Firstly the C64 was not a 'money is no object' project, the engineers were as frugal as the spectrum engineering team, but they worked for a semiconductor plant,
ooo, I disagree here. I read somewhere (maybe even an excerpt from your book!) that the commodore engineers would just make up another set of masks, fire off another sample run, test the chips and repeat until it worked. This is very, very expensive. Companies that do not own their own fabs make sure it works before they buy a new set of masks and make the chip. The C64's pricetag never reflected the actual costs of development that any other company would have incurred (it seems the chip fab was basically free) although Commodore itself would have lost $ in opportunity costs, assuming their fab wasn't underutilized. If the latter case is true then Commodore could have been a better run company :-)
In any case, the C64's special chips were originally destined for arcade machines but since that fell through the chips were rescued and stuck in a home computer in a hasty manner. What I mean by hasty is that the C64 is missing what the other home computers had -- a decent version of basic for home users, which is what the market was about at that time. Only later was it about the games and the C64 was very well positioned for that.
so they could design custom chips unlike Sinclair.
Sinclair designed its own custom chip as well, called the ULA. The ULA is a sea-of-gates semi-custom chip where the customer specifies the last metal layers to connect the already laid out transistors by the manufacturer. The manufacturer can inexpensively make a whole bunch of almost complete chips with regularly laid out transistors and customers like Sinclair can buy a batch, specifying the connections in order to make a custom circuit. The only disadvantage to a fully custom ASIC is perhaps in speed and density and cost once volume reached a certain large minimum level. Adding sprite hardware to a sea-of-gates chip wouldn't have been a problem, except in extended development costs. The SID is a bit different though, with its partially analog circuitry. Nevertheless the sea-of-gates chip option was an economical alternative to a fully custom ASIC, economic being something the Commodore engineers didn't have to worry about :-)
As for the Spectrum, they didn't have to squeeze, it is near the minimum you would need to make a computer, a CPU with a colour framebuffer. The ULA integrates some stuff that could have been done with some logic chips. They did however do a fantastic job on the Spectrum, but the level of integration isn't as amazing as you make out. If you made the C64 without those custom chips, it would have been a huge machine, so it shows just how much they squeezed into those chips.
Yes, the Spectrum is a simple computer but the integration of everything possible onto a single IC is what made it economical. Sinclair was the *first* company to do this with its ZX81 which is what made it possible to sell those computers for under $100 when others were still costing $1000+. Sinclair was about getting educational computers into the home, always making sure a decent manual and a good version of basic was present. When the home computer market changed from hobbyist to games machine it was only by pure accident that the Spectrum was fast enough to have quality games written for it. Had that accident not been possible it would have died commercially to the C64 in 1984 or so.
The C64 is a much more even machine than the beeb, so it is in no way FAR superior. The C64's price dropped a lot, while the beebs did not drop anywhere as near, which makes the price bracket argument moot.
As you say it depends on what you think makes a good computer. Personally, I didn't find the C64 all that interesting -- it is very slow (the slowest of all the 8-bits on the market), didn't have a z80 (yes this was a buying point for me as all the real computers had z80s), a poor version of basic (when I purchased I was a young kid interesting in learning computers), and the absolutely worst disk system out there (the slowest *AND* the most expensive at the same time, more expensive than the computer itself).
However, don't get me wrong -- I do enjoy the odd C64 game and had many friends with C64s as a youth (check out which side of the Atlantic I grew up on ;) ). As a games machine it was a very good piece of kit. The SID and VIC made sure of that.
I've never owned a BBC, and only used one at school, so I can't comment re: BBC vs C64, but one thing I have to correct you on is where you say that the BBC is known only for Elite (at least as far as games go). This is not true at all, as the BBC is till remembered for classics like Revs, Repton Mania and the Sentinel. I know this from the number of times I've read about the BBC, or discussed the 8-bit days with ex-BBC owners.
Yes I've owned a BBC and they were OK computers but like the C64 let down by a hideous case design.
The C64 was known as the "blocky computer" at our school, and with reason, all the games were blocky. They became more popular a bit later on 1988/89 as Amstrad lost interest in the Spectrum to the extent of just releasing the +3 ( which I owned, and loved ). The ratio amongst my peer group was 5-1 Spectrum:C64 and 3-1 Spectrum:BBC/Electron and 7:1 Spectrum:CPC. But instead of migrating to C64s, people migrated to Atari ST, and then Amiga and Sega Genesis.
The archimedes, when it arrived, was little short of a masterstroke. Zarch demo, played that for hours.
The Beeb is a very nice computer, but you are comparing two different machines. The beeb was designed to be an education/hobbist machine, that is why it had also those "important" features you mention. If you are considering it as home games machine, then it doesn't have all those "important" features. Its only superior depending on the role you want the computer to perform. As a games machine, it sucks. Elite was the only good game on the beeb and the only one it is known for. The C64 is a much more even machine than the beeb, so it is in no way FAR superior. The C64's price dropped a lot, while the beebs did not drop anywhere as near, which makes the price bracket argument moot.
I'll be the first one to admit being biased against the C64 - in my opinion, for the price they charged for it, it wasn't all that great. Since Commodore owned the semiconductor plant, for the price of a BBC Micro they should have had a vastly superior machine. But it wasn't. Hardware wise it was pretty uninspiring.
But you have to either have amnesia or be kidding if you think Elite was the only notable game on the BBC. The BBC Micro was a very good all round machine INCLUDING games: great games that originated on the BBC *other* than Elite include Revs, The Sentinel, Castle Quest, Repton - the list goes on. Many games were ported to the platform - favorites like Jetpac, Atic Atac, Knight Lore etc. There were a lot of very inventive games for the BBC. There were plenty of crap ones, too :-)
I even wrote one - I wrote a MUD for the BBC Micro. It was about the only 8-bit machine it was practical to write a MUD on because the BBC Micro was really the only machine to have a low-cost and ubiquitous LAN (IIRC, adding Econet was a matter of adding a single chip to the machine). The BBC Micro also fostered development - not only did it arguably have the best BASIC interpreter of the lot, but it also had a built in 6502 assembler, making the barrier to entry much lower than on competing machines.
I'd guess the thing with the BBC Micro was that it never quite sold enough to support a proper games industry in the same sense that the Spectrum, C64 and - to a lesser extent - the Amstrad CPC did. As such it lost out in terms of the sheer quantity of games as well as attracting few license deals and conversions.
Still, in terms of quality and originality it was up there with the best. You can add Thrust, Firetrack, Exile, Starship Command, and Twin Kingdom Valley to the list of games that started out on the BBC Micro. I dare say that I still play them a lot more than any of the old C64 games...
Thats right, and its image was that the people that bought one wouldn't really want their offspring to lower themselves to playing games on it ( at least, not for more than a few minutes before they went back to learning how to make money on the stock exchange from Pater on it ).
Plus the Electron, the one designed at a price point for kids to play games on, really wasn't up to much.
Of all the micros out there, pound for pound, you got most from the Spectrum until the Atari ST came along.
. If you made the C64 without those custom chips, it would have been a huge machine, so it shows just how much they squeezed into those chips.
Not really, it would have been a C16, or a VIC 20 with 64k. Or maybe just a lot cheaper as they might have used something off the shelf.
The custom chips weren't so revolutionary, they are more evolutionary from the work that the main designer did on an Atari that was released before it...
On another subject, wasn't there a computer released in Japan called the BabyMaker? Or was that some weird sex toy?
Comments
ooo, I disagree here. I read somewhere (maybe even an excerpt from your book!) that the commodore engineers would just make up another set of masks, fire off another sample run, test the chips and repeat until it worked. This is very, very expensive. Companies that do not own their own fabs make sure it works before they buy a new set of masks and make the chip. The C64's pricetag never reflected the actual costs of development that any other company would have incurred (it seems the chip fab was basically free) although Commodore itself would have lost $ in opportunity costs, assuming their fab wasn't underutilized. If the latter case is true then Commodore could have been a better run company :-)
In any case, the C64's special chips were originally destined for arcade machines but since that fell through the chips were rescued and stuck in a home computer in a hasty manner. What I mean by hasty is that the C64 is missing what the other home computers had -- a decent version of basic for home users, which is what the market was about at that time. Only later was it about the games and the C64 was very well positioned for that.
Sinclair designed its own custom chip as well, called the ULA. The ULA is a sea-of-gates semi-custom chip where the customer specifies the last metal layers to connect the already laid out transistors by the manufacturer. The manufacturer can inexpensively make a whole bunch of almost complete chips with regularly laid out transistors and customers like Sinclair can buy a batch, specifying the connections in order to make a custom circuit. The only disadvantage to a fully custom ASIC is perhaps in speed and density and cost once volume reached a certain large minimum level. Adding sprite hardware to a sea-of-gates chip wouldn't have been a problem, except in extended development costs. The SID is a bit different though, with its partially analog circuitry. Nevertheless the sea-of-gates chip option was an economical alternative to a fully custom ASIC, economic being something the Commodore engineers didn't have to worry about :-)
Yes, the Spectrum is a simple computer but the integration of everything possible onto a single IC is what made it economical. Sinclair was the *first* company to do this with its ZX81 which is what made it possible to sell those computers for under $100 when others were still costing $1000+. Sinclair was about getting educational computers into the home, always making sure a decent manual and a good version of basic was present. When the home computer market changed from hobbyist to games machine it was only by pure accident that the Spectrum was fast enough to have quality games written for it. Had that accident not been possible it would have died commercially to the C64 in 1984 or so.
As you say it depends on what you think makes a good computer. Personally, I didn't find the C64 all that interesting -- it is very slow (the slowest of all the 8-bits on the market), didn't have a z80 (yes this was a buying point for me as all the real computers had z80s), a poor version of basic (when I purchased I was a young kid interesting in learning computers), and the absolutely worst disk system out there (the slowest *AND* the most expensive at the same time, more expensive than the computer itself).
However, don't get me wrong -- I do enjoy the odd C64 game and had many friends with C64s as a youth (check out which side of the Atlantic I grew up on ;) ). As a games machine it was a very good piece of kit. The SID and VIC made sure of that.
Write games in C using Z88DK and SP1
Yes I've owned a BBC and they were OK computers but like the C64 let down by a hideous case design.
The C64 was known as the "blocky computer" at our school, and with reason, all the games were blocky. They became more popular a bit later on 1988/89 as Amstrad lost interest in the Spectrum to the extent of just releasing the +3 ( which I owned, and loved ). The ratio amongst my peer group was 5-1 Spectrum:C64 and 3-1 Spectrum:BBC/Electron and 7:1 Spectrum:CPC. But instead of migrating to C64s, people migrated to Atari ST, and then Amiga and Sega Genesis.
The archimedes, when it arrived, was little short of a masterstroke. Zarch demo, played that for hours.
I'll be the first one to admit being biased against the C64 - in my opinion, for the price they charged for it, it wasn't all that great. Since Commodore owned the semiconductor plant, for the price of a BBC Micro they should have had a vastly superior machine. But it wasn't. Hardware wise it was pretty uninspiring.
But you have to either have amnesia or be kidding if you think Elite was the only notable game on the BBC. The BBC Micro was a very good all round machine INCLUDING games: great games that originated on the BBC *other* than Elite include Revs, The Sentinel, Castle Quest, Repton - the list goes on. Many games were ported to the platform - favorites like Jetpac, Atic Atac, Knight Lore etc. There were a lot of very inventive games for the BBC. There were plenty of crap ones, too :-)
I even wrote one - I wrote a MUD for the BBC Micro. It was about the only 8-bit machine it was practical to write a MUD on because the BBC Micro was really the only machine to have a low-cost and ubiquitous LAN (IIRC, adding Econet was a matter of adding a single chip to the machine). The BBC Micro also fostered development - not only did it arguably have the best BASIC interpreter of the lot, but it also had a built in 6502 assembler, making the barrier to entry much lower than on competing machines.
I'm another BB, not the BB in question. :D
Still, in terms of quality and originality it was up there with the best. You can add Thrust, Firetrack, Exile, Starship Command, and Twin Kingdom Valley to the list of games that started out on the BBC Micro. I dare say that I still play them a lot more than any of the old C64 games...
Plus the Electron, the one designed at a price point for kids to play games on, really wasn't up to much.
Of all the micros out there, pound for pound, you got most from the Spectrum until the Atari ST came along.
Not really, it would have been a C16, or a VIC 20 with 64k. Or maybe just a lot cheaper as they might have used something off the shelf.
The custom chips weren't so revolutionary, they are more evolutionary from the work that the main designer did on an Atari that was released before it...
On another subject, wasn't there a computer released in Japan called the BabyMaker? Or was that some weird sex toy?
I don't think that's what he means. He's saying a discrete logic equivalent of the SID and VIC would have been huge, as in a lot of chips :)
Write games in C using Z88DK and SP1