Was the Apple-1 the first 'personal computer' ?

16 posts / 0 new
Last post
Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
Was the Apple-1 the first 'personal computer' ?

The Apple-1 is a singularity in the history of computing. Unlike any other vintage computer, the few surviving originals fetch phenomenal prices at auctions, up to 1 Million US$, which, even considering that the US$ has lost 98.8% of its purchasing power since the inception of the so-called 'Federal Reserve', still is quite a bunch of money, even at the time of this writing, in an era of galloping inflation caused by Green Toilet Paper "printing".

 

A currently running auction for a  partially  surviving early prototype Apple-1 (with about 1/3 of the PCB missing) here:

 

https://www.rrauction.com/auctions/lot-detail/345735106445006-steve-jobs-39-s-apple-1-computer-prototype/?cat=0

 

at the time of this writing stands at $278005, which is mind boggling. Why would anyone pay that kind of money for such an artifact ? (at the time of this writing you can get a nice single family suburbia home for that, small and humble, not a mansion like it used to be 50 years ago, but in a few years from now you might get only a garden shed for the same amount of money, if the "printing" continues.)

 

In the thread:

 

https://www.applefritter.com/content/prototype-apple-1

 

I floated the idea that the reason might be that some people may see the Apple-1 to be the first incarnation of what we call 'Personal Computer', and if so, examples of it would be priceless. I've opened this new thread to move this discussion out of the prototype auction thread. It deserved its own thread - and lots of thoughts and discussions - by many smart people on this list. Because it might be able to explain the "Apple-1 singularity" with all its consequences like a vivid clone building scene, and not only these mind-boggling prices at auctions.

 

Here are the criteria for the class of computers we call 'Personal Computer' (see my post #27 in the auction thread):

 

- a single motherboard computer (NO backplane !)

- with a microprocessor CPU capable of numbers and text manipulation

- with  expansion slots  for daughter cards in lieu of the backplane

- power supply unit plugs into motherboard by cable trunk

- full text keyboard plugs into that motherboard

- full text video output

- DRAM memory on the motherboard

- on board firmware to (at least) start it up at press of a button / no toggle switches for start up

 

I did not invent or make up these criteria - these come from a motion / petition at IEEE which sought to elevate the Apple-1 into the  IEEE "Hall of Fame" as a milestone in the history of computing. Take one of these criteria out and you will have something else, maybe an educational toy, or a computer for a mid sized business, or a programmable calculator, etc., a different class of machine., but no 'personal computer' anymore.

 

We could also work backwards in time from an example which indisputably is indeed a true 'personal computer', the IBM PC of the year 1981. It ticks all the boxes for the above criteria. It was affordable by a single person. And if you look inside you can see its key ideas were stolen from the Apple II: a motherboard, with a microprocessor as the CPU, plenty of expansion slots, separate power supply box with a cable trunk plugging into the motherboard, ... etc., read the above list of criteria.

 

And as we all know, the Apple-1 was the predecessor of the Apple II ... and it also ticks all the boxes.

 

So is the Apple-1 the first 'personal computer' ?

 

Comments invited - but please don't indulge in sophistry about the meaning of words, such as what the meaning of the word 'is' is (cit. Bill Clinton), or what a DRAM is, or what a microprocessor is. Stick to the above criteria and show us any computer which ticks all the boxes and came before the Apple-1, and was produced and commercially available at a price affordable for a person (such the machine could be used as a 'personal computer' and not as a time shared machine etc. of the prior computer age). Note that I did not mention a price tag in the above criteria list,  because in the age of fraudulent fiat currency, prices in form of numbers are meaningless. The logic goes as follows: if all the criteria in the above list are applied and leveraged, you can build an affordable and useful computer which can be used as a 'personal computer' by the average person. Omit one (or more) of these criteria and you end up with a less useful machine, or a too expensive one. This is the "secret recipe" for the personal computer. And I still think Woz was the first one who saw it, built it, and commercialized it (well, OK, the selling was done by Steve Jobs and, in the beginning, by Paul Terrell through his BYTE SHOP). Also keep in mind that "flights of fancy" in the patent literature or one-offs done by individuals don't qualify. The example in question must have been series produced and sold as a commercial product. 

 

Comments invited !

 

- Uncle Bernie

 

 

 

 

 

 

Offline
Last seen: 2 days 53 min ago
Joined: Nov 24 2013 - 13:23
Posts: 226
Why

Why does it have to be a single motherboard?  Or not be bus based?  Seems like criteria bent to fit Apple.

I feel like the Sphere 1 was a far more complete personal computer than the Apple 1 was - chassis, screen, power supply, keyboard all built in.  And it was out before the A1.  The Apple 1 wasn't complete - it was just a board.  You still needed to get your hands dirty and find stuff to make it workable.

Digital group's 8080 based machine was also out before the Apple 1 and had built in system monitor, video and keyboard I/O, and so on.  I feel like the bus vs single board with expansion slots thing is kind of an arbitrary distinction.

 

Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
Why a 'personal computer' must have a single motherboard

In post #2, falter wrote:

 

"Why does it have to be a single motherboard ?"

 

Uncle Bernie answers:

 

a single motherboard is the key trick to make the machine cheap to produce. Once you use a backplane like in the S-100 bus or DEC or DG minicomputers it gets more mechanically challenging and more expensive.

 

This has been made possible by progress in PCB manufacturing technologies. The Data General Nova of 1969 was the first minicomputer which pushed PCB technology to the limits, one Nova PCB had the maximum size of a PCB panel of the time being. This did cause DG some manufacturing headaches but contributed to the cheaper price compared to the DEC machines.

 

As for the Sphere-1, this came very close to the concept, it even had DRAM, but it was a multi PCB machine and this made it expensive. Again, a single motherboard is key. DRAM use, too. The Commodore PET2001 - other than its ridicolous 'chicklet' keyboard - screwed up because of the expensive SRAMs. Commodore had bet they could use in-house (MOS Technology) SRAMs at cost but these died like flies, and so they had to change the PET to take industry standard SRAMs bought from a real memory maker, and they had to jack up the price of the PET considerably.

 

This should not distract from the fact that many microcomputers of the 2nd half of the 1970s and the early 1980s did use SRAM - but they used trailing edge SRAMs which were already sold off at fire sale prices. The most notorious example is the Sinclair ZX-80 and ZX-81 which used cheap trailing edge ICs throughout. Ohio Scientific being another example. The Superboard II of 1979 used 2114 SRAMs - at the time trailing edge and already cheap.

 

However, not cheap enough ... I really struggled money wise to expand my Superboard II from 4k Bytes to 8 Kbytes RAM. Without writing some magazine articles featuring programs I wrote to run in 4kByte it would not have been financially viable. Now, even with the cheaper traling edge memory chips having a 32k expansion board was out of the question.

 

In the end all boils down to the money. How can you make a 'useful' computer at the cheapest possible price ? Having a backplane with many cards to get a complete machine makes it too expensive - this was the demise of the S-100 bus based systems ... all the S-100 manufacturers went out of business quick after the "1977 Trinity" (PET, TRS-80, Apple II) was featured in BYTE magazine.

 

I was after my own computer since the early 1970s, before the microprocessor, and never could afford one until after the 1st gen microprocessors went to the junk bin (where I got my first 8080 from). I got the BYTE magazines from a relative and yes I followed everything from the beginning. And I ran the numbers on the early S-100 machines. Despite the individual cards looked cheap, especially those great 32 kByte RAM cards devoid of any actual RAM chips (only empty sockets), all the numbers added up quickly into mind boggling heights.

 

So, single board computers like the Apple-1 truly were a game changer. KIM-1 was a contemporary microcomputer to the Apple-1, but more of an educational toy for engineers who wanted to get into the microprocessor technology.

 

I understand it's hard for younger people who were not around at that time to understand the  revolution  which took place from 1975 to 1981, fueled by the microprocessor. The story of the microprocessor began earlier in the 1970s but it took a while until the semiconductor technology had matured to a point where personal computers became viable - useful enough and cheap enough. The Apple-1 would have been impossible to conceive in the year 1974. It was not until mid 1975 when all the bits and pieces were available. And lots of pioneers jumped on that bandwagon and took these bits and pieces to put their microcomputers together for the first time in human history. The same kind of entrepreneurial spirit as with the first 'safety bicycles', the first motorcycles, and the first automobiles. But as always, most of these pioneers got something wrong, and failed in the marketplace. And as always, in hindsight, we can point out the criteria for success because we know how the successful products whose companies survived were designed and which key features they had that made the success possible. In hindsight, which is 20:20, we know everything. Or at least pretend to do so.

 

- Uncle Bernie

Offline
Last seen: 10 hours 10 min ago
Joined: May 4 2021 - 06:35
Posts: 79
Criteria

Uncle Bernie,

I am responding to your comment on this thread because it is more appropriate, and BTW thanks for replying to my post regarding the Olivetti Programma 101.

I respectfully disagree with the criteria you mentioned (I know you didn't write them...), for example I don't understand why DRAM is acceptable and SRAM is not...

The "single card" constraint is also rather obscure and cannot be motivated by production costs alone, even for the production techniques of that period.

 

As already mentioned here, this criteria seems to be quite arbitrary and tailored to Apple's implementation...

Claudio

 

Toolkitman's picture
Offline
Last seen: 1 week 3 days ago
Joined: Nov 21 2010 - 15:30
Posts: 243
In 1973 was  built a tv

In 1973 was  built a tv teletypewriter by Don Lancaster but was not a computer more a typewriter on the monitor screen. The other computers like the apple 1 use a similar circuit (not equal) to this one.

Offline
Last seen: 10 hours 10 min ago
Joined: May 4 2021 - 06:35
Posts: 79
UncleBernie wrote:KIM-1 was a
UncleBernie wrote:
KIM-1 was a contemporary microcomputer to the Apple-1, but more of an educational toy for engineers who wanted to get into the microprocessor technology.
 

This is not entirely true.

KIM-1 has been widely used in automatic looms used in textiles, for example, as an evolution of the Jacquard loom.

:-)

 

Claudio.

Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
On the SRAM vs. DRAM criteria

To comment on p-lab's post #4 above:

 

About the SRAM vs. DRAM ciriteria please see my post #3 above. A 'personal computer' needs a useful amount of RAM and we could argue over what that amount might be, I think 4kByte is barely enough, 8kByte better, and at mid-1970s semiconductor process technology it is IMPOSSIBLE to provide this amount of RAM at an 'affordable' price other than using DRAM. The reason is that all MOS transistors at a given process node are the same minimum size, and DRAM uses one MOSFET per bit, while SRAM uses six MOSFETs per bit (specialized SRAM processes added high ohmic resistors in higher layers and could do SRAM with four MOSFETs per bit). Alas, the YIELD of any semiconductor IC decreases exponentially with increasing die size and the NUMBER of die per wafer decreases quadratically with increasing die size. So you can see that DRAM is the only way to make affordable RAM with the lowest possible die size, the highest number of die per wafer, and the highest yield (of good die).

 

SRAM in a line voltage powered microcomputer system only was competitive for relatively small amounts of RAM, where the overhead for the DRAM (refresh logic, address multiplexing, more complex timing of the control signals) was larger than the savings brought by the DRAM's lower cost per bit. This is true until today.

 

And the criteria certainly were not hand picked to match the Apple-1. They also describe the IBM PC, although a few generations of microprocessors and RAM ICs later. As I said, these are the key criteria for a commercially viable (read: cheapest possible price for the capabilities) 'personal computer'.

 

The argument of single board vs. multi board is hard to grasp for people who never designed an industrial, mass produced system, but the cost multipliers involved with multiple PCBs in one product are mind boggling and, for the end user price, catastrophic. Just look inside any mass produced consumer product, they all tried their best to do everything with only one PCB, if possible at all. The reason is that they want to stay competitive - if they use more than one PCB, and their competitor can make a copycat product with only one PCB, he can undercut their price and take the market share away.

 

Microcomputers always have been a cuthroat business with razor thin margins (Jack Tamiel put it best: "Business is war !") and you simply can't compete when you use solutions that cost more money. I know only of one even more cuthroat business in microelectronics, and this is making ICs for TV manufacturers. These TV manufacturers were suffering from a brutal price squeeze and so they wanted feature set ICs for the cheapest possible price. Guess which type of RAM we designed into these ICs ... this was 25 years ago.

 

- Uncle Bernie

 

 

macnoyd's picture
Offline
Last seen: 1 day 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 764
II would have given the KIM-1 the first place prize IF

I would have given the KIM-1 the first place prize if it had an ASCII keyboard interface, Tape interface option and a (close to) NTSC type video output.

The Apple 1 was way ahead in terms of the on-board archetecture and relatively useful programs that were released with it.

The KIM-1 was a great 6502 learning tool with a built-in HEX keyboard but didn't have the same purpose in-mind.  The Apple 1 was intended to grow home enthusiasts at the time and became the perfect stepping stone to the Apple II.  Took me a year to save for that first Apple II.  I already had two KIM-1's when I bought my Apple II.  If I remember correctly, the KIM-1's were around $200 apiece. (?)

Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
About early microprocessor 'evaluation systems' / pricing

In post #8, macnoyd wrote:

 

"If I remember correctly, the KIM-1's were around $200 apiece."

 

Uncle Bernie comments:

 

I think the price was a tad below $200 and you even could get one for free if you pretended to be a OEM who would make a product using the 6502. I know that because the dad of one of my peers at high school once brought home a brand new KIM-1 which had been given to the company he worked at for free, from a sales engineer. I can't remember if it was an original MOS Technology KIM-1 or one of the knockoffs made for the 2nd source manufacturers of the 6502 (and family). I can remember the KIM-1 had an on-board cassette interface. Which was very slow but worked reliably. The same dad also brought home an obsolete ASR-33 Teletype which readily interfaced to the KIM-1 --- I think we needed to rig a small PCB with a few transistors. When it came to life making hex dumps from the KIM-1 it was like a jackhammer being used nearby. Guess where I spent my afternoons instead of doing my homework - which of course always "had been eaten by the dog". Oh, the fond memories of the early days of microcomputers ... I spent most of my time outside these indoctrination / brainwashing classrooms being around and programming computers. And I nearly got kicked out of school for masterminding the "borrowing" of the janitor's key ring and making duplicate keys for the school and then the same for the key to the computer room. So we could sneak into the school building and the computer room very early in the morning and late in the evening, when the school doors were locked. This allowed me and my co-conspirators to have more computer time than the officially rationed two hours per week. Oh, and the computer was a WANG2200. We were caught when a teacher had reserved the computer a whole afternoon for himself and one of us unlocked the door using the duplicate key while the evil teacher was there. I recently used my key making skills to make a new key for my storm door, just by having the cylinder lock out, and no key. So don't tell me that in socialist government schools you don't learn useful skills other than smoking dope and how to fight with a switchblade knife. But I digress. Nice computer kid story. Back to the KIM-1, SYM-1, etc...

 

These so-called "evaluation systems" typically were sold at or below cost, or for some prospective - and promising - customers given away for free. The idea was to lure OEMs into adopting a new microprocessor family. Since at these giveaway prices, nobody could produce the thing at a lower cost, they often were bought in bulk and ended up as controllers in some industrial machinery. This is the example given by p-lab in his post #6. I once got a IMP-16 card that was pulled out from an early automotive diagnostics computer. This card had been the "evaluation system" for National Semiconductor's IMP-16 family dating back to 1973. The manufacturer of said automotive diagnostics computer just had replaced the firmware EPROMs with his own code.

 

- Uncle Bernie

Offline
Last seen: 2 days 53 min ago
Joined: Nov 24 2013 - 13:23
Posts: 226
p-lab wrote:UncleBernie wrote
p-lab wrote:
UncleBernie wrote:
KIM-1 was a contemporary microcomputer to the Apple-1, but more of an educational toy for engineers who wanted to get into the microprocessor technology.
 

This is not entirely true.

KIM-1 has been widely used in automatic looms used in textiles, for example, as an evolution of

I actually had a KIM-1 that I sold a while back that had been set up for running some kind of heating system.  It even had the instructions to start it up in the event of power loss. 

 

Offline
Last seen: 2 days 53 min ago
Joined: Nov 24 2013 - 13:23
Posts: 226
UncleBernie wrote:In post #2,
UncleBernie wrote:

In post #2, falter wrote:

 

"Why does it have to be a single motherboard ?"

 

Uncle Bernie answers:

As for the Sphere-1, this came very close to the concept, it even had DRAM, but it was a multi PCB machine and this made it expensive.

 

The Apple-1 sold for $666, which in today's money is $3500.  That's hardly cheap, to start with.  The basic Sphere system (not the complete one with chassis and monitor) could be had for $650, and at least as it reads in their ads, it also displayed on TV.   Plus it came with a keyboard, something the Apple didn't.

Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
About the Sphere-1

In post #11, falter wrote:

 

"The basic Sphere system (not the complete one with chassis and monitor) could be had for $650, and at least as it reads in their ads, it also displayed on TV."

 

Uncle Bernie comments:

 

Maybe this is the reason that Sphere computers went out of business, quick, never to be seen again. It's a mistake many fledgling first-time entrepreneurs make: they forget to factor in an adequate profit margin. Oh, and it happens to larger companies, too. I don't think that Atari ever made a profit with their 8-bit computer line. For the TI-99, we know it only made losses for TI --- despite they were able to produce all the ICs in-house. For the C-64, the profit situation is unfathomable because Commodore was a conglomerate of shell companies, some of which being mail drops in remote places like the Bahamas. From the outside the whole structure of Commodore looks like a tax evasion trick. None of the big corporations, if properly managed, ever pays U.S. taxes on their profits, because all the profits magically appear in overseas places. So how much profit did Commodore really make ? Irvin Gould ousted Jack Tramiel for a reason... and for a financier / money man, pulling the strings from behind the curtain, the only reason (other than personal animosities) to cut off the figurehead is lack of profits.

 

But back to the Sphere-1: it ticks most of the boxes of the 'personal computer' criteria but not all of them. In the end it flopped on the marketplace. It's a close contender, though, so far the best one which was brought up by the various contributors on this thread.

 

- Uncle Bernie

Offline
Last seen: 1 week 2 days ago
Joined: Oct 9 2011 - 12:54
Posts: 1351
I find all this funny as

I find all this funny as people talk about the 1st automobile and usually say Ford Model-T.  Which was the 1st mass-produced automobile.   But Daimler had been making cars for years already.

 

the problem claiming the Sphere as the 1st personal computer is that you can see a direct line from the Apple-1 to Apple II (and the holy trinity of 1977) to modern machines, you can't see that for the Sphere or even the 1974 Scelbi, which predates both Apple-1 and Sphere. Only 50 copies were ever made and who knows how many survive, in full disclosure, I have an original Scelbi 8h and it's one of my prized possessions because of how rare, and while mine works (with a replica memory board) unlike my Apple-1, I never power it on and only use my replica from Mike Willegal because unlike an Apple-1, I will never ever find another original Scelbi 8h for sale.

 

The Scelbi was a full computer based on the 8008 and using either a terminal or an oscilloscope display could run programs and not just blink lights like the 1st Altair systems.  They just sucked to turn on since you needed to toggle in a loader to run software.  I use a replica memory board in my original because the 1101 memory suck and finding replacement white ceramic ones are impossible that work and I have a 1702 EPROM hard wired in so that if I did want to run it, I don't have to toggle in a loader.  The 1702 EPROM is period correct and the plastic 1101 chips are from 1975, still before the Apple-1 and Sphere were generally available.

Offline
Last seen: 2 days 53 min ago
Joined: Nov 24 2013 - 13:23
Posts: 226
Sphere wasn't first

Oh, I wasn't suggesting Sphere was actually the first personal computer.  I was just taking the criteria UncleBernie presented and looking for something that possibly met it while being earlier than the Apple-1.  My own criteria for 'first personal computer' is a machine that was affordable, intended for and usable by a single person.  I think the established winner based on that is still the Kenbak-1.

We can drive ourselves crazy with semantics.  I was thinking about this the other day while pondering what the first 'video card' was or what the first 'graphics card' was.  All depends on how you define it.  I just feel like the criteria UB has presented is a little arbitrary and set up to favor the Apple-1.

Offline
Last seen: 10 hours 7 min ago
Joined: Apr 1 2020 - 16:46
Posts: 476
The 'criteria' boil down to a 'pattern' (it's evolution !)

To comment posts #13 and #14 from Corey986 and falter:

 

In the end, those criteria for the 'personal computer' boil down to a 'pattern' as it's called in many fields of engineering.

 

To establish a 'pattern' in this sense, there must be a 'lineage' (as in genetics) from the 'first one' to those generations who followed and had the same pattern / genetics. Technological progress may subsititute certain components with more modern ones, but the pattern still is visible. As for cars, the 'pattern' is a box with four wheels, a steering wheel for the driver,  and foot pedals for acceleration and braking (remember the third petal ?). Anyone who wants to sell a car deviating from this pattern, not ticking all the boxes, will fail in the marketplace. Look at some of the first motorcars, including the one from Carl Benz ... ridicolous steering contraptions and controls ... they did not establish the winning  'pattern' yet which would have led to more success in the marketplace. Commercially, they all were a failure. Innovators and Inventors and Pioneers they all were ... like most of the early microcomputer pioneers. But most of these early pioneers failed and their microcomputers were one trick ponies, never to establish a 'pattern' or a 'lineage'. They were dead ends in the (technical) evolution tree.

 

As Corey986 has so aptly observed in his post #13, there is a direct lineage from the Apple I to the Apple II, and as I have stated earlier, to the IBM PC of 1981, which, if you peek inside, and check the boxes of 'the criteria', stole all the Apple II concepts, as far as 'the criteria' go.

 

So it is fair to say, that from the Apple-1 there is a lineage of 'genetic traits' to the Apple II and the IBM PC, and the latter indisputably  is  a 'personal computer'. But the (winning) pattern is there already in the Apple-1.

 

We see much the same with self-loading (aka semiautomatic)  handguns, I'm also deep into their history, and if we look at their history, there were some early ones (in the 1890s) which were peculiar and awkward, but worked reliably, and were a commercial success (Mauser C-96 and the Luger aka P08). But it was John Moses Browning who established the (winning) pattern with his early designs for FN (the model 1900) and Colt. The famous 1911 was the final incarnation of Browning's pattern which he did all by himself. It's still being made today, and it's the only handgun which is still in production, with the same pattern, for more than 100 years. Every successful self-loading handgun ever produced had to follow that 'Browning' pattern to be commercial success. Those who deviated from the pattern, failed. It was Gaston Glock who in the early 1980s delivered a complete reinterpretation of the pattern (actually, it was a close cooperation between Mr. Glock and Gen. Major Friedrich Dechant of the Austrian Office for Defense Technology - "Amt fuer Wehrtechnik"). This new, improved pattern is now ubiquitous - after the patents expired everybody "stole" it. The old pattern is fading away - except for the iconic 1911.

 

The same kind of evolution happened with the 'personal computer'. It's now an outdated pattern, but it had had total market dominance for decades. You might have noted that you can't easily buy any newly made 'personal computer' according to this pattern anymore ... you have to build it from components you can order online. Everybody else buys notebooks (no building) or rack mounted 'blades' for data centers (the "Cloud"). And almost everybody carries around a small box in his/her pockets which has more memory and computing power than all the computers in the USA of the 1960s combined. And this little box can tap into the vast data oceans of the world wide web instantly, where almost all the knowledge of human civilisation is stored, ready for you, if you can find it. A famous prophet once said "the people will dress crazy and you can't tell who is a man and who is a woman. They all carry little boxes around and look at those all the time. The box can answer all questions. Then comes the war." --- this was in the 1950s. He was ridiculed.

 

- Uncle Bernie

Online
Last seen: 44 min 57 sec ago
Joined: Mar 1 2019 - 04:38
Posts: 73
Its a moot and ultimatley

Its a moot and ultimatley pointless question, but hey, history is interesting.

 

 

I did pick up this comment in my build thread on the VCFD forum from user Chuck which was interesting.

 

"I remember Woz and Jobs pitching the thing at a Homebrew meeting: $666.66 back then. Most of us already had systems, so interest in a non-expandable board was lukewarm. Later on (ca 1977, I remember Apple setting up in an office building at the front of Bubb Road in Cupertino. We were on the other side of the street at 10101. 

Around December of 1975, I bought the "special deal" Altair kit. 8K of DRAM, SIO and CPU all for a measly $1000. You still needed a terminal to do anything useful."

 

Log in or register to post comments