Atari 800, a missed opportunity for Atari

[carmel_andrews]

They say that the Apple compared to the Atari 800, was more expandable

 

Fact is, the Atari 800 was just as expandable as the Apple II was but in a slightly different way and i am besides myself with surprise that Atari didn't expand on this and exploit this opportunity

 

Now, whilst the Apple II did offer expansion slots and the Atari 800 didn't, if you look at the way the 800 was designed, Atari put the processor, graphics/sound chips on seperate removable boards, the idea being that the 800 wasn't just limited to using a particular brand or family of processor and also the atari wasn't just limited to using the same antic/gtia etc chips since everything was on a removable board specific to that particular chip, if an improved version of that chip came out you could take out the board and remove the old chip and put in a new one, or alternatively, if atari had already put the new chip on a board that you could just replace that old board with the new board, just as simple

 

So having removable boards on the atari 800 was the atari equivalent of apple II's expansion slots

 

Additionally because the processor was also given it's own board, you were'nt limited (as previously said) to using a particular brand or family of processor, if Atari had understood how the 800 was designed and looked at the possibilities of having a processor on a removable board instead of being hardwired on a main/motherboard, it would have meant you could adapt the system to practically working with any type of processor

 

Now considering that various hardware manufacturers jumped on the pc/mac bandwagon due to it's expandability capabilities (i.e you weren't just limited to using a particular make of graphics/sound card, modem or ethernet card or processor) and probably made a ton of money from making/selling new and updated graphics/sound cards, processors, internal modems and ethernet cards, why couldn't atari have gone down this route with the atari 800, since as mentioned previously, the removable boards the 800 had was the equivalent of the apple II (and therefore the pc/mac's) expansion slots, it would have created a large market for the 800 hardware upgrades (especially if you look at the present pc market and also the apple II market from WBW)

 

And why Atari went froma largely 'open' and 'expandable' system like the atari 800 to something like the somewhat closed system (expansion wise) that the later xl/xe systems offered, remembering ofcourse that the PBI and ECI expansions were very few and far between and wasn't well (if at all) supported....probably due in large part to Atari's policy on 3rd party software and hardware support

 

Atari was basically telling 3rd party software houses to develop for the 800 was the base A8 system to program on, that was a good move ONLY when the atari 400/800 systems were only one's made by atari, this policy didn't make sence when the later xl and xe systems were being made or sold, remembering the fact that over 75 percent of atari software released during the A8's lifetime (i.e 1978/9 to 1992) is sub 64k, which means that the XL and XE systems got very little support in so far as software that supported these systems higher memory requirements

 

After all if your being asked to develop for the 800 was the base A8 to develop for, where's the incentive to develop for the xl (and therefore the xe) and the higher memory requirements, bearing in mind that software companies had NO PROBLEM supprting the commodore 64 and it's higher memory requirements...

 

Could you imagine if the c64 was vic 20 comaptible and commodore were telling software houses to write to the vic20 as the base commodore system to develop for (or if you want to go back that far, make the c64 PET compatible and ask software houses to develope for the PET as the base commodore system to develop for), would the c64 have been the best selling H/C if they had pursued that policy, no, infact it probably would have resulted in people buying more vic20's and giving the c64 an afterthought (remember what happened to the 1200xl and why it resulted in people buying more 800's)

 

That software support policy of Atari's not only killed off decent software support for the XL and XE systems, it also TOTALLY damaged and hindered any support for the PBI or ECI expansion possibilities (rememebering that atari were telling the software companies to develop for the 800 as the base A8 system to develop for and that the 800 didn't have this expansion the xl or xe had) like i said software houses and hardware developers has little or no incentive to supporting the xl or xe's additional capabilities (i.e more memory and expansions) because atari was telling the publishers and developers to stick with the a800 as the base a8 system to develop for

 

The thing that struck me about Atari's software policy is that Tramiel effectively continued that same policy as warners started, which again doesn't make sence especially when your going to release an updated version of the XL system and also a version of the updated XL system with twice as much memory, remembering that tramiel often referred to himself as 'the computer industries most experienced general manager', if that was so, when why did he continue with the same software support policy that warners started if it meant that his new systems and older systems with higher memory requirements (and expansion possibilities) didn't get the same level of support that cometitive systes like the commodore 64/apple II were getting

 

like i said, most of the software released on the A8 during it's lifetime was sub 64k...not exactly a good policy to have if you want support from 3rd parties for machines that have higher memory requirements and expansion possibilities, after all, unless atari were going to do a 'nintendo' like scenario and distribute/sell 3rd party software and hardware products itself, unless atari ditched the policy of writing/dev'ing for the a800 as the base a8 to write for, software comapnies were not and would not develop 2 versions of the same game (unless you are datasoft, novagen or synapse/synsoft) one version for sub 64k systems and one version for xl/xe systems...

 

After all how many software companies developed one game for the STFM/ST and then developed the same game but using the STE's extra features for the STE, bearing in mind that 98 percent of all ST/STE software released from 1985 to 1993/4 was written to the base system (i.e the ST/STFM...not the STE) according to a quote made by ex atari product manager Daryl Still that was published in ST format

[Bryan]

Boy, that's a long post. A few things:

 

The Atari 800 couldn't have the same style of externally-accessible slots as the Apple because the Atari connected to a TV natively. The FCC rules were very strict in the '70s.

 

Atari tried to keep the system details a secret for the first year or so which greatly hampered 3rd party development.

 

Developers will write for whatever hardware level will give them the best balance between features and potential customer base. I doubt very many developers consulted Atari on the issue.

[Rybags]

The 800 is nowhere as expandable as Apple II.

 

And it isn't more expandable than the XL.

 

As for putting other CPUs in - the motherboard doesn't have the signalling required to run many other CPUs. Z80, no go. Intel 808x no go. Motorola 6809 - maybe. Motorola 68008 - didn't even exist when the 800 came out and wouldn't have mattered anyway as it wouldn't work either.

 

The RAM used pre-XL was only good for a system speed of around 2 MHz. A Z80 at 2 MHz is a joke. An 8086 at the same speed wouldn'b be much better.

[Subby]

And remember that in the late '70s they were inventing this entire industry. Most companies really had no clue of what they doing, some reasons understandable, some not.

As I understand it the AppleII got "snuck" by the FCC, but they had woke up by the time the Ataris were ready to go.

And in the very early 80s there were persistent rumors that the A8 would get an expansion chassis to compete with the Apples. Which goes along with the 815 and the business software that Warners bean counters killed in order to make the Atari a game machine. (idiots)

And lastly, devs quite often still won't even do what you want them to do. CBM wanted more devs to support the C128's CP/M and 128 modes. Most devs only supported the C64 mode thinking more guaranteed sales in the short term, but IMO they sacrificed the long term.

I do agree with the general theme of your post though, Warner had a great opportunity to have their cake and eat it too. I think they could have very well competed with AppleII in -both- business and entertainment.

It was a very frustrating time to be an Atari fan.

[Willsy]

Could you imagine if the c64 was vic 20 comaptible and commodore were telling software houses to write to the vic20 as the base commodore system to develop for (or if you want to go back that far, make the c64 PET compatible and ask software houses to develope for the PET as the base commodore system to develop for), would the c64 have been the best selling H/C if they had pursued that policy, no, infact it probably would have resulted in people buying more vic20's and giving the c64 an afterthought (remember what happened to the 1200xl and why it resulted in people buying more 800's)

That is exactly what happened with the Commodore 128.

[JamesD]

As I understand it the AppleII got "snuck" by the FCC, but they had woke up by the time the Ataris were ready to go.

By not having RF out the rules were more relaxed for the Apple. That's why a 3rd party built and sold the RF modulator for the Apple. As long as Apple Computer didn't make or sell a device to hook it to a TV, they got away with it.

<edit>

Tandy wasn't so lucky. The Model I with it's expansion interface had too much RF even under the more relaxed rules.

Edited July 22, 2011 by JamesD

[ClausB]

The 800 is nowhere as expandable as Apple II.

Steve Wozniak said he insisted over Steve Jobs's objection that the Apple II have eight expansion slots. Jobs thought two would be plenty.

 

To be fair to Atari, they designed the SIO bus as their means of expansion. They were ahead of their time - fast serial bus with intelligent peripherals - sounds like USB!

[+Stephen]

The 800 is nowhere as expandable as Apple II.

Steve Wozniak said he insisted over Steve Jobs's objection that the Apple II have eight expansion slots. Jobs thought two would be plenty.

 

To be fair to Atari, they designed the SIO bus as their means of expansion. They were ahead of their time - fast serial bus with intelligent peripherals - sounds like USB!

Courtesy of Digital Press:

Joe Decuir:
One of the few people that really need no introduction. As one of Atari''s original "Imagineers", 
he helped design the Atari VCS/2600, as well as the 400/800 SIO interface.  
He went on to help develop the Amiga computer and later worked on the USB
architecture that we all commonly use today.

[jonnosan]

 

Now, whilst the Apple II did offer expansion slots and the Atari 800 didn't, if you look at the way the 800 was designed, Atari put the processor, graphics/sound chips on seperate removable boards, the idea being that the 800 wasn't just limited to using a particular brand or family of processor and also the atari wasn't just limited to using the same antic/gtia etc chips since everything was on a removable board specific to that particular chip, if an improved version of that chip came out you could take out the board and remove the old chip and put in a new one, or alternatively, if atari had already put the new chip on a board that you could just replace that old board with the new board, just as simple

 

 

Why is it you think the boards were made removable for the purpose of expansion? I would have assumed that removable boards were to facilitate troubleshooting & repair.

 

Although there are technical issues with interfacing foreign CPUs to the A8 bus, these would not have been insurmountable - the S100 bus was essentially the 8080 pinout, and there are certainly different CPUs available as S100 cards e.g. 68000, 8086, even the 6502. But the biggest issue I think would have been the chicken/egg issue of: an alternate CPU in the A8 would have no software available for it (apart from what the alternate CPU vendor provided), and without a significant install base, nobody would bother writing software for the alternate CPU, and without any good software available for an alternate CPU, nobody would want to buy the CPU board.

[Rybags]

To an extent the board system (esp RAM) was geared towards expansion.

 

But, like practically everything else marketed as "expandable", Atari never supported it properly in that regard. Repeat the story for XL, XE, 7800 and ST to an extent.

 

The 65816 was presented to Atari but after the 800 was discontinued. Naturally if that CPU was used as an upgrade path, the OS probably would have been updated to suit.

 

So in summary, the slot system as employed was a good idea in principle, albeit limited in capabilty, and by the time the opportunity came along to use it effectively the machine had been superceded.

 

Of course we have the 65C02 in the interim, but in reality the small extra advantage it would give is probably not worth it if you look at incompatability issues vs possible gains.

[+Faicuai]

They say that the Apple compared to the Atari 800, was more expandable

 

Fact is, the Atari 800 was just as expandable as the Apple II was but in a slightly different way and i am besides myself with surprise that Atari didn't expand on this and exploit this opportunity

 

(...)

 

 

And why Atari went froma largely 'open' and 'expandable' system like the atari 800 to something like the somewhat closed system (expansion wise) that the later xl/xe systems offered, remembering ofcourse that the PBI and ECI expansions were very few and far between and wasn't well (if at all) supported....probably due in large part to Atari's policy on 3rd party software and hardware support

 

(...)

 

The thing that struck me about Atari's software policy is that Tramiel effectively continued that same policy as warners started, which again doesn't make sence especially when your going to release an updated version of the XL system and also a version of the updated XL system with twice as much memory, remembering that tramiel often referred to himself as 'the computer industries most experienced general manager', if that was so, when why did he continue with the same software support policy that warners started if it meant that his new systems and older systems with higher memory requirements (and expansion possibilities) didn't get the same level of support that cometitive systes like the commodore 64/apple II were getting

 

(...)

 

 

 

...Your thoughts, generally speaking, are pretty much aligned with mine (and right on the mark, as well).

 

The JayMiner-800 (as I affectionally call it) was, indeed, more expandable and flexible than anything that came later after, especially from a practical point-of-view. Of course, there is a DEDICATED expansion bus on my 800XL, but it required an UGLY-LOOKING external box (like mating a horse with a dog), and it got NOWHERE near as exploited as the actual slots in the 800, which eventually saw (on their days) significant memory expansions in multiple forms/combination, system monitors, multiple ROM enhancements, integrated 80-col. boards, Ramdisks, etc. Even today, we have seen projects that have been able to provide nice RAM expansions on JM-800, albeit with tight/machine-centric banking-schemes, etc. NONE of this can be done as easily on the 800XL as it can be on 800.

 

Now, the CATCH here relates to a variable that no-one on this thread seems to have addressed, yet: DENSITY/SURFACE. This is probably the key, prime asset that the Apple II had/has, which the Atari did not. Even if the Apple had two (2) xpansions slots, you could stick a MILE-LONG board in-there, and with minimal Power-Supply problems either, without mentioning that cabling could be routed immediately through the factory-provided back-slots, and close the machine like nothing happened.Therefore, and thanks to its bus/slots design + physical space, it was possible to conceive Memory (and other) expansions of clearly larger capacity than the 800, even with run-of-the-mill components available during such days (e.g. the 800 required more advanced, denser and more expensive parts to achieve the same). This, in turn, eventually enabled more sophisticated applications (albeit more expensive).

 

On the marketing/positioning side, I think Atari started on the right foot, by poitioning/coining their machines as "Home Computers" (instead of another run-of-the-mill "Personal Computer" which, ironically, later adopted). The concept of "friendliness", "simplicity", "household integration/coexistence" was there, as you had a more useful, more capable machine, smaller / more compact machine than the Apple II, that could immediately be hooked-up to your TV, built like a tank, with much nicer sound, graphics, and reasonably expandable (including peripherals). Such connectivity model allowed products like the IndusGT (all-round best drive for the 8bits) to RUN CP/M from the drives' own processing resources and RAM, and use the host Atari as an intelligent/sophisticated terminal (although this came a bit late, though).

 

I wonder what would happen if all externally-accessible boards were removed, then replaced with some dedicated ribbons connected to each accessible slots on the 800, and then route them out through a new, after-market top-lid (OEM/factory looking) or simply internally, to the right-side of the machine, where there is a TON of unused, well-ventilated space, and then provide some nice expansions on a separate box/module, where you could do whatever you wanted / needed.

 

In any case, I agree Atari missed this window but, as pointed out on this thread, it was eventually applications and software what eventually determined the life and use of any of these inventions, back in the day (and continues to remain true, even in today times).

 

 

F.

[Rybags]

Physical space is only one aspect.

 

The main constraint on the 800 is lack of required signalling.

 

The second constraint is the closed architecture of the original OS in comparison to the XL.

 

With PBI devices you have built in hooks so that you can have seamless integration (from a software perspective) of things like HDDs, and dynamically loaded/linked handlers if required - and the SIO protocol was also expanded to cater for such instances.

 

Additionally you have the ability to overlay any location in the memory map with RAM, ROM or I/O space from an external device.

 

The fact that the whole thing might look like Cletus towing a uHaul trailer with his '76 Pinto might be a detracting factor, but that's the price of having a compact form-factor machine.

[+wood_jl]

The JayMiner-800 (as I affectionally call it) was, indeed, more expandable and flexible than anything that came later after, especially from a practical point-of-view. Of course, there is a DEDICATED expansion bus on my 800XL, but it required an UGLY-LOOKING external box (like mating a horse with a dog), and it got NOWHERE near as exploited as the actual slots in the 800, which eventually saw (on their days) significant memory expansions in multiple forms/combination, system monitors, multiple ROM enhancements, integrated 80-col. boards, Ramdisks, etc. Even today, we have seen projects that have been able to provide nice RAM expansions on JM-800, albeit with tight/machine-centric banking-schemes, etc. NONE of this can be done as easily on the 800XL as it can be on 800.

 

There were a couple 80-col boards, and some Axlon RAM upgrades that come to mind, and perhaps some of these other options you speak of for the 800, but it was FAR from "exploited." Any such 800 boards are indeed RARE today, and that's (of course) because it was not exploited.

 

I'd say the XL's PBI got more "explotation" (not really) through stuff like the MIO, HD interfaces, RAM upgrades, etc. But I'd still have to say PBI stuff is kind of rare. I don't think any of the expansion options on the Atari (regardless of generation) were "exploited." But if running a hard disk is your bag on the 8-bit Atari, the XL/XE is your machine, is it not? How many people really ran the expensive Corvus kludge through the joystick ports of a beige 800?

Edited July 30, 2011 by wood_jl

[flashjazzcat]

The JayMiner-800 (as I affectionally call it) was, indeed, more expandable and flexible than anything that came later after, especially from a practical point-of-view. Of course, there is a DEDICATED expansion bus on my 800XL, but it required an UGLY-LOOKING external box (like mating a horse with a dog), and it got NOWHERE near as exploited as the actual slots in the 800, which eventually saw (on their days) significant memory expansions in multiple forms/combination, system monitors, multiple ROM enhancements, integrated 80-col. boards, Ramdisks, etc. Even today, we have seen projects that have been able to provide nice RAM expansions on JM-800, albeit with tight/machine-centric banking-schemes, etc. NONE of this can be done as easily on the 800XL as it can be on 800.

As time goes on, I suspect the PBI and ECI of the XL/XE are being pretty well exploited, although I'll concede there probably wasn't a great deal around back in the day, unless you had a Supra HDD, Black Box or MIO. In any case, even these three were completely unusable with an 800. As for ugly looking boxes and RAM upgrades: the RAM320XL which was produced a year or so back was a very unobtrusive plug-in upgrade for the 600/800XL. The only caveat was that the 800XL has no power supply on the PBI by default.

 

I find it hard to imagine anything which says "interface me" more than the ECI on the XE (which has a pretty good set of signals to complement the PBI code in the OS), although the question of physical positioning is another issue entirely. The 800 had one big-ass case, and overall I think I prefer the slim lines of the XL and XE, with the option of running a ribbon cable out the back.

Edited July 30, 2011 by flashjazzcat

[Subby]

One of the things you have remember is that when the original 800/400s were designed, they matched the general design philosophy of the time.

In other words, the general public was still comfortable with the typewriter looking things, so let's give them typewriter looking things, Atari 800, Apple II, BBC something-or-another, you get the idea. Of course there were machines that didn't go that route, TI-99, VIC20, CoCo, more. Yes this is some gross generalization.

[Stormtrooper of Death]

To an extent the board system (esp RAM) was geared towards expansion.

 

But, like practically everything else marketed as "expandable", Atari never supported it properly in that regard. Repeat the story for XL, XE, 7800 and ST to an extent.

 

The 65816 was presented to Atari but after the 800 was discontinued. Naturally if that CPU was used as an upgrade path, the OS probably would have been updated to suit.

 

So in summary, the slot system as employed was a good idea in principle, albeit limited in capabilty, and by the time the opportunity came along to use it effectively the machine had been superceded.

 

Of course we have the 65C02 in the interim, but in reality the small extra advantage it would give is probably not worth it if you look at incompatability issues vs possible gains.

 

I dont know if my reply fits the topic, but i had an Atari Microsoft Basic cartridge. It was 8KB or 16KB and you could use another basic on your atari. I tried it on my Atari 800XL and it did work. I dont know if you replaced the Atari 800 basic module with the Microsoft one, if the ATari 800 would be a real native Microsoft BASIC atari.

 

Did anybody on this very nice Atari forum ever tried that ?

 

PS: I bought this Microsoft BASIC cartridge in Eindhoven in the 90s. Lucky for me, it also included the disk version that was a bit more extended. (with more commands etc.)

 

PPS: I have to admit, that there were not many Atarians that ever made programs in Microsoft BASIC for this add-on.

[+Stephen]

I dont know if my reply fits the topic, but i had an Atari Microsoft Basic cartridge. It was 8KB or 16KB and you could use another basic on your atari. I tried it on my Atari 800XL and it did work. I dont know if you replaced the Atari 800 basic module with the Microsoft one, if the ATari 800 would be a real native Microsoft BASIC atari.

 

Did anybody on this very nice Atari forum ever tried that ?

 

PS: I bought this Microsoft BASIC cartridge in Eindhoven in the 90s. Lucky for me, it also included the disk version that was a bit more extended. (with more commands etc.)

 

PPS: I have to admit, that there were not many Atarians that ever made programs in Microsoft BASIC for this add-on.

I don't think MS BASIC was ever very popular. Truth be told, the only different version of BASIC I heard about was Turbo BASIC XL, and the somewhat occasional mention of the OSS cart BASIC XE.