan on it, and they’re half the value of the entire freaking system… uhm yeah we’re gonna have to talk about that!)

    3. >Outsourcing your operating system of BASIC to another company just shows they weren’t willing to innovate or reinvent themselves.
      ⌈ ⌊ + ÷ × − ∣ ∼ ⊖ ⋆ ⌹ ⌽ ⍉ ⍋ ⍎ ⍒ ⍕ ⍟ ⍳ ⍴ ○ ¨ × ÷ + ⊢ ⊣ ⊤ ⊥ ⋆ ○ ⌊ ⌽ ⍟ ↑ ↓ ∈ − ∣ ∧ ∨ ≠ ≤ ≥ ⊖ ⌈ ⌹ ⍉ ⍕ ⍱ ⍲ ⍳ ⍴ ⌿ . ⍀

      There is the Big Blue Elephant in the room.
      IBM made the PC using off-the-shelf IC even if they had a IC foundry.
      They outsourced the operating system even if they had fine OS and developers.
      Instead to use their own preferred language APL they opted for Basic and didn’t put the graphics chars to write APL.
      They had also a BASIC interpreter but they outsourced it from Microsoft.

      The IBM PC was quite succesful and was strongly supported by IBM.

      The cheap guy in computer area was Sir Clive Sinclair that designed their PC with the goal to save the last pence.

      1. IBM PC was a “stealhworks” project inside IBM to avoid scrutiny by the higher-ups and it would have been immediately aborted if they had used internal IBM resources, so they had to farm out everything. They set up the lab in Boca Raton specifically to avoid the management types in Poughkeepsie and Armonk.

        IBM PC was most certainly not “strongly supported” by iBM, they had an extensive and interoperable product line and the IBM PC did not fit into it at all. IBM PC support for IBM LU6.2 networking was always terrible, a disaster compared to their mainframe offerings.

        1. It was in fact the general frustration with IBM’s lack of support and innovation of the IBM PC which led the Compaq founders to start their competing company with their famous restaurant napkin business plan.

    4. As far as I know almost everybody outsourced Basic to Microsoft. It’s why IBM went to talk with them when they wanted to compete in the space. I think it’s just that Commodore didn’t buy any support/customization so they had no APIs.

    5. Hi. Article author here.

      None of what you’re saying is accurate.

      > The problem with the 6502 is that it cannot be sped up much

      Sure it can. Later models ran at 4MHz, 8MHz, 10MHz and even 14MHz. There was also the 65C816 (as used in the Apple ][gs and Acorn Communicator) which was 16-bit — the CMD SuperCPU ran a C64 at up to 20MHz: https://en.wikipedia.org/wiki/SuperCPU

      > was 40 columns

      Nothing to do with the CPU or MOStek at all. Several early-generation 6502 machines had 80-columns, including Commodore’s own PET 8000 series. Your comments about abandoning MOS because it didn’t improve the chip show a total failure to understand which companies and which components do what.

      >The Commodore 64 was limited to 38K

      Only in BASIC.

      > Jack was making computers that cost less and dumbed down on their capabilities by making it as cheap as he could.

      Not at all. There were far cheaper, more dumbed-down machines around the C64’s time. It was quite a premium machine in 1982.

      What’s the relevance of the C16? CBM _could_ have made it more C64-compatible, as I said in my article. Various early-80s 8-bit machines had some kind of startup menu or setting to allow enhanced modes or slower backwards-compatible modes for old software. My point in the blog post was that CBM didn’t realise why people bought their machines, which is why it failed to capitalise on it with better models, and brought out incompatible duds instead.

      As for using MS BASIC — almost _everyone_ used MS BASIC. That is no bad thing. What is bad is that CBM already had access to later, better versions but decided to save money on a $595 computer by fitting a measly little 8KB ROM.

  3. I didn’t get into the Commodore game until the Amiga. By then, they had long left BASIC behind and C was the language of choice. But, as was stated above, they never understood how to market good hardware. What they failed to realize is that it was (as gebhardm said) those nerdy peeks and pokes that set folks apart from everyone else. With the Amiga, it just worked. But the industry didn’t have terms like “Multimedia” yet, and the general public… having never seen it before… really didn’t know what they were looking at (James Burke: the mind cannot see something it has never seen before). Add to that another twist in which the contemporary automobile was becoming very complex, but with no 3rd party support. So, the shade tree mechanic was becoming extinct. Those folks needed something to tinker with… and along comes Microsoft who understood that you can’t simply build something awesome, you have to bring along your user base as well. They had to increment it in steps that people could understand and that it is okay to charge them for it. So, in the end, eventually MS was able to build a user base and a whole industry while Commodore never figured out what went wrong. Eventually they sold their most valuable patent, the two button mouse, and that was the end.

  4. Weirdly I never was able to program in BASIC. I think I ended up starting on languages that weren’t pre-compiled (PHP and Python) then quickly started working on Tcl/Tk and C.

    Ironically, given it’s intent, BASIC gave a young me the impression that I couldn’t wrap my head around coding. Turns out I just can’t follow BASIC well for some reason.

  5. Well the Commodore Basic was at that time for the majority the entrance into programming. Those who wanted to step deeper into tickling the limits of machines went into coding with assembler which is still for those machines state of the art !

    Even for the Commodore Amiga (and keep in mind with the Vampire Stand Alone, its still alive and kicking) Assembler is the non plus ultra machine language when it comes to code a software that is getting the maximum out of its hardware.

  6. From the link:
    “Which is what has now landed us with an industry centred around huge, unmaintainable, insecure OSes composed of tens of millions of lines of unsafe C (& C derivatives), daily and weekly mandatory updates in the order of hundreds of megabytes, and a thriving industry centred around keeping obsolete versions of these vast monolithic OSes (which nobody fully understands any more) maintained and patched for 5, 10, even 15 or so years after release.”

    So, because CBM used a cheap version of BASIC, all of today’s C programs, and C based OSs are crap?
    That is a leap of logic I have trouble following…

    In my fledgling experience (at the time) I saw the keyboard of the TS-1000 a bigger handicap for it than the BASIC that I used in the VIC-20, nor did I see/feel much difference with the BASIC used in TRS-DOS machines.
    Yeah, I saw everything as a nail, because my only tool was a hammer.

    1. The whole diatribe is poorly founded. If anything, those who were around at time are all better programmers for having only that level of tool, be it Commodore, Acorn, MSX, whatever. If you wanted anything doing, then you _had_ to do it yourself. No internet, no downloadable libraries, no SDKs…

      1. There were books! Big fat hold the door open sized books like the PRG. If you had a C64, you had to have the PRG! Which, amazingly, could be found in actual stores. Ok, maybe University bookstores, but hey.

    2. I think I am missing the point either. Sure we were not going to write operating systems in basic so some other language had to be chosen. So no matter how good implementation of basic would be it would not spawn reliable portable OS. Also basic was good for beginers and some apps but critical software was made in other languages. No matter if those were COBOL, Fortran, Forth or ASM some very smart people found them not good enough to write operating systems even on much stronger machines. Those options (together with basic) were available long before C and still C was developed because was needed not by accident.
      Other thing is that monolitic OS is not a matter of language but OS design – please correct me if I am wrong.
      Same goes for being unmaintable or insecure. You can’t expect portability, functionality, efficient design, stability and security while not giving time. And security is least visible aspect of software for a common user. It actually became a common issue last years with popularity of internet and online payments (and people still don’t care about updates).

  7. The 6502 was the Edsel of that era. Few people made any money on it. The giants sucked all the blood out of the upstarts. The real chips came from Intel, and everyone could make money. NetWare and dBase II on top, and manicured lawns with a maid became the new norm.

    1. “Few people made any money on it. ” Many Apple investors made a LOT of money from the 6502. Electronic Arts made a LOT of money from the 6502. All of those dongles and boards sticking out of your old Apple II? All of the people who made that stuff made a LOT of money. You really under-estimate the Apple II market.

    2. People made shedloads of money off 6502 systems. In the early days of the Atari 400 there were folks I knew who sat around hand soldering memory expansions and basically printing money. The software industry also made a ton out of it and in the early days so did the programmers. After that it got taken over by people who couldn’t quite make evil enough to succeed in the music business and it went to shit. They cut direct sales out of the market, blew tons of cash on big names and paid crap for people to churn out safe repeats of old games with a new big film or other product time in.

      It’s true the 8bit computer industry folks on the whole didn’t make a lot but that’s because for once there was something resembling a highly competitive functioning market so the consumer got to pocket the win not a bunch of fatcat monopolists as has been the case for most of the period since.

      That is how markets are *supposed* to work.

      1. As someone who was in the games industry in the ’90s, the comment about people who couldn’t quite make evil enough to succeed in the music business rings SO true.

  8. I would argue that the C64’s bad basic actually produced better programmers in the end. Because of Commodore’s limited basic I was forced to start programming in assembly to do the things I wanted to do (write games) As a result me and most of my friends who had C64’s at the age of 12, 13, 14 years old had all learned assembly, and all of the prerequisites such as binary and hex math, how registers worked, better problem solving skills from working with a limited number of registers and limited memory. I as a high school drop-out was still able to become a successful professional developer because of the skills I picked up trying to overcome the limitations of that bad version of Basic.

  9. We know that Commodore outsourced basic, the design team for the Sid chip only worked a month on it, Jack Tramiel cut the MOS design team, Jack Tramiel had money issues that let him bring Irving Gould on board. If you are a worker at Commodore, you would most likely be told “we can’t do this” or “we can’t do that” because there was no good design team. The money issue could be why they were always arguing for less memory or silicon and saving face on blaming it on Japanese competition which never materialized..I believe it was an excuse.

    If you have money issues like this, you gouge which is what some of their marketing team did to the products. Increased price means less sales and less a company has to do because some of them were lazy. It is not a solution to run a company from a sustainability aspect which is why they had a tank, it was leaking and was never replaced.

    They didn’t invest in the C-64 line or C128 with bigger ram expanders because I feel they were favoring the Amiga line. When you want to favor one customer and discourage another customer, you are basically ruining your company by ending a product line that you could sell and ending a relationship with C-64 customers.

  10. “we agree wholeheartedly with him that Commodore were less than competent in marketing their hardware.”

    You can agree all you want but the fact remains that they dominated the personal computer market until the late 1980s.

      1. The computer magazines of the time were chock full of advertisements for Commodore peripherals: better joysticks, cooling fans for 1541 floppy drives, printer adapters, etc. The ads made up most of the magazine so it was hard to miss them.

        1. And the terrible tape drives where the screening was so bad our loading instruction sheet actually told people that if the game wouldn’t load to get the commodore tape deck as far as physically possible from the TV and computer as they could. That usually fixed the loading.

          It was designed to a price point, and the other vendors had equally amazing fails, like the original Spectrum Plus keyboard where the keys tended to fall off if you turned it upside down.

  11. The 6502 is still being produced today because they are making money at it.

    Ever hear of the EPYX Fast Load Cartridge? A lot of people became rich because of the Commodore 64.
    Ever hear of the CMD hard drive for the C-64?

    The whole computer revolution was based on Chuck Peddle. There were many different manufacturers of 6502 based computers.

    “The 6502 design is actually still produced today and is used extensively in embedded systems with production volumes in the hundreds of millions.”


  12. I’ve been told that less sophisticated systems produce more sophisticated users. It was true in my case. I developed skills to get around programming limitations that my friends with Apples never did. When I took programming in college my C-64 learned skills helped. My C-64 itself also helped in college because I was able to buy and use implementations of FORTRAN, PASCAL and Forth for a fraction of what IBM PC versions cost.

    1. Early Apple Mac was very sophisticated for its time but was also a very good assembly language development platform, the assembler was available for super cheap and there was plenty of documentation. 68000 has powerful and simple addressing modes and is a much better learning platform than 6502. It was very expensive for its time but with the educational discount it was an economical learning investment.

  13. As a VIC20/C64 user back in the day I recall lamenting the lack of sound and graphics commands in the BASIC. The super expander cartrifges added them but in general they weren’t available. It would have been nice to have a BEEP or a DRAW command like Sinclair BASIC.

    1. I recall my school had a few PET 4032s, which had BASIC ver 7.0, as opposed to my 64’s ver 2.0, and the thing I most coveted was disk commands. Just to see the directory, the 64 made you type load “$”,8 and then list — total kludge! Whereas the PET would let you just say ‘catalog’, which of course could be shortened in that weirdly C= way by typing C shift+A and boom, done.

      Funny thing is, I don’t think the PET had more ROM than the 64 did. The 64 used 8K for the ‘kernal’ and 8K for the basic. I forget the details of the PET but it seems like it had about the same amounts of each. AND it included a rudimentary ML monitor, albeit the one w/o the ‘assemble ‘ command

  14. The quality of Commodore BASIC is an interesting argument, though very non-apparent at the time. Basically, a better language allows better apps, and better apps makes a system more popular. But on a practical level, I don’t think they could have made the C-64 that much better.

    And when they did make a vastly better computer, the Amiga, they still went under.

    Commodore was a failing business machine company that happened to make some great and successful home computers (for the era and price). Sadly, Commodore execs never took that end of the business seriously. The word incompetence comes up a lot, but I would blame its demise squarely on the leadership’s inability to adapt.

    Marketing was an issue, sure. No doubt it was really bad. But that’s not what ultimately killed them. It was the giant bankrupt company they were dragging around.

    1. Yes better tools make more apps, but at the time nobody caredabout BASIC. Serious apps were not written in BASIC. Instead they pushed free hardware, services, support and tooling that people actually wanted – like cross assemblers and in 16bit days compilers. The questions we actually cared about were: documentation that didn’t suck too much, reference code, stuff to get our code onto the machine and free hardware.

      Commodore didn’t make the Amiga, it was designed by some of the senior Atari people who left to form Amiga Corp. Atari bought into it then shortly before they Warner were trying to flog Atari off. At the time Tramiel was trying to get Commodore to produce a 32bit machine and for his trouble got the boot. Tramiel then went on to start the ST project and bought the consumer electronics section of Atari off Warner.

      That’s why the Amiga is technologically a sequel to the Atari 8bit machines, and the Atari ST was largely built from off the shelf components and a mostly bought in OS (although that’s complicated since the OS ended up a joint project). The Amiga also ended up with a partly bought in OS (Tripos based) because they failed to get that bit of their OS done in time.

      The designer of the Atari 8bit SIO interface btw went on to be one of the lead designers of USB …

      So whilst there are plenty of reasons Tramiel was disliked he was on the ball about Commodore getting off 8bit, and he wasn’t the one responsible for the final car-crash ending.

  15. Ever hear the saying ‘If all you have is a hammer”. If all you know is Basic then sure compare Basics. But your forgetting that Basic was intended for Beginners and there are other tools out there better suited for different tasks. Don’t limit yourself to a Basic only toolbox.

    1. SO right. Beginners’ Allpurpose Symbolic Code. The fact that it was so often the first language people wrote compilers for suggests that it’s also the easiest language to write a compiler for. I learned BASIC on an HP2000E then taught myself Fortran (which was trivial, nothing to brag about) then learned Assembly when I got a PC-XT, not because I had to but because it was interesting. PEEK and POKE had taught me that there was something interesting there.

  16. I still would like to know why the author blames the Commodore 64 for Basic and not the Amiga. There were plenty of other computers and…

    Amiga users pretty much abandoned BASIC because it was too slow. In some ways, what the Amiga did could be felt today. The Amiga got programmers into C language.

    I disagree that BASIC can be felt today because computer clubs are mainly gone. The library teaches people how to get on the internet or maybe some clubs teach Word or Excell. I believe only 2% or less of the American population knows how to code so BASIC failed.

    The bias of the article is about Commodore and not the IBM PC, Dartmouth College or the PDP-11. It’s like saying Commodore owned the world of computers and disregards BASIC on other computers because of their accomplishments.

  17. The Commodore-64 was responsible for an entire generation of hackers, even more so than Apple II. Way back then as a pre-teen I couldn’t afford an Apple II, but a C64 was (barely) manageable (after some other computing platforms). It was an absolute blast.

    The point of the C64 was that you were put face to face with the hardware. None of those inefficient ROM-based crutches. If you wanted to make noise, you had to know what you want and tell the hardware to do it. It was f*cking awesome. Hacking the platform became a joy like nothing else. 80-column display? No problem. Fast floppy drive reads? Which option do you like? It taught how to get every last clock cycle and piece of performance out of the hardware. And 64kB was a lot of memory for its time, even if you couldn’t use all of it.

    I still remember those mornings when — after an all-nighter — I finally figured out how to do something during a single retrace interval after squeezing code that couldn’t possibly do it and counting clock cycles to get it just right. I’ve gotten close with later platforms, but it’s never been the same.

    All Commodore had to do was add more clock cycles to the darn thing. They could’ve kept the same cr*ppy basic, the same cr*ppy peripherals. Just make the CPU faster. That would’ve been cheap and easy. But noooo, they had to mess with the formula.

    What killed them was looking for the next “big thing” instead of milking the cow in front of them.

    Oh, and I had two generations of the Amiga, too. That thing was such a let down. When it worked, it was the absolute awesomest thing out there. But then the next thing would be a guru meditation. Developing code for it seemed to be an endless cycle of code-compile-gurumeditate-reboot. It was just a bridge too far for the big C.

    What happened with C64 and didn’t happen with later Commodore products was the app support. There seemed to be thousands of apps for the C64. None of the later Commodore products had the same following (or at least I don’t recall it being that). Had they figured that one aspect out, things might’ve been different. But that’s history.

    1. I identify so well with what you wrote. I cut my teeth on the C64 and I wrote many programs which I compiled it BLITZ! I wrote “magic hacker” after watching wargames (I am 49 now). It would dial Sprint or MCI and then furnish a random or sequential code of the correct length and then call Comp-U-Serve. The only 24/7 thing available to verify the dialing sequence. I will never forget the line of code that made it all work. IF PEEK (56177) 127 then. The program would pause for up to 45 seconds to confirm the carrier was there. If it was, it would print the code on the screen. If not, it would hang up an recycle. I could now call all of the west coast BBS’ free of charge and download anything I wanted. My parents got me my own phone line because I constantly tied up theirs. I just ended up using them both! If it was not for the Commodore 64, I would not be the IT pro I am today. My dad always told me this was the best investment he ever made. Oh and that the C64 was not a toy and I was not to touch it!

    2. ‘All Commodore had to do was add more clock cycles to the darn thing’ … ‘What happened with C64 and didn’t happen with later Commodore products was the app support’

      So true aki009, you nailed it, that was the problem not a basic interpreter that was rarely used once 6510 machine code was learnt.

  18. BBC BASIC was strong enough to survive as a language independent of the BBC Micro. Indeed it thrives to this day in the form of ‘BBC BASIC for Windows’, ‘BBC BASIC for SDL 2.0’ and ‘Matrix Brandy BASIC’, which are greatly extended and updated dialects that nevertheless retain good compatibility with their ancestor of nearly 40 years ago (for example all still support the BBC Micro’s MODE 7 teletext-compatible display). https://www.bbcbasic.co.uk/

  19. It is practically impossible to teach good programming to students that have had a prior exposure to BASIC: as potential programmers they are mentally mutilated beyond hope of regeneration.

    1. That’s not been true since structured BASICs were introduced in the 1980s. Indeed BBC BASIC was one of the very first such structured BASICs, it having been mandated by the BBC (under the guidance of educational specialists of the day). In the UK, schools and colleges taught BBC BASIC as the introductory language of choice for decades (and some still do) and until recently it was recommended by the Oxford, Cambridge and RSA examining board.

  20. That would mean programmers got a job without any schooling or programming experience.
    In college, my professor taught us to do loops and think differently so its not impossible.

    In College, they make you take other courses and languages.

  21. “I hate to be the one to burst your bubble but this is the standard operating model for the US corporation and it has been for hundreds of years. You can read any number of history books about US industry and this is always the case, you can be talking about cars or canned meat or furniture or prescription drugs, so maybe you can cut the fake indignation.”

    When you go to a computer club and a salesman or programmer from a company has courtesy and supports you and shows up, you know you are in a computer cult when an Amiga user screams at the company representative and says, “we have to be supported!” even when there is no money or product interest.

    1. It wasn’t so much of a cult than just what was left over from the user base when all the sane people had already seen the writing on the wall and moved on. When everyone else leaves, you’re left with the Chris Chan types.

      1. This behavior started with CBM and from users from the beginning. It started with a lie that Commodore users were more loyal than other platforms.

        It started with the computer wars. It started from the capabilities listed from the computers themselves.

        The C64 was a computer you could buy at Kmart or Toys R Us. Sears got into the mix and some bookstores at the mall had the Amiga featured.

        BASIC is basically a bootstrap that makes the whole system usable. It made people who had no knowledge believe they were invincible. Take away Basic and languages and what do you have? Lots of people with no ability to make it work let alone do anything useful.

  22. Commodores demise wasn’t BASIC and BASIC lead to a computer revolution. Micro managing, greed and failure to innovate or recognize innovation in their back yard. Then stiffing sales, repair, and even supply lines.
    That was demise. Amigas brilliant light not enough amongst and against the darkness. Overpriced slow IBM PC compatible only sad. Even the aftermarket boys laughed at that.
    $300 dollar computer started People who would never have considered programming starting to code. Even more so with $100 computer(s). I used VIC-20 as a replacement KIM when it got to that price.. Actually cheaper and available. Picking on CBM BASIC is so sad. Picking on BASIC just wrong. I liked it more than any other BASIC at the time. Easiest editor even on VIC20. PET … Meh. Spaghetti code and rats nest drop though programming but the shit worked. It is for Beginners and quickly advanced to combining assembler. Maybe its ‘digressed to’ but squeezing every bit of a 64KB programming model and a quarter MIP processor required some Black Magic. Many didnt have anywhere near 64KB actual RAM until the mid 80s. PC-1 (5150) Had 16KB and you freakin paid for it. Soldered to the motherboard. 4116 and the soldering hell of another time. Followed by DIP chip popping and why you only turn off your PC when leaving for days.
    Compiled languages were of course next up. Interpreters too slow on already slow microprocessors. Minis had the resources to build and run a compiler. Once compiled would run much faster and take less resources ideally. Micros had to wait until the 64KB and storage were more common. Turbo Pascal and the< 64KB compiler another story. Then Turbo C. Is nice to hear Billy upset and have cheap fast compiler. Cake and Ice cream.

    1. Hi. Author of the original blog post here.

      I don’t think you understood what I was trying to say at all. I wasn’t attacking BASIC in general. I *like* BASIC. I was attacking Commodore’s use of MS BASIC 2.

      Already, a year before the C64 was released, there was a machine with same CPU whose BASIC had named procedures, REPEAT…UNTIL, IF…THEN…ELSE and graphics, sound and music commands. Have you ever used any other machines from that era with better BASICs, to understand the context at the time?

      1. You may have noticed someon else in the comments speaking of “demise” but theres no reference in your article or jenny list article. Huh. Howd that happen? Is a mystery.
        Like a any good BASIC interpreter of the day I dont understand anything. Only to interpret and move on the program from there using the Data and experience I have in my collective albeit finite memory. Ha.
        If you said 70’s to early 90s and excluded M$ buggy source crap pushed on pretty much every manufacturer under the BS that its the best BASIC to have in your machine then YES. Abso-freaking-lutely there were better.
        Most of which were in non 6502 machines. From Mainframe, to mini to micro and back again there have been many BASIC implementaions. Compiled and Interpreted. Most of the lack of command enhancements and more specific peripheral commands wasnt due to manufacturers but their licensing agreement with guess who? Wanna pay for that? No? Then you cant use our Name or our BASIC. That we wrote. Didnt steal it. It was stolen from us. We invented micro based BASIC. Ya thats it. The litigious nature of certain larger corporate entities were very quick to shutout and shutdown. Many systems came across my way for ‘evaluation’ that were shutout and folded overnight as soon as any one of those corpy scumbags found out that there’s a machine with better BASIC firmware. But what about our NDA? Who squeeeled?” cries the startup. Gales of stupid laughter eminate from dark corner shadowy figures of the hollow abyss.
        Non M$ BASIC implementations were on two actual Robots. Though one of those didnt have much more than 7-seg LED display so no graphics commands but it did have full motion commands for motors. Sound , Speech, time and various sensor commands too.
        Could go into things like Pick Basic but thats just painful and rude. No not Pic Basic.

  23. I translated and investigated simple basic programs across platforms and I saw only about up to 3 differences.

    For the Commosore 64 to be singled out as the problem is rediculous. Lots of kids buy what they are familiar with and kids were more exposed to the Apple Ii, Iii+ and Apple Iie in school.

    The whole argument does not have enough proof because you would have to show all the other manufacturers updated basic. I am not talking about 3rd party sellers.

    1. Sinclair actually didn’t for the spectrum, it was really just a ported ZX81 basic with add ons for color, the engineers wanted to rewrite it from scratch to double the speed but Clive just wanted it out the door.

      However, it may be seen as a business move also for Commodore to not break backwards BASIC listing compatibility with the PETs which had some penetration in the educational sector, so was probably currying favor with the educational purchasers.

      Nobody really wanted to go too far off in the weeds with their BASIC anyway and make it too much different, minor dialectical differences were okay, but if little Bobby couldn’t get his Usborne book or magazine listings to work no way, no how, then Daddy wasn’t gonna buy that machine.

      1. BBC BASIC was more different than most: no PRINT USING, no PEEK/POKE, non-standard file I/O etc. I think there was less pressure to be compatible with Microsoft BASICs in the UK, and Acorn’s earlier Atom BASIC was even more strange, for example it used semicolon rather than colon as the statement delimiter! That was too much for the BBC to swallow, and we insisted that the more extreme peculiarities were reined in for BBC BASIC, but it remains a notable outlier.

  24. Programming is messy

    “But that is not the way programming works. Programming is messy. Programming is a mix of creativity and determination.”

    “There are no books that teach you how to solve a problem no one has seen before. This is why I don’t want my kids to learn syntax. I want them to learn to solve problems, to dive deep into an issue, to be creative.”


    1. Interesting article, thanks for posting the link.

      One thing that author may have overlooked, is how any programming language (taught to kids today) may be obsolete by the time they hit the workforce. I thought Java was the future, a dozen years ago when I was in college.
      Of course, these programming camps for kids COULD be teaching them COBOL (which I believe will still be around when today’s kids are ready for retirement- or euthanasia, which might be how people are “retired” in their future)

      1. > One thing that author may have overlooked, is how any programming language (taught to kids today) may be obsolete by the time they hit the workforce.

        Not overlooked; it’s a widely-held view.

        But I think it’s more important to distinguish between different kinds of education and educational institution. Schools and universities are not vocational training. They don’t teach work skills, such as how to do accounts or book-keeping, or how to plumb pipes, or how to wire up an electrical ring-main.

        No, they teach reading, writing, literacy, numeracy. Cognitive skills.

        Teaching programming was a part of many curricula once. Not teaching whatever language is currently used in business: teaching how to analyse a problem, break it down into steps, and how to show a computer how to perform those steps. And if you do that well, it doesn’t matter what language you use. BASIC is a decent starting point, but so is Logo. For more advanced kids, maybe Smalltalk. For undergraduates, a Lisp such as Scheme.

        Schools are for teaching methods, not work tools.

    Did I understand correctly: someone gave Jack a “good deal” (note that this imply that Jack was not responsible for good deal but someone else) and then Jack, even if someone gave him something, was not capable to keep things together?!?

    This is so absurd, and childish, that it hurts!

    Jack enable to millions of people to get a computer in home (something that Bill Gates like to brag) in 80s. ST was also great computer, first computer in the world that had 1MB of RAM and cost less then 1000$!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.