What spec would the A1200 had launched with if it was up to YOU

[Bruce Abbott]

Originally Posted by Samurai_Crow

I would rather have kept the PCMCIA for Ethernet controllers and wireless, eventually.

As for the SIMM socket for Fast RAM, not if it came at the expense of a belly-slot. Don't underestimate how cheap Commodore could get with base configurations. Having a Fast RAM card removable but standard in the belly-slot as stock would have been great enough.

Having both would be tricky anyway, since there was very little room left on the motherboard. You might get a trapdoor area too small to take much more than just the CPU.

Of course this wouldn't be such a problem in a larger case, as we saw with the A4000. But then you get the down side - this onboard RAM would be slower than a fast CPU, so you end up having to put RAM on the accelerator card anyway for best performance.

[abu_the_monkey]

for me I would have preferred the A1200 came with a HDD ~40mb minimum as standard, the cpu on a card in the belly slot with an option for fast memory on it rather than on the main board, and options for alternative configurations from the resellers.

if the memory could have been configured at start-up to 1mb chip and 1mb "real" fast (as in the cpu has full access al the time) it would have been a nice touch as a lot of older software didn't need 2mb of chip ram but would run better with the addition of fast ram.

[AestheticDebris]

Originally Posted by StevenJGore

My 120MB 2.5" HDD in 1993 cost £250. That was considered a good price at the time, and it included installation (i.e. taking my A1200 to the shop for it to be fitted).

Well yeah, but 2.5" hard drives were really expensive back then and the A1200 case even as shipped could just about fit a much cheaper 3.5" model. Making that fit properly, even if not available by default, would've made a big difference.

I do think a 1MB Fast, 1 MB Chip split would've worked better overall. I know Commodore supposedly asked coders what they wanted, but I think the final call to go that way given how much it slows computation was definitely a mistake.

[Bruce Abbott]

Originally Posted by eXeler0

The obvious push-back here is that it makes a lot of difference for devs whatever the machine can do out of the box as opposed to whatever is available in the after-market...

Yes, and the A1200 could do a lot more 'out of the box' than the A500 could.

But when the A1200 was released (after the A600 and everyone thought nothing else was coming) developers were still working on A500 games. They weren't going to drop what they were doing to work on AGA titles instead when there were hardly any users. So they had plenty of time to 'grow into' the A1200 before becoming concerned about whether it had enough power.

When we see what is being developed for the A500 today, it's obvious that the stock A1200's power was not fully utilized. Take Dread for example, it runs smooth as butter on a stock A1200 - my 50MHz 030 is too fast for it!

No PCMCIA,

Yes, this would save a few bucks.

But the PCMCIA slot turned out to be quite useful. The only problem was a lack of suitable cards. To fix this I would have produced a PCMCIA 'breakout' board to help developers (and hobbyists!) make products for it. I would also explain that Amiga PCMCIA cards didn't have to be PC compatible so they could be made much simpler and cheaper.

no built in RF-shield.

= no sales in the US.

Sell it with PSU sold separately. A500 owners could reuse the A500 PSU.

A500 owners usually either kept their old machine as a complete system or sold it. Even if they didn't, they might want that more powerful PSU when they got an accelerator card.

Let's say this spc would have added $150 to the price at launch but would probably drop fast a year later...

$150 is way too much. The A1200 had to be not much more than an A500 to get maximum sales volume. The most important thing was to grow the userbase as quickly as possible. That's why I believe getting it out sooner was more important than making it more powerful.

[eXeler0]

Originally Posted by Bruce Abbott

Yes, and the A1200 could do a lot more 'out of the box' than the A500 could.

But when the A1200 was released (after the A600 and everyone thought nothing else was coming) developers were still working on A500 games. They weren't going to drop what they were doing to work on AGA titles instead when there were hardly any users. So they had plenty of time to 'grow into' the A1200 before becoming concerned about whether it had enough power.

When we see what is being developed for the A500 today, it's obvious that the stock A1200's power was not fully utilized. Take Dread for example, it runs smooth as butter on a stock A1200 - my 50MHz 030 is too fast for it!

Yes, this would save a few bucks.

But the PCMCIA slot turned out to be quite useful. The only problem was a lack of suitable cards. To fix this I would have produced a PCMCIA 'breakout' board to help developers (and hobbyists!) make products for it. I would also explain that Amiga PCMCIA cards didn't have to be PC compatible so they could be made much simpler and cheaper.

= no sales in the US.

A500 owners usually either kept their old machine as a complete system or sold it. Even if they didn't, they might want that more powerful PSU when they got an accelerator card.

$150 is way too much. The A1200 had to be not much more than an A500 to get maximum sales volume. The most important thing was to grow the userbase as quickly as possible. That's why I believe getting it out sooner was more important than making it more powerful.

The topic was not "you are a time traveller and you go back in time and change stuff, so releasing things sooner isn't really an option in this scenario." ;-)

[TEG]

I would required a really usable VGA mode from the techies and market the A1200 around this new capacity. Cheap VGA monitors where flooding the market and it was now the norm.

The A1200 have minimal VGA capacity and was delivered with an adaptor, which show Commodore thought of it, but it was unusable. At the slightest opportunity, the Workbench was falling back in OCS resolution. And the mode was only in 4 colours (Productivity mode) and slow. So, yeah, a bit of fast RAM and a least 16 colours so 2 more bitplans.

I would have initiated a change in the politic too: Several configurations in stock available from the start. Thanks to the cpu on board. So you show your platform is sustainable for the futur because expandable like a PC. And you make money by selling the upgrade.

[Cris1997XX]

Ugh, when will all these what-if scenarios stop...?

[eXeler0]

Originally Posted by Cris1997XX

Ugh, when will all these what-if scenarios stop...?

Hey buddy, there are 50,000+ threads on EAB, you can probably find what interests you while staying away from those that dont. Easy, ey?

[khph_re]

C= has a tiny of money and time: 512k/1MB fast ram, bodged HD disk drive like A4000

C=Has a bit more money & Time: 512k/1MB fast ram, Bodged slow HD disk drive, buffered IDE for master and slave, increase case size as necessary to fit a third party CD ROM (optional purchase with machine, bundle demo disk and a couple of the best CDTV titles).

C= commodore some how has cash and time: As above-full speed HD floppy, remove PCMCIA,
expand clock port into modem slot. update custom chips, 16 bit sound, new sprites,increase copper speed somehow to get Low Res chunky etc. Increase ROM, OS boots from rom, put battery backed up storage RAM on the mother board (to save WB prefs, update libs, save games etc - mabe 128/256k). Pretty up the OS, make WB Asynch. Hope for the best it isn't to late. Maybe scan doubler for Amiga TV modes, MIDI ports.

Gt ST devs onboad - (it's not doing great, they have some great software, same processor should make it a bit easier) , to late to grab MAC most devs.

[lmimmfn]

Change chip RAM to VRAM to increase bandwidth, chunky mode(bypassing DMA slots for video), 68030/25MHz, 2 Meg Fastrsm, and room in the case/compatibility with 3.5Inch harddrives.
Charge £550 quid for that config snd it would have been a winner!

[Bruce Abbott]

Originally Posted by eXeler0

The topic was not "you are a time traveller and you go back in time and change stuff, so releasing things sooner isn't really an option in this scenario." ;-)

Originally Posted by eXeler0

Its 1991 and you are head of the development of the next Amiga models...

You didn't say when in 1991 I was head of development of the 'next Amiga models'.

From Commodore the Final Years:-

The AA taskforce

Earlier in the year, Irving Gould had expected the AA chipset to be done by the fall, with the release of the A1000 Plus computer to follow shortly after. The first revisions of the chip had been produced earlier in the year. But by early October, the chip was not ready for release. "AA was just sort of languishing," says Lew Eggelbrecht. "The design was done there were a lot of bugs". Bill Sydnes needed to deliver a computer soon. Rather than wait on the AA chipset, he altered the design specs for the A1000 Plus by replacing the AA chipset with ECS, something Jeff Porter had previously warned would be a terrible mistake...

Eggelbrecht, Sydnes and Jeff Frank eventually restarted AA development in October.

That's the 'terrible mistake' I would avoid if I was in charge of Amiga development in 1991. I would prioritize fixing the bugs in the AGA chipset. With luck the A1200 (or whatever it was called) would ship in late 1991.

In doing so I would simply be doing what Irving Gould was expecting. Gould was let down by the engineering department. After stating in an interview with Amiga Computing magazine in early 1991 that a new chipset was expected 'this Fall', they delivered nothing. Amiga fans were expecting it too. Imagine how differently we have felt about Commodore if the engineering department had gotten the job done instead of dashing our hopes and embarrassing Gould.

Sydnes' (or was it Eggelbrecht's) decision to switch the A1000 Plus to the ECS chipset was a bad one, 'forced' by the slow progress of AA. The main reason for that was that AA was just a backup in case AAA didn't get finished in time. 15 engineers were wasting their time on AAA. I would have switched the priorities around to put AA first, letting go of the 'deadwood' and just keeping the best engineers for the job. Ali would have been pleased with that cost cutting and greater commitment to completing AA on time.

Dave Haynie had his A3000+ prototype booting with an early version of the AA chipset in February 1991. It shouldn't have taken over a year to get it finished. All that fancy stuff they were trying to do with AAA could wait. Gould had all but publicly committed Commodore to getting something out in 1991. Why couldn't they?

Bill Sydnes was hired on May 1990 to head the PC engineering division, but was soon poking his nose into the Amiga side too. As we all know he was a PC guy with no knowledge or the Amiga or enthusiasm for it. Had I been in that role in 1991 it would have been a different story. I would get AA out on time. Gould would be happy, Ali would be happy, and Amiga fans would be happy.

You said,

One of these will be the A1200 aimed at the lower end of consumer market and it has a targeted release date of late 1992.
What do you do differently

My answer is that I would change the targeted release date to late 1991, a year earlier. It should have happened. With the right person in charge it would have.

[Promilus]

15 engineers were developing AAA because in the long run that was the only project which was revolutionary enough to maintain a technological lead in a future to come. AA and even AA+ couldn't do that. And since it was already a placeholder for bigger AAA which was not directly compatible with AA anyway (and only partially with ECS) there really was no actual need to move team from more perspective AAA into placeholder AA. If anything it was wrong to do AA in the first place instead of using all those resources to actually fix and finalize AAA on time. 68EC030 is no MMU and going even higher than already bottlenecked 14MHz. Obviously to meet the demand for memory access it only makes sense to get on-board fast ram as well. And that requires some additional logic. Some bigger PLD or enhancements for other chip inside chipset to do DRAM memory interface for 25MHz 030. With accelerators in trapdoor "fast ram" would be kind of a s"low ram" instead - just like on board fast is on A4000 because timings wouldn't suit even faster CPUs well. But that's beside the point because Amiga memory management can handle that fairly well anyway.

[trixster]

I would say ship it with the following:

1) keep 2MB chip and add some fast ram
2) a 20MB or 40MB HDD
3) a faster cpu (28mhz 020 maybe?)

ideally do all three but that would probably be too expensive for it to be viable as an A500 upgrade, so prioritse the first two. A Dec 1992 advert in AF from Power Computer listed an IBM 40MB IDE HDD at £200, so could C= have done a 20MB one for an extra 100-150, or a 40 for 150-200 extra?

[pixie]

I would remove CPU from the mainboard and provide it with a standard 68020 board with ram slots. Then I could introduce as standard at least with 1mb fast ram

[eXeler0]

Originally Posted by pixie

I would remove CPU from the mainboard and provide it with a standard 68020 board with ram slots. Then I could introduce as standard at least with 1mb fast ram

Ive been thinking along these lines too. The concept of no cpu on the mainboard was used on the A4000 which came before the A1200 after all…

[Wachi]

In those years. Anything that would have made a 1200 run DOOM and MP3s smoothly would have been decisive in computer sales.

[eXeler0]

Originally Posted by trixster

I would say ship it with the following:

1) keep 2MB chip and add some fast ram
2) a 20MB or 40MB HDD
3) a faster cpu (28mhz 020 maybe?)

ideally do all three but that would probably be too expensive for it to be viable as an A500 upgrade, so prioritse the first two. A Dec 1992 advert in AF from Power Computer listed an IBM 40MB IDE HDD at £200, so could C= have done a 20MB one for an extra 100-150, or a 40 for 150-200 extra?

Probably they should have made the A1200 case a wee bit larger so you could fit a 3.5” drive.
My first hdd in the A1200 was a 60 MB and it was fine, a friend had a 20MB and that was too small, even back then. 40MB is the absolute minimum.

[pixie]

Mine had a 3'5" HD, I think the problem back then was that not all hd were supported driver wise

[Tigerskunk]

Same specs, but two years earlier...

[kremiso]

plus, Muzza Turrican2 AGA released back then