So long and thanks for everything your company did to enhance my childhood years. A true pioneer of modern day computing.
The founder of Commodore, one of the driving forces in the early history of the personal computers, has died at the age of 83. Tramiel, born in 1928 as Jacek Trzmiel to a Jewish family in Poland, emigrated to the US after the Second World War after losing his parents in Hitler's camps. Tramiel spent time at Auschwitz and at a …
I owe you my career. Without your invaluable and incredible contribution to mainstream computing in the form of the C64, I wouldn't have got into computers and wouldn't have the skills I have today, and I'm sure many of us here can say the same. Rest in peace, sir, and may your legacy long endure.
This post has been deleted by its author
I owe you my career.
Another me too. First computer I owned (shared with my brother) was a Commodore 64. I used to thoroughly enjoy working through the tutorials and code listings in Commodore User before it became solely a games review magazine. Without that start, I wouldn't have been in a position to work as a professional computer programmer some ten or twelve years later.
Like Steven Roper, I also owe Jack Tramiel my career, starting with the VIC-20, them moving through the CBM-64, Commodore 128, Amiga A-1000 and onto the A-2000, I was a major devotee of Commodore computers, and consider them to be as important in the history of our industry and the BBC Model 'B' and the Sinclair ZX-81 in terms of getting young people enthusiastic about computers and working in IT. I learned more about programming and hacking on these machines than just about any other platform, and as such I can say Jack Tramiel's company and it's products have had a truly profound and important effect on my life.
In losing Jack Tramiel we have lost one of the great's. Most can only wish they can have the effect on as many as Jack Tramiel had. May you rest in peace, Sir. I am sure many on this forum and in IT would wish to convey their condolences to his wife and children for their loss.
Whenever I see documentaries on the beginnings of home computers they invariably focus on Apple/Steve Jobs and Microsoft/Bill Gates. They usually ignore the likes of Tramiel/Commodore or at best make them out to be people who produced "games" machines of little relevance.
Yet as someone who was around at the time, Steve Jobs was a fringe figure selling hardware at eye watering prices to a tiny market and Bill Gates sold dull business software for use on equally expensive computers. It was the likes of Tramiel who ignited the publics interest in computers. Tramiel, Sinclair, Sugar, Curry were all people who bought computer to the masses. They produced hardware that was accessible and affordable.
Tramiel was a real power. At one stage Commodore were building 400,000 C64's a month, selling 17 million in total. He then went to Atari and had a major feud with his former company. He also launched the ST and turned around Atari who had never really recovered from the early 80's video game crash.
Granted he wasn't an engineer, but he recognised the need and made the business decisions to get Commodore into the market and pushed for the development of machines like the C64 because he saw the need.
Yes the man had faults and was a hard nosed businessman but now isn't the time. Apple, Microsoft et al all owe Tramiel and people like him a huge debt of gratitude for putting the foundations in place for their empires.
It's a Silicon Valley bias thing; Tramiel dared to be over on the east coast, Sinclair, Sugar and Curry weren't even in the same country.
Tramiel was as important to the business as any of them. His cutthroat approach to price cutting can just as easily be cast as striving to include more features at the same price, hence pushing the industry forward.
Simple reason - Apple and Microsoft are still around. The others aren't. With the exception of Tramiel, all the others were one-hit wonders. And Tramiel was a two-hit wonder.
None of them saw the need to sell their existing product *AND* build on that success sustainably. So Acorn died by not working sustainably, and Sinclair, Commodore, Amstrad and Atari died by producing a load of shite that no-one wanted.
And on the other hand Apple could have easily gone bust if Jobs hadn't returned, and Microsoft's success pivoted on some luck with their original DOS deal with IBM. Arguably if Microsoft had been pulled up more on some of their anti-trust stuff they could have been in serious trouble.
There but for the grace of god.
>> "And Tramiel was a two-hit wonder."
I guess, technically, he was a "four hit wonder", if you want to argue the toss (PET, Vic20, C64, Amiga). All four were cutting edge at their time, and prolific (and relatively speaking, cheap). There was no competitor to the PET, the ZX80 and 81 were blown away by the Vic20 not long after they came out, the Spectrum wasn't a patch on the C64 (the BBC-B was about similar or slightly better, for non-gaming, but much worse for just gaming), and the Amiga was probably on par with the ST, but a huge hit and sold plenty.
>>"None of them saw the need to sell their existing product *AND* build on that success sustainably".
You don't think a 12-13 year span for Commodore, counts for anything (selling 10s of millions along the way)? The razor thin margins though does explain why they didn't last (along with the commoditisation of the (IBM) PC (clone) through the mid-late 80s and largely complete by the mid-90s).
Until the (affordable) PC came along (in the late 80s/early 90s), many small businesses were running themselves using Vic20s, then C64s, then Amigas! (not to mention the thousands of PETs installed in schools, as far back as the very late 70s).
The Amiga wasn't a hit for Jack. It was initially developed for Atari while Jack was at Commodore. Ironically it only became an Commodore product once Jack had taken over Atari.
Jack then spent a hell of a lot of money preventing Commodore from releasing the Amiga which bought his engineers time to produce the ST.
That is the very very basic version of events. It's rather more complicated than that but you can read up on it here if interested:
Bringing computing to the masses is a non-trivial accomplishment. Offering things that no one can afford to buy is not really that meangingful. It took the PC industry 20 more years to reach the level of price acessability that Commodore created. Commodore didn't stand the test of time but it did help prime the market and create a generation of computing professionals.
The PC market is built on the bones of Commodore, Atari and the rest.
GEM - Graphic Environment Manager
Copyright Digital Research - 1985
(For Atari ST systems)
At that time, Pc's were in text mode only and Apple II was sold with a similar graphic OS on a much cheaper hardware. Guess what, the Apple machine was the more expensive yet the most underpowered (arguably with Pc's).
I could never understand how they could do it but damn thing worked fine. You could buy a mouse for c64 and use it too.
I will say a very interesting fact. GEM guys actually released a Y2K update back in 90s. That is the degree of responsibility, understanding of "customer relation" such companies had.
They ask why people feel nostalgic you know.
I could never use Atari ST my own but I know one of its legendary shareware developers. Graphic converter author. Using his millions downloaded software, I found a cosmetic, mega cosmetic typo bug. Not to forget it, I mailed the issue (!) to him on first new years day trusting to "on demand" nature of mail. Used "trivial", "cosmetic" words on subject line so developer won't flame me.
On first day of new year, 20 minutes later, 9:20 local German time, I had a mail reply thanking for report and a beta link to updated version.
People miss that kind of scene. Device (Atari St never had Miner genius) specs, although way more modern than PC isn't that relevant.
You'll all be thinking of GEOS rather than GEM, surely? GEM was the one that came with the Atari ST and was also available for all of the other 16-or-better bit machines from Digital Research, GEOS was — I think — the Commodore 64 one with the surprisingly complete set of applications. I've seen other 8 bit GUIs but never anything that went significantly beyond a basic proof of concept, with a calculator, text editor and nothing much else of production use.
Please, El Reg & Iain Thomson, please fix that implication that the PET was some kind of reaction to the Apple II; chronology doesn't support it:
January 1977: in the mainstream (then as now) CES show in Chicago, Commodore introduces the PET.
April 1977: Commodore ships the first PETs, priced at $595.
April 1977: The Apple I's price is reduced from $666 to $475 (which of course doesn't include many of the PETs features, such as a screen or a cassette deck).
April 1977: at the West Coast Computer Faire, Apple introduces the Apple II.
June 1977: Apple ships the first Apple ][, priced at $1298 excluding monitor, etc.
These days the fiction that the Apple II was "the first" home computer seems pervasive, just as the equally fictitious notion that the Mac was "the first" WIMP machine seems to be common.
And that's unfortunate, because it's people like Jack Tramiel that showed Apple what to do.
I remember working in a science lab as late as 1992 and there was a PET in the corner doing data collection from the experiments we had running. My super told me that PETs had a lock on data collection in labs, which was either him winding me up about a tiny niche market of a niche market, or was true...i never found out. Any Reg readers out there care to comment?
>> My super told me that PETs had a lock on data collection in labs ...
Maybe not a lock (by which I assume you mean a commanding market lead), but I recall they were popular for one good technical reason ..
Fitted as standard on the back of the PET was an IEEE-488 bus (aka HPIB or GPIB) - although they used a cheap and nasty card edge connector rather than the quality ones in the standard.
There were (at a price) many lab instruments with an IEEE-488 ports on the back, so it was easy to hook them up and send/receive commands/data. So it would be fairly easy to (for example) send a command to a power supply to set a certain voltage on it's outputs, and then read in the outputs from the device under test using a GPIB equipped voltmeter.
There were other computers that could do this, such as some rather neat jobs from HP - bet you can imagine the relative prices ! That gave the PET an edge.
There were lots of other stuff that would hook up to GPIB - printers, plotters, floppy drives, ... In some respects it was to connectivity back then as USB is now - and in between we've had the good (SCSI), the bad (pick any one from many proprietary systems), and the ugly (parallel port disk drives, yuck). I can think of 3 reasons IEEE-488/GPIB never became ubiquitous - it probably was "not cheap", it was probably seen as "belonging" to HP, and the cables/connectors were fairly bulky (but they were stackable which made daisy chaining very easy).
IIRC it was also fairly easy to program with as well. I recall "some years ago" getting paid for a data transfer job off a PET based system (running a database off floppy disks !). I borrowed an IEE-488 card for an Apple II and wrote a program to handshake like a printer and slurp the data to a file on disk. A quick report to dump all the data out from the database, "printed" to this other computer, then load into new database and job done.
But there's more. There were (from memory) three competitors for the desktop at this time. Two are well known - the PET and the Apple II. IIRC the third was the TRS-80. My memory is a bit vague from that long ago, but I think any one of them could have made it - but Apple won for one simple reason. When Visicorp brought out Visicalc, they looked at what was available and decided that Apple had the better specs - basically more memory with up to 48k on-board and up to 64k on an add-in card. Visicalc brought automation to spreadsheets (yes, they existed before computers and calcs were done by hand) and was snapped up by ready users (especially department managers keen to escape the tyranny of their IBM centric MIS departments). That meant sales of Apple IIs, and the others lost out. The rest, as the saying goes, is history - the TRS-80 disappeared into oblivion along with a myriad of others, the PET struggled along for a while (especially in niche areas such as lab work), Apple did "fairly well".
In my summer holidays, I had a temporary job with the first computer shop where I lived in Devon. The shop sold PET systems, peripherals and software. The majority of the customers were small businesses who used PETs for accounts, invoicing, etc.
As I remember, there had been some hobbyist microprocessor kits around but most of these had only hexadecimal keypads and 7 segment displays - the PET was the first affordable system available in the UK that had the associated peripherals and software that made it more than just a toy for enthusiasts. The Apple ][ and TRS-80 came a bit later.
The 68000 had a nice 32-bit architecture inside, but the bus width was 16 bits. For this reason it was usually called a 16-bit processor at the time.
If this seems unfair, recall that the Z80 was considered a 8-bit processor, even though it had 16-bit registers and could do 16-bit arithmetic in one instruction. Its main competitor 6502 (used in the PET and Apple II) was completely 8-bit: All registers were 8 bits only, except the instruction pointer.
It was even named as such.
ST = Sixteen/Thirty-Two
STF = ST with a Floppy drive
STFM = ST with a Floppy drive and built-in TV Modulator
STE = ST enhanced (for certain values of "enhanced", anyway)
I believe there was also an STM model, but I never saw one in the wild, and there was a Mega ST variant which was an STF with the internals in a separate box to the keyboard.
This was followed by the TT (Thirty-Two/Thirty-Two) which had a 68030 processor and a 32-bit bus.
Finally there was the Falcon, 68030 processor with a separate 56001 DSP, strangely reverting to the 16-bit bus though. This was potentially a lovely bit of kit but it never really went anywhere - Atari killed it off so they could build the Jaguar but then found themselves competing against the Playstation 1, and the rest is history.
Anyway, thanks for everything Jack. Hope you're happy wherever you are.
I heard that as well (and also that TOS stood for "Tramiel Operating System"), but the Atari official line was "Sixteen/Thirty-Two" and "The Operating System".
It may well have been that these were backronyms - the engineers and marketing guys were told it had to be called the ST and thought "well, it's got a 16-bit bus and a 32-bit processor, we can make that name work" - but you'd have to ask someone who was actually in the meeting to get a definitive answer to that.
I fondly remember reading the Silica Systems calalogues about 1988 and thinking "I'd love to have an ST and the 520STM is fairly cheap". Of course I was about 10 at the time and didn't realise the value of money. Still having an Atari 65XE I often daydreamed about having an ST, until about 2 years later when I got a 520STFM :-)
I wouldn't be surprised if the 520STM are worth a few quid now being more rare than the STFM.
I had a Falcon, awesom bit of Kit.
Mine had a 040 upgrade, a Maths coprocessor upgrade, 16mb of RAM (I think) and 1GB HDD!
Running Cubase VST (with multitrack D2D recording!) it utterly destroyed anything the other manufacturers could muster.
Was tempted with the 060 upgrade, but I would need to sell the car to pay for it!
Ah happy days.
heh, still remember 1200 with 68020 board poking out from underneath with a 4mb 72pin simm and 3.5 400mb drive stuffed next to the keyboard . King of the hill (no monkey island 2 floppy swapping for me and I could play sim city 2000).
This was the machine that got me into soldering, hacking stuff to bits and programming. As most people on here have already said, thanks for my career training!
If you are going to call the 68000 a 16 bit CPU because the external bus was 16 bits, then the 386SX is also a 16 bit CPU, as the external bus is 16 bits. But the 386 would be a 32 bit CPU. Since the instruction set is the same, the core is the same, and only the bus changes, that is a nonsensical result.
Moreover, as the CPUs move to high speed serial interfaces (Hypertransport, PCIe) you move to an external bus that is 1-4 lanes wide - will you call those 1 bit CPUs?
Not the same thing at all, the 68000 was a mostly 16-bit implementation of a 32-bit ISA, 386SX was (I believe, never owned one) a 386 in a cheaper package suitable for cheaper motherboards. In other words there is no 32-bit bus version of the 68000 core because the core couldn't utilize it. The 68020 was the first 32-bit implementation of the architecture and is a completely different beast internally. So the 68000 really was a 16/32-bit processor, or a 16-bit implementation of a 32-bit ISA. (Still that's a very elegant way of doing things as ISAs obviously tend to live a lot longer than particular implementations.)
"the 68000 was a mostly 16-bit implementation of a 32-bit ISA,"
Sorry. Dead wrong. The 68K family ALL had 32 bit data registers (D0-D7) and 32 bit address registers (A0-A7).
The only real difference was the 020 and up added an MMU to allow for virtual memory, whereas the 68000 didn't have the MMU.
Resp a David D. Hagood
Sorry David D. Hagood but you are totally wrong: Motorola 68000 is a 16 bit CPU (and one I loved).
True it has plenty 32-bit registers (D0..D7 32 bit data registers and A0..A7 32 bit address registers being A7 the stack register), but it doesn't make 68000 a 32 bit CPU becuse internal bus (inside CPU) is 16 bit wide and ALU (arithmetic-logical unit) is 16 bit wide so it can only do a 16 bit operation each time and if you need for expample add two 32 bit numbers that operation was done in 2 steps, 16 bit each time.
16 bit internal bus width and 16 bit ALU width determine that Motorola 68000 is a 16 bit CPU.
Motorola 68008 is also a 16 bit CPU becasuse althought external data bus is 8 bit, internal data bus and ALU are 16 bit.
For the same reason Intel 386SX is a 32 bit CPU because athought external bus is 16 bit wide, internal bus and ALU are 32 bits wide.
The 1st Motorola 680xx CPU wich is a 32 bit CPU is Motorola 68020: 32 bits internal bus (and external) and 32 bit ALU, ie its ALU can perfom one 32 bit operation in one step while 68000 ALU would require two steps (low 16 bits in one step and high 16 bits in other).
For your consideration I show other examples (CPUs I have programmed): Zilog Z80 (CPU used in most Sinclair computers, and MSX and lot of others) has some 16 bit registers but it is an 8 bit CPU and if you use for example the "ADD HL,BC" instrucction wich adds tow 16 bit registers that operations is made in two steps inside CPU because ALU is only 8 bits wide (you can watch cicle count in 8 bit add vs 16 add in Z80 and you'll se that it takes at least double clock cicles).
Other CPU example: Saturn CPU, used in lots of HP programmable calculators like HP-48 series, has 64 bit registers and instructions to operate in 64 bit registers but ¡¡it is a 4-bit CPU!! http://en.wikipedia.org/wiki/HP_Saturn_(microprocessor)
I recomend you this book excellent book on M68000: "MASTERING THE 68000 MICROPROCESSOR" by PHILLIP R. ROBINSON
You can download it from http://www.bombjack.org/commodore/amiga-books.htm
Page 12 speaks on "Is the 68000 a 16-bit chip?".
Like a rule (not theorem): for a processor to be x-bit wide it must have at least x-bit internal data bus and at least x-bit data ALU and at-least x-bit registers. M68000 only has 32 bit registers but internal data bus and ALU are 16 bits.
Thanks for the C64 Jack - my love for tech started there, and my lifelong enjoyment of tinkering has kept me from purchasing a Mac ever since; I've got close, but dammit I love pulling things apart.
Thanks for the vision to include wonderful things like the MOS 6851 SID chip - in my view one of the most important pieces of silicon ever designed.
Thanks for the machine that allowed me to enjoy things like Elite, Mercenery, Ultima 4, Uridium, Paradroid, Impossible Mission, Flight Simulator 2.
Thanks for an architecture a 17-year old of medium intelligence could work out assembly for, letting me discover things like fractals all by myself, two years before James Gleick's book.
Thanks, Jack for your works.
This post has been deleted by its author
Thanks for all the enjoyment your machines brought me during my childhood years, sparking my current enthusiasm and love for all things IT.
You were largely responsible for the meteoric rise of the home computer market by providing computers at a realistic price point and in doing so, stepped on the toes of your contemporaries who in return ensured that your place in history has largely been forgotten.
Thank you, and so long.