Was anyone else disappointed with the A1200?

[Spongehorn]

Thread seems to have veered off into arguments, but to answer the original question, YES. I loved mine! And dearly wish I still had it now! Lovely design, just gorgeous. First computer I ever had with a hard drive. First computer I got into BBS's with... Just loved it and if I wasn't utterly broke I'd be trying to procure another right now! Emulators are great but give me real hardware any day!

[Bruce Abbott]

Originally Posted by grond

There never was an 040 at 50MHz. Also, those CPUs were ridiculously expensive. I think the real problem wasn't the high end but rather sticking to the 7MHz 68000 and 16 bits for seven years.

I think the real problem was the cost of more powerful CPUs (and associated harware) during that period. But this 'problem' turned out to be an asset for Amiga owners, because it meant they didn't have to buy a new machine every 2 years.

Many of my friends had A2000s with accelerator cards, and some had A500s with expansion units. But they could still switch them off for high compatibility when they wanted to play games. In comparision, PCs of the day were a nightmare to get games running on - you never knew whether your machine would be compatible or not, even after playing with system settings etc.

Ironically, one of the major disappointments of the A1200 was a lack of compatibility with A500 titles. Would this have been less of a problem if Commodore had released an 020 based system earlier? I think not. I bought an A3000 in 1991 and just had to suck it up if a game wouldn't work (had no other Amiga to play games on because some prick stole my A1000). If I was an ardent gamer that would have been a big problem.

From the start the biggest problem with the Amiga was a lack of installed userbase, and fracturing it into machines with different capabilities would not make it any better. We see the problem with other 'enhanced' home computers in that era - eg. C128, CoCo 3, Apple II GS, Spectrum 128, Amstrad CPC Plus. Few titles were produced that made full use of them because the market was too small. Only PCs and Macs escaped it, mainly because their primary focus was businesses (who could afford to upgrade their hardware and software regularly).

Truth is, we were lucky that the Amiga's designers decided to go with the much more expensive 68000 rather than use a 6502 variant, which would probably have been outdated in 2-3 years. That 7 years of 68000 based Amigas gave developers a powerful and stable platform spec to work with, which wasn't found wanting until 386-SX PCs hit the market in the early 90's. So the A1200 was actually released at about the right time.

[AmigaHope]

Originally Posted by Bruce Abbott

I think the real problem was the cost of more powerful CPUs (and associated harware) during that period. But this 'problem' turned out to be an asset for Amiga owners, because it meant they didn't have to buy a new machine every 2 years.

2 years is too short a cycle, 4-5 years is much more reasonable. 7 years is way too long.

In comparision, PCs of the day were a nightmare to get games running on - you never knew whether your machine would be compatible or not, even after playing with system settings etc.

That has a lot to do with the crappy all-over-the-place design of IBM upgrades/clones in the early/mid-80's. Once they got hardware compatibility ironed out (base VGA-then-VESA/Soundblaster/XMS) and once developers stopped using CPU loops for timing, stuff stayed more or less compatible. Then it mostly became a matter of "your machine must be this fast to run this" which held out fine until OS-friendly software became the norm.

Ironically, one of the major disappointments of the A1200 was a lack of compatibility with A500 titles. Would this have been less of a problem if Commodore had released an 020 based system earlier? I think not. I bought an A3000 in 1991 and just had to suck it up if a game wouldn't work (had no other Amiga to play games on because some prick stole my A1000). If I was an ardent gamer that would have been a big problem.

If they had released an *affordable* 020 machine earlier it would have helped a lot. The A2500 and A3000 didn't make much of a difference because they were only targeted at the professional market. The A500+ (with a 68020 and maybe a new name) and A3000 should have shipped together in 1989 or 1990. By then the 68020 was a budget CPU.

Compatibility was not really a huge problem on the 020. 80% of stuff worked fine with the caches turned off. (Can confirm as my main machine was an A500 w/68020). This would mostly be a mid-cycle refresh with some games having zero enhancements while others would, with only some requiring the new hardware.

From the start the biggest problem with the Amiga was a lack of installed userbase, and fracturing it into machines with different capabilities would not make it any better. We see the problem with other 'enhanced' home computers in that era - eg. C128, CoCo 3, Apple II GS, Spectrum 128, Amstrad CPC Plus. Few titles were produced that made full use of them because the market was too small.

These are bad comparisons as the only real upgrade there is the IIgs, and it suffered because it was competing with a platform from its own company (the Mac). The IIgs actually had some good releases right when it came out but the writing was on the wall by the time the Mac II came out.

The C128 was a shit upgrade because its best feature (the faster CPU) meant turning off most of the graphical capabilities. Otherwise it turned into a really expensive RAM expansion. The Spectrum 128 was also basically an expensive RAM expansion. In general these setups don't get good support unless there's also an easier/cheaper way to expand the previous model with that RAM.

The Amstrad Plus and CoCo 3 were modest upgrades but their markets were already dead.

A better comparison is the upgrade from MSX to MSX 2, which was hugely successful (albeit not including a CPU upgrade, it did work as a RAM expansion plus huge video upgrade)

Truth is, we were lucky that the Amiga's designers decided to go with the much more expensive 68000 rather than use a 6502 variant, which would probably have been outdated in 2-3 years. That 7 years of 68000 based Amigas gave developers a powerful and stable platform spec to work with, which wasn't found wanting until 386-SX PCs hit the market in the early 90's. So the A1200 was actually released at about the right time.

It was already found wanting as soon as the Megadrive had been released.

The A1200 was released at about the right time for a much better, faster system than the A1200 was.

[grond]

Originally Posted by Bruce Abbott

I think the real problem was the cost of more powerful CPUs (and associated harware) during that period. But this 'problem' turned out to be an asset for Amiga owners, because it meant they didn't have to buy a new machine every 2 years.

An asset for Amiga owners means that it was just the opposite for Commodore: their customers had no reason to give them more money because they didn't offer enough of an upgrade.

Ironically, one of the major disappointments of the A1200 was a lack of compatibility with A500 titles. Would this have been less of a problem if Commodore had released an 020 based system earlier?

Yes, of course it would have! A500 games breaking because of the Kickstart version not being 1.2 stopped soon after 1.3 was introduced. If an 020/ECS-based successor of the A500 had been in the market together with the A3000, programmers would have adapted to it earlier when the game catalogue was still smaller.

[AJCopland]

Everyone talks about A1200 incompatibility but I've never actually encountered any

[d4rk3lf]

Originally Posted by AJCopland

Everyone talks about A1200 incompatibility but I've never actually encountered any

There is some (I encountered Moonstone, for example), but it's way exaggerated then it really was.
Maybe 5 to 10% games haven't worked, but 90% of these games you could make work with various boot options.

Magazines are also to blame here, because, back in the day, some of them writes numbers like 30-40-50% of incompatibility.

[Gorf]

Originally Posted by Bruce Abbott

Ironically, one of the major disappointments of the A1200 was a lack of compatibility with A500 titles. Would this have been less of a problem if Commodore had released an 020 based system earlier? I think not. I bought an A3000 in 1991 and just had to suck it up if a game wouldn't work (had no other Amiga to play games on because some prick stole my A1000). If I was an ardent gamer that would have been a big problem.

Yes it would have been less of a problem!
Commodore should have upgraded the Kickstart ROMs on a yearly basis ... no need to improve much, but they should have changed the locations of the Libs every time on purpose, so developers would learn NOT to use direct addresses..

It would have been a great help if they would have switched to 14Mhz 68K in 1989 on A500/A2000 ... (of course with optional turtle mode). So developers would learn not to use CPU cycles as timing ...

Same goes for the Blitter, which schould have been gradually upgraded to make use of faster RAM (and later wider bus)...

Yes all that would have helped, to educate developers, so the jump to AGA or AAA would not brake much..

And it would have helped to keep the PC at distance all the time!

Few titles were produced that made full use of them because the market was too small.

Here is the Atari ST to blame ...

as I mentioned somewhere earlier: the world was just not big enough for both..

[grond]

Originally Posted by Gorf

Yes it would have been less of a problem!
Commodore should have upgraded the Kickstart ROMs on a yearly basis ... no need to improve much, but they should have changed the locations of the Libs every time on purpose, so developers would learn NOT to use direct addresses..

In hindsight they could have randomised addresses even for one and the same kickstart version for different production lots...

[AmigaHope]

Originally Posted by grond

In hindsight they could have randomised addresses even for one and the same kickstart version for different production lots...

It amazed me how many early developers jumped willy-nilly into the 1.2 ROM even though they already knew that changes had already happened from 1.0 and 1.1. I guess they just never even considered that the A1000 existed and were too set in their ways from the C64 where the ROMs never changed.

1.3 was a big wakeup call and almost everything written after 1.3 came out works fine on 2.0 and up.

[Kyle_Human]

Originally Posted by Gorf

So you are telling us now, that Jay Miner was right out lying about his work?
And everyone else working with him too?

All the Los Gatos people ever said is they were working on something called a ranger. That they were, an A1000 chipset machine in a tower with slots, bearing no resemblance to the rumour version. This still exists and it's whereabouts are known, with the owner having put pictures up of it online.

Originally Posted by Gorf

sure ...
what "improvements" would that be?

please stop trolling - seriously!

A500 brought a lot of improvements.

If you look at A1000, it's across not only a motherboard but also a daughterboard, and has lots of 74xx and PALs everywhere. A500 compressed this down to one board by replacing all that glue logic with Gary chip, making the whole computer half the price. It also gave Angus the 512K slow ram capability. Those are the two most famous improvements, but they're worth paying particular attention because the Los Gatos people were actually against doing them. Go figure.

People think the original Amiga design team were some geniuses who would've lead you to the promised land if they weren't betrayed by evil commodore. They think that they're supergeniuses like tony stark or something. Really they were just average joes who had some ideas before other people did. Once they'd implemented those ideas they were no better than anyone else, they were mostly just young people still learning. Which is why the original chipset wasn't even finished by Los Gatos people.

I mean seriously they were trying to put the entire analog video circuit inside Denise. Los Gatos weed hookup must have been amazing.

[Gorf]

I just linked to the interview with Jay Miner here in this thread.
It does not matter if he called it "ranger" or not - it was the next generation of chips he designed.
Your "1000 tower" is nonsense... that is not what we are talking about.

The A500 has a lot of cost reductions, but no technical improvements over the A1000.
The "slow ram capability" was the worst decision ever made, combing two disadvantages.

The A1000 "daughter board" is just RAM and solely commodores fault, since they wanted a 128kb version, not accepting that a bit mapped GUI needs lots of RAM und also not accepting, that the OS was not ready for ROM yet.
(Btw: the first Atari STs from 85 also had no ROM and needed a "kickstart" TOS disk...)

All the other cost reductions could have been done for the A1000 as well .. making it a affordable desktop...
the A2000 benefited from the same cost reductions... but was sold at a much higher price, for no real reason ... making it dead in the water.

As for the glue logic an Gary: of course they had nothing ready at Los Gatos, as they could not know they where going to use Commodores CIAs as interface chips ...
So they designed the clue logic with PALs and integrated it later into Gary

Commodore Germany did exactly the same for the Buster - first discrete logic and later integrated into one chip ... Dave Haynie did the same again for the A3000 and many CPU boards...

[chb]

Originally Posted by Kyle_Human

I mean seriously they were trying to put the entire analog video circuit inside Denise.

AFAIU they were planning to put NTSC composite signal generation on-chip, which is in fact a very reasonable choice for a game console. The PPU in the NES/Famicom has analog composite out, the VIC-II in the C64 has analog Y/C outs...

Or do you want to insinuate they wanted to do RF modulation on-chip? Then please give a source.

Originally Posted by Kyle_Human

Los Gatos weed hookup must have been amazing.

Projection, projection, projection...

[Kyle_Human]

Originally Posted by Gorf

Your "1000 tower" is nonsense... that is not what we are talking about.

.
It's ranger. That's what ranger is. There is no other. Not even a shadow of one, just imaginations.

Originally Posted by Gorf

The A500 has a lot of cost reductions, but no technical improvements over the A1000. The "slow ram capability" was the worst decision ever made, combing two disadvantages.

Cost reduction is in itself a technical improvement by definition, and slow ram was proven a very good idea. It opened the doorway to higher chipram capable agnes as a plugin upgrade, the connections are already there. Without the slowram revision the A2000 would have been permanently with only 512k chipram. Good luck video toasting with that.

Originally Posted by Gorf

The A1000 "daughter board" it just RAM and solely commodores fault, since they wanted a 128kb version, not accepting that a bit mapped GUI needs lots of RAM

The daughterboard isn't for system ram, and it isn't because of some mythical 128K version of the A1000. The reason that the WCS lives on a daughterboard is the A1000 glue takes up too much space.

It's wrong to say Los Gatos didn't design Gary because of not knowing about CIA chips, because at the time Commodore bought Amiga the chipset was nowhere near finished, and Commodore themselves finished it. Before Commodore there was only TTL breadboards, no chips.

Originally Posted by jay miner

Commodore was very good for AMIGA in the beginning. They made many improvements in the chips. Commodore made a lot of improvements in the things that we wanted but we did not have the resources to accomplish. The AMIGA originally only had three hundred and twenty colours across the screen, even in the six forty mode. They helped us put in full colour in the six forty mode. They also improved the colour by moving the NTSC converter off the chip.

[Gorf]

Originally Posted by Kyle_Human

.
It's ranger. That's what ranger is. There is no other. Not even a shadow of one, just imaginations.

No it's not ... only your "A1000 tower" is imagination.

When we talk about the "ranger chipset", we are talking about the last chips designed by Jay Miner. Here AGAIN his interview from 1988:
http://www.amigahistory.plus.com/jayminerinterview.html

Cost reduction is in itself a technical improvement by definition, and slow ram was proven a very good idea.

No it definitely has proven to be a terrible idea. For obvious reasons.
Making it real FastRAM would have cost only cents.
Making it real ChipRAM would have added no costs to the board at all (as the A500plus proves)

It opened the doorway to higher chipram capable agnes as a plugin upgrade, the connections are already there.

They had TWO YEARS to modify Agnus (NOT Agnes) to support more than 512KB, but did not manage do so...
So calling Commodore beging its typical incapable self "opening the doorway" is ridiculous.

Without the slowram revision the A2000 would have been permanently with only 512k chipram. Good luck video toasting with that.

??? How did slow ram help with video editing?

The daughterboard isn't for system ram, and it isn't because of some mythical 128K version of the A1000.

Really: please stop telling all that nonsense here.
It is not so hard to get some correct informations about the history of the Amiga. I really don't understand, what your goal is, bringing up all that fake news..

There ist nothing "mystical" about a 128kB version - that is what Commodore was seeing in the original Apple Mac 1984. So management thought: if it works for Apple, it will work for us.
Not recognizing that the original Mac was close to useless because of the lack of RAM (it got upgraded to 512KB in 85).
And not understanding that colour needs more RAM than just one monochrome bitplane ...

So the A1000 case and motherboard size was not planned big enough.
It is actually smaller than the A500 board.
(The A1000 is wider than it is deep ... so it would have been no problem to make it more square and gain some space...)

Also the space under the Floppy-Drive is almost empty in the original board revision ... later that space is used to get rid of the piggyback-board.

The reason that the WCS lives on a daughterboard is the A1000 glue takes up too much space.

no - is just because it was not planned to hold that amount of RAM.
(See above)
Or to have a "WCS" at all, since that was planned as ROM ...

It's wrong to say Los Gatos didn't design Gary because of not knowing about CIA chips, because at the time Commodore bought Amiga the chipset was nowhere near finished, and Commodore themselves finished it. Before Commodore there was only TTL breadboards, no chips.

Again: please stop spreading this nonsense here.
just some posts above you were told, that Amiga managed to integrate it's chips already before the purchase.

These chips where manufactured by a different manufacturer (Synertek) and where demonstrated in public.

[AmigaHope]

When Brian photographed the Ranger it seems he only got good photos of the backplane and of the "Garganturam" card (presumably a large fast ram expansion). There were two other cards plugged into the backplane that I can't find good photos of, but they seem to have more interesting chips on them that I can't identify due to not enough resolution/bad angle.

This is the best shot I could find of the two cards (the garganturam is the third one in the back)

http://obligement.free.fr/gfx2/ranger_cartes_zorro.jpg

[sandruzzo]

Even with few money they could have pumped ECS/OCS a lot and bring as somenthig a lot better than Aga.

They loved more share bonus, than working on Next Amiga... Come on, better and faster blitter and copper, and Paula of course wasn't so difficult to do

[Gorf]

Originally Posted by sandruzzo

Even with few money they could have pumped ECS/OCS a lot and bring as somenthig a lot better than Aga.

They loved more share bonus, than working on Next Amiga... Come on, better and faster blitter and copper, and Paula of course wasn't so difficult to do

Exactly!

and so we finally can come back to the thread: and that is why I was disappointed with the A1200 (or AGA in genreal - but this goes hand in hand).

[eXeler0]

Originally Posted by sandruzzo

Even with few money they could have pumped ECS/OCS a lot and bring as somenthig a lot better than Aga.

They loved more share bonus, than working on Next Amiga... Come on, better and faster blitter and copper, and Paula of course wasn't so difficult to do

This is a the part of the story I don't think I've heard too much about. We've all heard that AAA was taking too long and marketing ppl got tired of waiting so AA /AGA was done, but I've never heard an explanation from an engineer involved in the process about the reason why they couldn't fix a seemingly easy thing as doubling the 16-bit data transfer to 32-bit with Copper & Blitter (I would have forgiven them same old Paula ;-)

[sandruzzo]

@eXeler0

I think the overestimate themselves and underestimate others' capabilities. If the add full 32bit 14mhz support for both blitter and copper, better ram access and new paula, you'lle get a real powerfull Amigas' chipset with easy.

[grond]

I think Commodore assigned ridiculously small engineering teams to product development. They started lots of projects only to cancel them. For some they cancelled engineers were even threatened to get fired if they were found working on them in secrecy. These stories always sounded to me like there was a horrible lack of management, vision and leadership in Commodore. They clearly had the means to do much better.