was there a reason why atari limited the memory size of 2600 games

[kool kitty89]

By 1983 we had NES quality hardware on the market.

Do you mean the knowledge that NES quality hardware existed in the world or that people in the USA could run out and buy their own NES? For any young newbies out there reading this thread who may not know, the NES wasn't available nationwide in the USA until 1986

 

I think he meant the hardware was out at the consumer level by this time, the Famicom launched in mid 1983, granted it took a while for them to get it to the US. (due in part to the delays with negotiations with Warner/Atari, and the subsequent gaming crash in the US)

 

I was thinking just a plain vanilla flavoured 2600 but with a 6502 and 64k games (or 48k even)

 

They could have done that in 1980 or 81. That would have been a wise move, probably, however it would have alienated the original installed base of 2600 users. Remember that videogames were really just out of the starting gate by then. Even the 2600 was a big purchase decision. People were not really ready for the hardware upgrade treadmill which is so routine today. If they were going to offer a next gen system it was going to have to be truly next gen and we all know how that played out.

 

The main advantages of a 6502 (or Atari's custom 6502C) would be the higher clock speed and interrupt capabilities, addressing really isn't that big of a deal for mass produced games with bank switching. (as was mentioned previously) I think they'd have to modify things somewhat to actually take advantage of the faster clock speed (so the 6502 wouldn't have to drop to 1.19 MHz as it did in the 7800). Without additional onboard RAM and with the CPU still driving the display, it would still be pretty limited. If you really wanted an updated system that was heavily derived from the 2600 hardware you'd add a decent amount of RAM, and most importantly, a dedicated chip to drive TIA's graphics. (equivelent to ANTIC in the 8-bit computer line) Now, withoth any other modificaions you'd still be limited to TIA's 2 players+missiles (compared to 4 with CTIA/GTIA) and TIA's audio, but you'd now have far more CPU time (driving the display ate up at least 1/2 of the 6507's processing power) on top of the 6502's faster clock speed. (having the ANTIC like chip woould also mean there'd be no more "chasing the beam" for programmers)

 

On top of this, keeping the original hardware would make backwards compatibility relatively simple and inexpensive. (and you could still use 7800-ish carts to add more pins without compromising 2600 cart compatiblity) A lot more pracical than including compatibility by tacking on 2600 hardware in addition to the new system's, or offering an adaptor module containing the 2600's hardware.

 

Some of this was already brought up here: http://www.atariage.com/forums/index.php?showtopic=146992

Edited July 14, 2009 by kool kitty89

[mos6507]

If you really wanted an updated system that was heavily derived from the 2600 hardware you'd add a decent amount of RAM, and most importantly, a dedicated chip to drive TIA's graphics. (equivelent to ANTIC in the 8-bit computer line

 

That's exactly what the A8 was, though. The designers took everything they learned from the 2600 and improved it with the A8. The only features that were dropped with the A8 was the sprite copies and mirroring. Otherwise it could do just about anything the 2600 could do as you can see with the near 1:1 ports of Activision 2600 titles to the A8/5200.

 

 

 

In theory they could have made something like a 7800 architecture with the A8, so that it could play both 2600 games and A8 games. I think by and large hardware designers are not fans of backwards compatibility. It's out with the old, in with the new, and certainly the thinking at atari circa 1978 when the A8 was in R&D was that the 2600 wasn't going to build up such a huge following and catalog that such backwards compatibility would be desirable. That's something that only became evident after the 3rd party boom, which is what led to the 7800 project that Atari conspicuously said was guided by consumer feedback from the 5200 flopping.

 

The problem in those early years was cost. In the early 80s it was Commodore that really innovated on the cost reduction front. The A8 and the Amiga after that were much more bleeding edge architectures compared to the 2600 and therefore the ability to package them at a console pricepoint was next to impossible.

Edited July 14, 2009 by mos6507

[kool kitty89]

In theory they could have made something like a 7800 architecture with the A8, so that it could play both 2600 games and A8 games. I think by and large hardware designers are not fans of backwards compatibility. It's out with the old, in with the new, and certainly the thinking at atari circa 1978 when the A8 was in R&D was that the 2600 wasn't going to build up such a huge following and catalog that such backwards compatibility would be desirable. That's something that only became evident after the 3rd party boom, which is what led to the 7800 project that Atari conspicuously said was guided by consumer feedback from the 5200 flopping.

 

Yeah, providing compatibility can be limiting on the design, however, the cancelled "10-bit" 3200 "Super Stella/PAM" was supposed to be backwards compatible (from the limited available for it), and that was supposedly planned for release some time in 1981, so they seem to have shown interest in this. (even if it was at the insistance of management and not the engineers) And there's certainly plenty of examples of this, specifically in the VG industry, Sega was very accomidating in this respect up until the Saturn. (despite the very limited success of their home game systems in Japan, the SG-1000 Mk.III/Master System were backwards compatible with the original SG-1000 and the Genesis/MD was backwards compatible with the Master system via a -largely passive- adaptor, this extending to America snd Europe as well in the form of the Power Base Converter) Of course there are much more extreme examples of continuing compatibility, x86/MS DOS compatibility lasted a very long time, with Win9x OS's still built on top of DOS. (and even XP having rudimentary support for DOS programs, more tacked-on though, not integral to the system) Along with this of course went compatibility of x86 processors. (of course, in both cases, there's significant limitations imposed by doing this, especially given x86 wasn't the best architecture to begin with)

 

The most important reasons for including this feature (from a competitive and consumer standpoint) was that new customers, not having yet bought an Atari system, coud buy the new system, and have the entire existing 2600 library at their disposal. Also, there's the problem with Coleco offering compatibility via their adaptor add-on (followed by Mattel), and offering out of the box compatibility would one up the competition.

 

I've heard that Warner did some consumer studies (surveys, focus groups, etc) before releasing the 5200 as well, but largely ignored them. Then again, with the failure of the 3200 design, mounting pressure from Mattel, and the ColecoVision on the horizon, the 5200 seems to be more of a rushed quick fix than anything else. The controllers are more or less designed to one up the intellivision's with 16-direction disc and keypad. Perhaps there was internal issues with the computer section of Atari as well, preventing a direct XEGS-ish conversion of the 8-bit series.

 

The problem in those early years was cost. In the early 80s it was Commodore that really innovated on the cost reduction front. The A8 and the Amiga after that were much more bleeding edge architectures compared to the 2600 and therefore the ability to package them at a console pricepoint was next to impossible.

 

Yeah, that'd be another reason the A8-bit line wouldn't have necessarily been a good idea in pure console form as a successor to the 2600 ~1980, instead of breaking into the home computer market. (from what I've read, the 8-bit chipset was originally being designed specifically for a home game system, later being worked into computer consoles)

 

That would be one ting about having a more direct evolution of the 2600 hardware, such as adding an ANTIC type chip, full 6502, and some RAM. (even with just TIA for sound it could be significantly better than on the 2600, due to the increased CPU time, hence why 7800 TIA-only sound is often significantly better than the 2600 -there's even some cases where TIA+7800 music is a bit better than the coesponding game on A8/5200) Without modifying the controller connections, they could even have added another 2 buttons (with the standard digital joystick) via the pot lines.

[kskunk]

I don't know the exact details, but the basic idea is that one was a true ground and the other was an input from the cartridge. Let's call that the NO_CARTRIDGE line. Without a cartridge plugged in, the NO_CARTRIDGE line would be connected to +5V through an internal resistor. With a cartridge plugged in, it would be shorted to ground. So this signal could be used as an enable line for the internal ROM.

 

That makes perfect sense. The whole feature would cost just one resistor and one ROM chip, and be cheaper than a pack-in game.

 

- KS

[carmel_andrews]

Well i did rumours, just before atari went ahead and re released the 7800 in europe that atari were planning an xe with either a vcs adapter or 7800 compatibility

 

This was because atari UK said that the UK market and the euro market were still mostly tape/disk driven, so it made sense to pursue the above appraoch rather then launch a whole new system

 

Going back to the 10bit '3200' (or, project 'Sara') this could have got the go ahead if the programmers that rejected the original spec, worked with the big A on a revised spec, i think even the programmers realised that for atari to compete with inty, the system needed to be 10bit as the inty itself was also 10bit

[kool kitty89]

Going back to the 10bit '3200' (or, project 'Sara') this could have got the go ahead if the programmers that rejected the original spec, worked with the big A on a revised spec, i think even the programmers realised that for atari to compete with inty, the system needed to be 10bit as the inty itself was also 10bit

 

Why would you need a 10-bit CPU architecture? (I beleive the programming difficulty came from the new, and rather uncommon archetecture, the Intellivision had similar drawbacks)

Existing 8-bit designs were perfectly adequate for what they needed, the 6502 would be simple to configure for a 6507/VCS compatibility mode (as with the 7800), hence whay they continued to be used for the ColecoVision, Master System, NES, and TurboGrafx 16. Other parts of the hardware, like graphics/video are far more important in this respect. (hell, if you went with the ANTIC like chip for TIA and kept the old 6507, you'd still have around double the CPU resourse available of the 2600, of course you'd still lack the additional 3 address lines and intterupt capabilities of the 6502 -and more with Atair's custom 6502C which reduced the number of necessary support chips -mainly a cot saving measure for the system I beleive)

[carmel_andrews]

I don't think inty's problems revolved around the processor, it was more to do with the marketing and getting 3rd parties to produce software

 

remembering that nintendo's nes system was released in japan in 1983, which upped the stakes in relation to better hardware (but still using an 8bit proccy) and higher mem. cap games

 

Perhaps insteady of the 10bit (whatever) processor, atari could have jumped the gun on everyone in 1982 (i.e both the console and the computer market) and used something like a 68008 (the same processor they used in the sinclair ql) basically a cut down version of the 68000 but accessing less memory. 1 meg...i think if atari could have released a 68008 based 2600 (with slightly extended gfx/sound hardware) but with a fallback 6507 mode (so that you could still play your old 2600 games) not only would atari have seen of the likes of the inty and the CV...it would have kept nintendo away from competiting in the US market since nintendo would not be able to compete against an extended 2600 and say 2 years later, go down the 68000 route and considerably update the existing 2600 hardware, totally blocking out or limiting nintendo's presence in the upscale console market

 

Atari could have had this system retailing for about 180-190 usd (and remember, back wards compatibility with the old 2600 software base) then in 1984/5 as mentioned just go down the 68000 road (as the market would have been geared down that road, what with the st/amiga, lisa/mac and the extended 2600) with considerably expanded 2600 gfx/sound hardware perhaps, combining Maria with an updated gtia chip and quad pokey (from atari coin op) as a 2600 come 7800 type system and replace the existing 5200 with a 5200 based on xl/xe spec's (i.e xl/xe bank switching and high mem. cap games) that would have seriously kept nintendo in check

[kskunk]

atari could have jumped the gun on everyone in 1982 (i.e both the console and the computer market) and used something like a 68008 (the same processor they used in the sinclair ql) basically a cut down version of the 68000 but accessing less memory. 1 meg...

 

I think you're right that 1982 was the correct moment to act. Commodore did act. The C64 may have been the perfect 'console' of that generation, at least based on sales and number of published games from 1982-1986.

 

I love discussing console design, but I don't think processor power or ROM size was very influential in that generation of consoles.

 

Graphics and sound capabilities were far more important. The NES was a monster success and it had the same CPU power as a 1978 Atari 800, with a 32KB limit on program ROM (without additional hardware in the cart).

 

It's a matter of diminishing returns. If a game publisher has to pay $6 for a 32KB ROM, it will be pretty hard to talk him into paying $12 for 64KB. And if he can cram his game into a $3 16KB ROM, you won't be able to talk him out of it. That's six or seven figures profit on a decent-selling game!

 

How much ROM do you really need to win the market in 1982? The best selling game in the period 1976-2006 was Super Mario Brothers. It was 40KB in 1986.

 

For most games, CPU power didn't really matter much until the transition to 3D graphics. Even the Super Nintendo in 1990 only ran at twice the clock of a 1978 Atari 800 and on many codes the SNES's 65816 is no faster per-cycle than the 800's 6502. (Which says something about how good the 6502 is... but that's another topic.)

 

In 1982, if you put a 5200 with a 68008 next to a C64, a lot of people would buy the C64. Not only because it is cheaper to make and thus cheaper to buy, but c'mon... we all had friends in the 80s who were seduced to the dark side of Commodore because they only cared about superficial things like SID music and big sprites.

 

- KS

[carmel_andrews]

only way the 2600 could have gone was by combining the 7800s maria with something like an updated gtia... i.e to add more player/missiles (sprites) and more colours (either something like ste/amiga or falcon) and higher rez (600/480)

 

If the maria chip wan't feasable, another possible avenue would be an all in one antic/gtia chip...but using up some of the duplicate registers with new features...or dual antic/gtia's in one package with new features (i.e'. like blitting and hardware scaling...a'la lynx)

 

 

Sound would probably be something like a quad pokey (from atari coin op) with some of the capabilities of amie/amy thrown in for good measure

 

Thing is though, you would still have to eventually think about upgrading the processor (remembering that atari were planning a 68000 based console in 1984)

 

Now i do remember hearing or reading in some book on 80's computers back in the day that commodore had MOS working on some 16bit version of the 6502 processor (not to be confused with similar being designed by WDC)

 

Now if atari wanted some sort of presence in the console market post 1982, perhaps a 16bit version of the 6502 in some sort of updated 2600 hardware might have worked out

[mos6507]

I don't think inty's problems revolved around the processor, it was more to do with the marketing and getting 3rd parties to produce software

 

There is more to a game machine than the processor. The entire gfx architecture of the Intellivision is not that all it's cracked up to be. It had a smaller color palette, and based on my recollections, difficulty moving sprites around very fast. The games for the INTV played to its strengths and I admit some envy back in the day for things like AD&D or B-17 bomber, but the 2600 still had the lead in twitch-style action gaming.

[mos6507]

Atari floundered because they lost their in-house R&D team so they had no direction anymore. If they had direction they wouldn't have to have contemplated warming over the old hardware to milk it a little more. They would have just leapfrogged over the competition. There was talent in the company, still, but not being applied in the right place. Just look at what the coinop division continued to do in the early 80s, culminating with stuff like the polygons in I Robot. Really, it's all about braintrust.

[kool kitty89]

atari could have jumped the gun on everyone in 1982 (i.e both the console and the computer market) and used something like a 68008 (the same processor they used in the sinclair ql) basically a cut down version of the 68000 but accessing less memory. 1 meg...

 

I think you're right that 1982 was the correct moment to act. Commodore did act. The C64 may have been the perfect 'console' of that generation, at least based on sales and number of published games from 1982-1986.

 

Too bad they screwed things up with the 1200XL... Had they had the 800/600XL out at competitive price points and marketing compared to Commodore, perhaps the A8-bit line would have fared better against it. (the C64 certainly has its advantages, but as far as Graphics go the A8's are good enough, though the resolution's a bit low, and POKEY sounds great -granted many people love/hate the SID, the C64's got a smaller palette to work from as well)

 

The hardware wasn't as cutting edge as it was back in '79, but it would be more accessible and affordable now (and the 600/800XL rectified many issues of the 400/800 designs as well as the botched 1200XL), and of course they'd have the Atari label on them. In fact, they may have been able to get away with promoting the A8-bit line (cheaper 600XL particularly) as their new main game system, and the successor to the 2600. (rather than going the modified XEGS like route, though that's basicly the same idea) Probably come out with an adaptor module as well to counter Coleco.

 

There would still be the problem of price though, if they couldn't get the 600XL down low enough that could cause trouble. (however, given that it'd generally be similar to what the 5200 had on board, but with a significantly smaller case, built-in keyboard, smaller/DE-9 controller ports, and the Parallel+Serial ports, cost shouldn't be that much different)

The 600XL being the game system with computer capabilities, and the 800XL being the higher end machine. (rather like what the 400 and 800 were originally intended to be) The joystick could still be somewhat limiting, though there's no reason they couldn't have introduced a new analog joystick with more buttons. (or a digital joystick using one or both pot lines for additional action buttons)

 

 

I love discussing console design, but I don't think processor power or ROM size was very influential in that generation of consoles.

 

Graphics and sound capabilities were far more important. The NES was a monster success and it had the same CPU power as a 1978 Atari 800, with a 32KB limit on program ROM (without additional hardware in the cart).

 

Certainly, that's the major differences in these systems, that and how the memory is mapped (and how much RAM is onboard), but the Graphics in particular is probably the most significant. (granted the memory mapping would effect this too) In the sound department POKEY was pretty competitive with most of the 3rd generations consoles. (the Master System's PSG was rather simple)

 

Plus, there's no need to change to a different processor architecture necessarily, specifically something like the 68008 (which is similar to or slower than a lot of older 8-bit procs at similar clock speeds, the main advantage being some of its additional features, the larger address bus, and forewards compatibility with the full 68000). The 650x line is rather good to stick with I think (from my understanding, it had a relatively simple and flexible instruction set and is relatively cycle efficient), even for later machines it could be a good choice. (like on the TG-16, basicly a fast 6502 with integrated bank switching to provide 2 MB of address space) I think a major reason Nintendo went with a custom 65816 derivative over a fast 65C02 was the address bus (24-bit, same as the 68000 and 80286), then again they probably could have gone with someting like NEC did with integrted bank switching, but I'm not sure what other issues there were.

[carmel_andrews]

Atari should'nt have bothered with the 1200, what the consumer was looking for was something that was MUCH better then the 800 and the c64 and atari failed to deliver (and i count the xe and later xl series as part of that)

 

whoever that stupid prat of an a-hole was as warner's who told Jay Miner that atari were'nt interested in what Miners idea's were regarding the future of the A8 (or words to that effect)...needs to be outed, named and shamed and treated like a paedaphile for the rest of his/her natural life, because if you think about...that is what killed atari, everything that happened after that, was as a consequence of that mentality and that decision, because that decision affected not only the fututre of the a8 but also any future atari console as well (from the warners period)

 

This is because everything was reliant on R&D and hoping that whatever tech R&D come up with, could be packaged into a consumer product and that product bought to market via the normal processes of mass production, manufacturing and distribution and as we all know warners basically killed R&D so far as the A8 and atari games systems were concerned prefering instead to getting r & d to develop tech that was 2-5 years down the line

 

The problem is though, is that R&D need people (like Jay and his like) that know how to not just develop the tech. but create a product around that tech and get that product and technology into the market via the above mentioned processes and in a small time frame (say 6 months)...unfortunately the way that warners were running atari r&d, it either seriously forgot or overlooked this small but vitally important aspect of r&d..i.e creating consumer product from technology developed by r&d and getting that product into the market via the above processes...

 

these people like Jay were leaving atari faster then they were coming in...atari were not replacing these type of people with similarly experienced outsiders to get atari tech and product out there in the market...these people within r&d were being replaced with sales/marketing people that did'nt know a jot or sod all about creating a product from tech developed by r&d and getting that product into the market place by the above processes...and as we all know you can't sell anything unless you have a product (and warner's r&d were NOT coming up with the products)...so what was the point of hiring these sales/marketing people for if they have no product to sell

 

In short...warners were basically handing over huge budgets to people running these r&d units (aparrently they were called 'engineers' or 'designers'....more like wannabe, so-so or has been engineers or designers) and basically told to get 'high on coke' (or some other illegal substance...but lets not go down that road)

 

The problem also is that atari were willing to invest in this form of r&d more then they were willing to invest in updating the product tech. that atari's existing produxct lines were using....as we all know product based technology has a shelf life a lot less then the shelf life of the product itself ...so if we look at say the atari 800, atari should have started doing something major with the 800 back in 1981 (hardware wise, and i am not just saying, put it in a new case add 16k and a semi compatible o/s) i am talking a major update here, better gfx/sound capabilities etc...remember the c64 already had atari alike features so it was pointless just releasing a 64k modified version of the 800 as that wasn't what the consumer wanted

 

what atari should have been doing instead of setting up these r&d spin offs that had no hope of creating product that could enter the market place via mass production/manufacture and distribution (within a 6 mounth period), atari should have spent their money buying out amiga type companies, as amiga type companies had something that warners r&d lacked...that is people within the r&d unit that knew how to create a consumer product from r&d's technology and getting that product into the marketplace within a small time frame via the above named processess

[kskunk]

The problem is though, is that R&D need people (like Jay and his like) that know how to not just develop the tech. but create a product around that tech and get that product and technology into the market

 

That's a common problem with high tech companies. When technology is brand new and poorly understood, and suddenly starts selling like crazy, management can look at it two ways:

 

1) Wow, I really don't know why this is selling so well. I need to derive the formula for this success so I can repeat it. What did my engineers do and why? Why is the market responding so well? How long will it last? How soon before our competitors steal the market from us?

2) Wow, I must be really smart to be making all this money! Now I can afford to hire all my friends and family as VPs! I don't have to listen to people who think my golfing buddy Jeff is a bad VP of R&D! Money talks: I'm the smartest!

 

Being a CEO is a job that appeals mainly to egomaniacs, and most egomaniacs will jump to option #2 before even pausing to think about #1. That's why the really successful tech companies have paranoid CEOs. Only the paranoid survive.

 

- KS

Edited July 17, 2009 by kskunk

[kool kitty89]

How do the 68000 coputer projects fit into this then? (ie Sierra and Gaza)

[carmel_andrews]

Atari shat all over their book of success, that why they never repeated any of their successes

[Rybags]

What do Atari's 2 biggest successful products (and possibly their only real successes from a marketing standpoint) have in common?

 

Easy. They were designed without meddling from Warner. The 2600 and ST had their initial successes because they were brought into fruition with a minimum of fuss and a relatively clear focus of what they were to do and who they wanted to sell them to.

 

Almost everything else the company ever did after 1980 was a case of a rowboat designed for 6 people having 10 people in it, and all trying to paddle in different directions.

[kool kitty89]

Atari shat all over their book of success, that why they never repeated any of their successes

 

Weren't those projects developed well after Warner taking over Atari? They were prototyped then shelved, but I'm not sure why, given the timeline it may have had to do with the crash. The information about Gaza abd Sierra is rather vague, but they could have been among projects that werepostponed while Morgan was restructuring Atari during the crash. (we'll never know if those efforts wuld have been successful, NATCO etc, as the Tramiel deal ended it)

[carmel_andrews]

Not too sure about that koolk. remember, during that tranisition period whilst tramiel was 'evaluating' what he took over from the big W... he closed down all R&D and tried selling off what could'nt be turned into viable products and replaced R&D with his own team...and ofdcourse...as wel all know tramiel wasn't interested in atari technology or existing atari products (i.e. surplus to requirements)

 

Perhaps if tramiel had organised all atari's existing r&D and combined it into atari's existing product line and made another division...it might have saved a heck of a lot of that technology and we still might be reaping that benefit even today.....unfortunately tramiel was way too small minded and short sighted to see that

[kskunk]

How do the 68000 coputer projects fit into this then? (ie Sierra and Gaza)

There's a big difference between what's cool and awesome, and a product you can build a sustainable business around. Occasionally you'll get engineers who can do both. But most engineers just build cool stuff, and management/marketing need to discipline them all the way to a solid product design. (And this is why engineers are considered grumpy!)

 

If I came across a canceled Vectrex-alike it would be easy to imagine how it could have altered the course of video games, spurning earlier adoption of 3D graphics and allowing game genres to be explored that were simply impossible otherwise.

 

A canceled NES would just look like more of the same old sprite and character hardware that defined the 80s. The NES technology is really not that special.

 

Now which was the better product? Ask the customers, not the engineers. What makes things technically 'superior' often has nothing to do with what customers actually want.

 

So to get back to your question: Sierra and Gaza could have been tragic victims of the management upheaval at Atari. Or more likely, they were half-baked engineering adventures lacking the discipline to get anywhere near production.

 

Atari's R&D oversight was lousy, so it was littered with neat-o projects that had absolutely no hope as products. (E.g., ATW.) But that's what makes Atari so fun to study!

 

- KS

Edited July 18, 2009 by kskunk

[kool kitty89]

Not too sure about that koolk. remember, during that tranisition period whilst tramiel was 'evaluating' what he took over from the big W... he closed down all R&D and tried selling off what could'nt be turned into viable products and replaced R&D with his own team...and ofdcourse...as wel all know tramiel wasn't interested in atari technology or existing atari products (i.e. surplus to requirements)

 

Perhaps if tramiel had organised all atari's existing r&D and combined it into atari's existing product line and made another division...it might have saved a heck of a lot of that technology and we still might be reaping that benefit even today.....unfortunately tramiel was way too small minded and short sighted to see that

 

I was mainly referring to what might have happened to atari under Warner with the changes/reforming being done by Morgan bofore Tramiel acquired the company.

 

I'm aware the Tramiel pretty much dumped Atari's existing R&D (including the computer portions geard more toward his interests), though a lot of shelved/cancelled Atari projects probably weren't very useful for retail (don't know enough about Gaza or Sierra to say where they'd fit in), but he did at least show some intrest in their leftover projects, but as you mentioned, tended to replace the engineers. The AMY sound chip is one such case, which failed after loosing its designers, though the idea of putting one imto an 8-bit seems rather dumb to me. (it could have been an asset on the ST, particularly with its weak sound hardware and given that was one of the Amiga's strong points)

 

 

So to get back to your question: Sierra and Gaza could have been tragic victims of the management upheaval at Atari. Or more likely, they were half-baked engineering adventures lacking the discipline to get anywhere near production.

 

Atari's R&D oversight was lousy, so it was littered with neat-o projects that had absolutely no hope as products. (E.g., ATW.) But that's what makes Atari so fun to study!

 

- KS

 

Yeah, though we won't know for sure until more information surfaces on these particularl projects.

 

Gaza sounds like it may have been less practical, being workstation oriented and (supposedly) featuring a dual 68k multiprocessor layout.

 

Seirra might be more in line with something practical like the Amiga, but there's no way to be sure at the moment. (had Warner been a bit more level headed managing things, Jay Miner may have never left and the Amiga, or something like it may have been developed internally by Atari R&D, perhaps the 8-bit line would have been changed/improved as well)

Of course a lot of theing could have been different with a change in management, that being the major problem in the Warner years...

 

 

 

Ok, that asside, hypothetically speaking, if Atari had succeeded the 2600 and 8-bit line ~82, should they have focused on a single new product, or release new computer and game systems seperately? Also, while the popularity of the 2600 may have been significant enough to warent backwards compatibility, would it be woth bothering with this for a successor to the 400/800 (or likewise working on expanding/improving the 8-bit hardware -more than the simple changes on the XL and XE) or is making a clean start a better idea?

Edited July 18, 2009 by kool kitty89

[kskunk]

Ok, that asside, hypothetically speaking, if Atari had succeeded the 2600 and 8-bit line ~82, should they have focused on a single new product, or release new computer and game systems seperately? Also, while the popularity of the 2600 may have been significant enough to warent backwards compatibility, would it be woth bothering with this for a successor to the 400/800 (or likewise working on expanding/improving the 8-bit hardware -more than the simple changes on the XL and XE) or is making a clean start a better idea?

In the 80s, Atari had an identity crisis. Were they a consumer electronics company or a business equipment company? It's really hard to do either well when you're trying to be both.

 

Commodore illustrates why you should pick your market. The Amiga was a next-gen console aspiring to be a business computer. Sure, that was PERFECT for a few niche markets (like creative professionals, TV studios, teenagers with rich parents, etc). But it was too expensive to be a game console and too hamstrung by its console DNA to evolve as fast as business PCs and workstations did.

 

I think Atari would have been better off splitting consoles and computers into separate companies, just like it spun off its arcade division. After all, the arcade division was still publishing hits 5 years after Atari was dead.

 

With that said, I still think any console released in 1982 should have been marketed as a 'home computer'. A lot of companies figured that out and put keyboards on their consoles that year.

 

The problem with evolving the Atari 800 line into a console is that there is a huge, existing, 3rd party software market you can't earn royalties on. Atari knew that. That's why the 5200 (out in 1982) was as deliberately incompatible with an 800 as they could make it while still using all the same chips.

 

Commodore was so vertically integrated they could sell C64s for console-like prices and still make huge profits. Atari wasn't in that position, so they needed the royalties.

 

In 1982, Atari was also hurt by trying to market two incompatible consoles at once. 2600 carts were still selling like gangbusters, so Atari's own development and marketing teams focused on the more profitable 2600 while the 5200 withered in the stores.

 

So I think the best console plan for Atari in 1982 would be... surprise, the 7800. I think the 7800 was a good design for 1982 because it was created with perfect hindsight of all the things Atari did WRONG in 1982. They learned a lot, but by 1986 it was way too late to apply those lessons.

 

After the video game crash in 1983, stores would not sell game consoles. Nintendo very cleverly avoided the video game stigma by selling 'entertainment systems' with that toy-like robot. The 7800 might have weathered the storm by sharing home computer shelf space with the C64, if it was sold as a 'keyboard bundle'. The keyboard had an SIO port, which was the right amount of 800 compatibility IMO -- enough to leverage a big ecosystem of peripherals without opening the door too widely to unlicensed 800 software.

 

It would have been even better if they included 'forward compatibility' too. That is, allow 2600 games to authenticate and take advantage of 7800 features. That way, a 2600 game could offer '7800 enhanced' graphics. This would help nudge customers to the 7800 without forcing publishers to abandon the 2600 cash cow. (Nintendo did that for two GameBoy generations -- many games were 'Color Enhanced' but still played on the huge installed base of classic GameBoys -- they sold over 100M consoles over 15 years without the painful generational transition others suffered.)

 

- KS

Edited July 18, 2009 by kskunk

[kool kitty89]

So I think the best console plan for Atari in 1982 would be... surprise, the 7800. I think the 7800 was a good design for 1982 because it was created with perfect hindsight of all the things Atari did WRONG in 1982. They learned a lot, but by 1986 it was way too late to apply those lessons.

 

Do you mean a console with the 7800's general characteristics, or that specific hardware arrangement? (like some of the alternitives I proposed)

 

The 7800 was cost optimized so taking the 8-bit/5200 hardware and smacking 2600 hardware onboard for compatibility wouldn't be an ideal option. (though adding a POKEY would probably have been a better idea in the long run than the on-cart expansion -might have been used for the planned keyboard interface as well)

But instead of MARIA, adding an ANTIC like chip, POKEY, and 6502 (or 02C) in place of the 07 could have been a good route to go with. (maybe enhance TIA as well, but with the ANTIC-alike you'd have a lot more options and a lot more CPU time already, include a lockout/licence protection system, of course)

 

 

 

It would have been even better if they included 'forward compatibility' too. That is, allow 2600 games to authenticate and take advantage of 7800 features. That way, a 2600 game could offer '7800 enhanced' graphics. This would help nudge customers to the 7800 without forcing publishers to abandon the 2600 cash cow. (Nintendo did that for two GameBoy generations -- many games were 'Color Enhanced' but still played on the huge installed base of classic GameBoys -- they sold over 100M consoles over 15 years without the painful generational transition others suffered.)

 

- KS

 

Would that have been possible to arrange? (specifically with the current 7800 hardware -even if interfaced to the cartridge port differently)

[carmel_andrews]

Atari made 2 big mistakes relating to it’s dealings with amiga (one from warners/one from tramiel)

 

Warner’s should have put in the original agreement that if the comapny is sold to a third party (regardless and irrespective of any or all the loan being repayed) Atari retain all rights to any technology that amiga develops with that 3rd party

 

Tramiel should have said to amiga (apon them repaying the loan)...sorry, i am not interested in accepting the payment, furthermore the contract doesn’t stipulate or insist that atari has to honour accepting repayment of the loan...atari want that technology period, now if you want to do a deal with this other company, that is upto you, just bear in mind that we won’t accept your repayment, we only want the technology

 

Now, one of 2/3 scenario’s would have occured in the former

 

Original amiga, if they didn’t make an agreement would probably have gone insolvent,atari would have probably come in and snapped up the developemnt team and probably the technology

 

Original amiga would have probably had to accept that as part of any deal with atari and would have had to inform other parties (like commodore)that atari would still have rights to any or all technology that amiga and their new owners come up with

 

Amiga’s new owners (commodore) would have to either do a deal with atari (allowing them so manufacture some version of the finished lorraine c/s) or basically get into bed with atari, sharing R&D costs and both coming out with variants on the amiga chipset, commodore, going down their amiga route and atari either integrating the lorraine c/s within a 68k modifed A8 o/s or a modified ST alike o/s and in a bid to mute the presence of the pc and mac dominance in the upscale market to agree that whatever amiga type system was the most successfuil after 5 years, for both manufacturers to just develop that system only, but have backwards compatibility with the dropped platform and get 3rd parties to start making 3rd party versions of the successful amiga type system and sharing any R&D costs if anyone develops new variants of the hardware (i.e gfx/sound chipset etc)

 

2 scenario’s would probably eminate from the latter (tramiel) scenario

 

For amiga to have to go back to commodore and tell them that atari won’t accept the loan repayment, the only want the technology

Commodore would have one or two choices to make, either withdraw the offer (unlikely as atari would have ended up with the spoils and also the ST to boot) or accept the offer and try working out some deal with atari (bearing in mind that atari won’t accept the loan repayed which basically means they won’t accpet monetry compensation) or get into bed with atari

 

For amiga to accept the commodore offer but get an atari agreement that they either don’t manufacture any amiga compatible machine that is compatible with a cbm/amiga agreement for a period of 2 years (possible as tramiel would simply implement a ST alike o/s onto the existing lorraine chipset)

 

3rd possible scenario would be that commodore buy out amiga, make it insolvevent and sue atari over the rights to amiga’s technology, a high rise strategy for commodore as even if atari were to accept some level of monetry compensation it would be significantly higher then simply repaying the loan (remembering that commodore were just starting to lose money and also the costs of getting whatever amiga were developing into hard sillcon) again only scenario that wouild apply here would be for commodore to do some sort of deal with atari

[kskunk]

But instead of MARIA, adding an ANTIC like chip, POKEY, and 6502 (or 02C) in place of the 07 could have been a good route to go with. (maybe enhance TIA as well, but with the ANTIC-alike you'd have a lot more options and a lot more CPU time already, include a lockout/licence protection system, of course)

 

TIA's graphics abilities wouldn't improve dramatically even with an ANTIC-like feeding it. You would have to enhance TIA to be more like GTIA. GTIA can accept a new background pixel from ANTIC every color clock. A stock TIA is not near fast enough for that.

 

But enhancing TIA seems like a lot of work for no particular gain. The Atari 8-bits have separate ANTIC and GTIA because it was the only cheap way to build such a capable video system in 1978. By 1982, much more capable video systems fit on one chip. (See VIC-II.) So it would make sense to put the 'new' graphic system in one chip instead of depending on TIA for anything.

 

I'm definitely on the fence about whether MARIA was good or bad for the 7800. It's a very capable chip that's pretty tough to program. But in the eyes of 2600 programmers in 1982, it might be intuitive and very easy to work with. Traditional character mode systems (like ANTIC) are easier to use but less capable.

 

ANTIC was also 4 years out of date by 1982. I think ANTIC is really cool, but it was before the 'discovery' of attribute RAM, the feature that made so many C64/NES/Genesis/etc games look so colorful.

 

In the end, if Atari included an ANTIC-type character graphics system in their 1982 console, they would need to upgrade it to be state of the art. Going with 4 year old technology (I'm thinking of you 5200), is not a recipe for success in the console industry.

 

It would have been even better if they included 'forward compatibility' too. That is, allow 2600 games to authenticate and take advantage of 7800 features.

 

Would that have been possible to arrange? (specifically with the current 7800 hardware -even if interfaced to the cartridge port differently)

 

The lock-out system in the 7800 is designed to prevent this. But you could allow hybrid 2600/7800 games by removing a couple of traces on the PCB and making a change to the boot ROM. I don't think a different cartridge port would be required, although there are some other details with addressing that could be better.

 

If they had wanted to support this it would have been easy. From the lock-out design, it looks like they specifically wanted to _prevent_ hybrid 2600/7800 games. My best guess is that they were thinking about some day dropping the TIA, and wanted 7800 games to avoid depending on any 2600 features beside TIA sound.

 

- KS

Edited July 18, 2009 by kskunk