Alternate History: What Atari Could Have Done Differently?

[pcrow]

1 minute ago, mytek said:

Unless I'm mistaken, owning MOS's chip fabs was one of the main reasons that Commodore could undercut the competition with lower pricing.

 

I would imagine so. Or at least it would have been considerably cheaper and faster when Atari R&D needed a new custom chip.

Good points.  Besides cost, this would have really helped get FREDDIE out sooner, as well as CGIA (combined Antic and GTIA), KERI (combined Antic, GTIA, and Pokey), and MUFFY (FREDDIE and MMU).  I also don't know if Porkey (dual/stereo Pokey chip) was used in coin-op consoles or was never released.  All of those would have made for smaller and cheaper systems.

[phaeron]

Would have been nice if the 8-bits had gotten an 80-column video mode.

 

Problem is, most of the good ways of doing this would have required an AGA-like respin of ANTIC and GTIA to produce an Alice+Lisa like upgrade, since there isn't enough memory bandwidth to double the existing IR mode 2, and GTIA gets all playfield data through ANTIC. That includes doubling the memory bus speed to 3.58MHz, interleaving memory banks to produce a 16-bit video bus, double CAS, or taking the character generator off the main bus. Probably the smallest change would be to add a 640x192 graphics mode and software render text to it, requiring only the ANx bus to be upgraded, but it would have taken 16K RAM and would have been very slow due to the DMA overhead.

 

That leaves the C128-style hack approach of gluing a separate 80-column display output on the side, integrating poorly with the rest of the machine. Best approach would have been to build it into the machine, but the 1090 never made it and it wasn't in the 1400XL plans. The XEP80 shipped but was hobbled by using the awkward and slow joystick port interface instead of SIO, not to mention the marginal video timing.

 

[eegad]

I always wished Atari would have released the 800xl earlier (basically release it instead of the 1200xl), and also have released a cart compatible gaming system instead of the 5200 (the xegs was the right idea, but years too late...there should have been a xlgs back in 1982, 1983).

 

[_The Doctor__]

There were licensing issues for machines with a keyboard vs machines without a keyboard, so you might save money licensing software for one vs the other. This  is also why Atari could have brought out the 800 lines architectural successor (loraine) which is what we come to know as an Amiga now without a keyboard. Loraine was only slightly different.

[+mytek]

I just thought of one additional reason that Atari, not Commodore, should have bought MOS Technologies. Chuck Peddle.

 

It was Chuck Peddle who was working at MOS that persuaded Jack Tramiel to create an 8-bit computer based on the 6502, hence the reason the PET came into existence, with the VIC and the C64 to follow. So it's kinda like one of those SciFi movies where someone goes back in time and stops someone from meeting someone else that changes the whole course of history.

[+bf2k+]

1. Fire upper management...

2. Fire marketing...

3. Let engineers run wild...

 

 

[pcrow]

This predates Atari, but I wish the 6502 had several more opcodes:

 

I would have loved JSR (indirect) to allow for tables for OS routines and handlers (implemented with tables with 3-byte entries starting with the JSR opcode instead using JMP (indirect), but having it would have made the code a lot nicer).  I would think this would have been very easy to implement.

 

I would also have liked an addressing mode where you specify an offset from the stack pointer, which would have made pushing parameters on to the stack much more convenient.  This might have been more tricky to implement.

 

Oh, and any of the 6502C opcodes would have been nice, especially having STZ (store zero), letting you write zeros without having to clear a register first.

[xxl]

- pay the designers extra money for homework and buy their Amiga,
- remove the lock on 2KB RAM under the HW registers (1 additional cable and MMU reprogramming) = MapRAM
- stronger emphasis on the development of SIO (USB)
- publish HW documentation for bedroom programmers

[pcrow]

11 minutes ago, xxl said:

- remove the lock on 2KB RAM under the HW registers (1 additional cable and MMU reprogramming) = MapRAM

[pcrow]

15 minutes ago, xxl said:

- remove the lock on 2KB RAM under the HW registers (1 additional cable and MMU reprogramming) = MapRAM

I thought you could control which bank was mapped from 32K-48K.  Can you not map the upper 16K into that range?  There's really no good way to use the machine without access to the hardware registers.  I could see a good argument that they should have gone with smaller windows of 128 bytes per device or even 64, though that would have been a pain for some modern PBI and cartridge tricks.

[+kheller2]

On 3/23/2024 at 4:19 PM, pcrow said:

I really wish Atari had put out a 720K 3.5" drive back in 1984.

no you don't.  The format wasn't even decided upon back then and Atari had enough problems getting 5.25 Tandon drives working.  Apple dealt with the 3.5" drive at the time.  Now Atari should have planned to release one in 85/86 yes.

[+kheller2]

Atari could have handled their finances better and tried to grasp the concept that money was not going to keep being printed on 2600 bills.

[Hawkeye68]

I think if we could go back in time, prevent all the mistakes and create the perfect Goldilocks scenario, Atari computers still would have been obliterated by PC clones.  People wanted at home what they used at work.  The market for MS-DOS software was enormous.

 

And I've always wondered if the Warner Atari really made money on the 400/800/1200/XL, or if financially, they would have been better off sticking to games.

[pcrow]

54 minutes ago, Hawkeye68 said:

I think if we could go back in time, prevent all the mistakes and create the perfect Goldilocks scenario, Atari computers still would have been obliterated by PC clones.  People wanted at home what they used at work.  The market for MS-DOS software was enormous.

Oh, that's almost certainly true.  I don't even see a plausible scenario where Atari's 68000 system surpassed the Macintosh (even if Atari had bought MOS, Commodore never went into the computer business, and Atari's offering was a combination of all the best of the Amiga and ST).  But in starting this thread, I was more thinking what changes would have given us today a larger community with even better systems.  And honestly, I was hoping for more technical than business ideas, but that's just my bias.

 

But suppose Atari had recognized the inevitability of the PC?  What if they shifted to making games and add-on hardware for the PC?  Put a PIA and POKEY on an ISA card in 1984 with audio and a pair of joystick ports, and then port their more popular games to the PC platform.  (Or probably a better sound chip by then, but you get the idea.)  Where might that have taken Atari?

[cathrynm]

My view, most of the company destroying mistakes were on the console side. What they might have done is hire top management who 'actually played the games.'  If this had been the case, someone might have noticed that the 5200 joysticks made Pacman, which would have been the killer app for 5200, not quite what it should have been. This was the start of the end. All the pieces were there, but no.  I did work with someone awhile back who had worked at Atari around that time, and the project he was on was the "Mindlink joystick."  Somehow you were going to strap sensors on your brain and control games with your mind.  We knew about this at Synapse because we were doing something simlar with 'Relax' -- so we knew what was going on. This was never going to work for games, but according to the guy I talked to they had huge hopes the Mindlink controller would be the next big thing, but no, it was dumb and I suspect nobody was willing to say anything. The final company destroying blunder was not releasing the 7800 when they had it.  It might have saved the company. Basically 7800 + a Pokey chip would have been a perfect successor to the 2600. The system had decent graphics, it had 2600 compatibility, but no, they sat on it. I guess they just didn't know what they had.

[thank you]

i went from 130xe to 386dx in 1994, so i missed a lot of things.

 

i don't know if there would have been an appetite for vendor specific PC add-ons in the 80s...

 

But it seems there were only a few survivors:

 

Apple/Mac was inferior for so long but made it through somehow.

 

Software companies like EA moving to PC... Atari tried this to some extent (not sure when this started)

 

Nintendo was maybe the best case for Atari but the NES and Gameboy were compelling vs. 7800 and Lynx (Tetris, Mario).

 

Maybe they could have put out a crappy PC-DOS and saved us from fifty years of MS, but now I'd probably hate Atari instead

[reifsnyderb]

1.  Pursued the education market that Apple was camping on.

2.  Had a larger 8 bit computer with internal expansion slots.

3.  Explored changing the 400 and 800 shielding out for injection molded plastic with stamped metal shielding to reduce it's price as a cost-saving stop-gap measure until the XL's were released.

4.  Released the 800XL instead of the 1200XL and/or had a PBI with the 1200XL.

5.  Eliminated the internal competition so as to cut costs.  i.e.  The 5200 should have been a 600XL, without a keyboard, in the 5200 case.

6.  Finished a better sound chip.  i.e.  Amy

7.  Produced an improved graphics chip.

8.  Integrated 3rd party upgrades in later products.  i.e.  Axlon compatible RAM banking in the XL series computers

9.  80 column display!

10.  Released the 1090XL so as to upgrade the 8 bit computers that were without the internal expansion slots in #2.

 

[xxl]

11 hours ago, pcrow said:

There's really no good way to use the machine without access to the hardware registers.

With MapRAM you can use hardware registers and RAM that is covered by them because this RAM is mapped similarly to ROM with SelfTest. Blocking it was ridiculous.

[Rybags]

800 should have been more open architecture, as in slots closer to PBI capability.

 

400 probably would have been better done XEGS style with keyboard optional later purchase on a low end model.

 

XL needed to be sooner.  They could have gone single board for the huge production saving cost probably by mid 1981.

 

Amiga - instead of dealing with them and slipping a loan for their chipset they'd have done better to just purchase the company outright.

 

ST - the release spec of the ST should have been the STe - as in DMA sound, blitter, fine scrolling and 4096 colours at the least.  That way it would practically match the Amiga.

 

Falcon - needed to be sooner and needed to be full 32-bit.

 

Jaguar - maybe hold off release at bit and do it as a CD system from the onset.

[zzip]

Atari should have committed to a single path for 2600 successor.   First it was going to be the 400, then the 5200, then the 7800 and XEGS.   And in the interim they were talking to Nintendo about releasing the NES as an Atari-branded console.   That's insane!   Since they put so much marketing behind the 5200 as the 2600 successor,  that's the console they should have stuck with through the 3rd generation.   Fix the controller issues, make a cost-reduced "slim" model.  And release exclusive games to make people want it!    Not only was much of its library a rehash of what the 2600 had,  the AtariSoft effort was porting Atari's best games to every system under the sun.  And some of those ports were as good as or even better than the ports on Atari's own hardware!    How are they supposed to sell their own hardware?

 

The 7800 - send it back to the drawing board for enhancements so it could be released a few years down the line as a true 5200 successor instead of a panic move.   Having Atari's most loyal customers spend a hefty amount on 5200 hardware only to kill it after 18 mos was extremely bad for goodwill.   If the internet was around back then, that would have been a huge gaming controversy.

 

Just because your competition is doing something doesn't mean you should too.    Atari and the other guys were copying each other's bad ideas instead of checking with the customers.  Did the customers want a keypad controller, or did you copy it from Mattel?   Did consumers really want a mountain of pricey peripherals that are only supported by a couple games each?  Or were you just locked in an arms race with Coleco and Mattel to see who can produce the most peripherals.

 

Better management--  Because Atari was printing money during the gaming boom years of the early 80s, it wasn't obvious how bad many of the management decisions were until it was too late and Atari became a huge liability to Warner.

 

"Consoles are dead" - Because Tramiels believed this post-crash conventional wisdom (which they helped spread at Commodore),  they neglected the console market for a few years, giving Nintendo a huge opening to exploit and steal Atari's "king of gaming" crown.  Turns out that proprietary computers like the ST/Amiga were actually a dead end but proprietary consoles are still alive and well.

 

There's also a few things outside Atari's control that could have gone differently:

* What if Jack Tramiel was never terminated from Commodore?   What would late 80s/90s Atari be like?

* What if Coleco never showed Donkey Kong running on ADAM?   Would the falling out between Atari and Nintendo have happened?   Would Atari have released the NES console, which would have changed the course of gaming?

* What if Atari actually acquired the Amiga chipset?  What would Atari Corp do with it?   ST design was already far along at this point, so would he have integrated into the ST, delaying release?  Introduce it into a later model ala STe?   Release it as a separate computer?    Or would it become an abandoned footnote like the  AMY soundchip or Abaq Transputer?    And in an alternate universe where Tramiel doesn't leave Commodore,  what would Atari Inc do with it?  Release their planned Amiga based gaming console?   The 1850XLD?   

 

I'm not sure getting the Amiga chipset would have mattered much.   Both Atari and Commodore fought and increasingly uphill battle against the PC market and both gave up on their computer lines around the same time.

 

 

[Hawkeye68]

10 hours ago, cathrynm said:

My view, most of the company destroying mistakes were on the console side. What they might have done is hire top management who 'actually played the games.'  If this had been the case, someone might have noticed that the 5200 joysticks made Pacman, which would have been the killer app for 5200, not quite what it should have been. This was the start of the end.

The overpriced 1200XL and the 5200 controllers were the main signs things had gone disastrously wrong.

 

I'll never forget the first time I tried the 5200 at the local mall.  I tried Super Breakout and just assumed the controllers were broken.  How could Atari, the premier gaming company, have let this happen?  Mind boggling, then and now.

[reifsnyderb]

47 minutes ago, Hawkeye68 said:

The overpriced 1200XL and the 5200 controllers were the main signs things had gone disastrously wrong.

 

I'll never forget the first time I tried the 5200 at the local mall.  I tried Super Breakout and just assumed the controllers were broken.  How could Atari, the premier gaming company, have let this happen?  Mind boggling, then and now.

The 5200 was our first, and only, console.  It always worked well for us.  Eventually, the mylars were replaced in the controllers but it was after years of not being used.  We played lots of PacMan, Star Raiders, and Frogger on that machine.  The only issue I remember was that we had two bad Star Raiders cartridges.  Eventually, we had one that worked.

[chevymad]

How about instead of fighting Commodore to the bottom, they had merged?

[MaxZorin]

I believe all 16-bit computers were a mistake. I think the culprit was... Steve Jobs (even though I liked him), who believed that in the mid-80s you could just "reset" progress and make a computer "from scratch" (Macintosh '84), breaking backward compatibility (with Apple II). The same mistake was later repeated by Jack Tramiel when he made Atari ST and later by Commodore when they made Amiga. In the late 80s they realized they had made a mistake and tried to fix it (Apple IIGS, C65). But it was too late - their 16-bit computers did not sell that well.

 

If you break compatibility, you lose your user / fan base. Some of them waited out the time of 16-bit computers and later bought a 32-bit IBM-compatible.

 

I think Atari should have continued 8-bit series - the XE series should have had a fast 65816, new hi-res graphic modes, POKEY should have been more modern. Just enough...

[CommodoreDecker]

Back in the day, my only gripe was the lack of new game titles that the C64, then later with the NES and others would enjoy.  Atari computer marketshare going down only created a possibly avoidable spiral if more recent games would have been ported.  

 

The 8-bits did need another hardware upgrade - either a better POKEY or improved graphics coprocessor.  I recall the 1400XL and 1450XLD and both looked sweet in the catalog, but those were shelved... would the intended voice synthesizer have saved the 1400 line? 

 

The 7800 should have had POKEY built in to allure more game developers, and the 7800 should have had composite output by 1984 as well.

 

The 5200 would likely have been clobbered by the crash anyway - computers were overtaking consoles being one rarely sung factor behind the crash - but if the joysticks weren't poorly built, more games made to take advantage of them instead of just the usual arcade style games that didn't need the sensitivity, and - if nothing better - giving the power brick a user-accessible fuse replacement, since that otherwise innovative all-in-one RF switchbox had the teensy problem of shorting if you don't plug things in a certain order (read: "into the outlet as the very last step", but that's not too far off from how some people will look at an outlet and find the nearest convenient fork to shove in as well.  😕  )

 

Didn't Commodore control microchip production, hence the C64 being able to be sold less? Were there no other producers of the 650x CPUs? (I just looked up and that was confirmed.)  The price undercutting didn't help and, IMHO, the C64's only strengths were the SID and more sprites.  It had higher graphics resolution, sure, but the color palette wasn't as anywhere as good as Atari's. "Masters of Time" is a great example of C64 being a net-worse experience with limited color palette and, yeah, the Atari seemed to have the better music despite the limited octave range compared to SID.

 

Which reminds, games like "Panther" that did come across better on the C64 often felt like cheaper conversions for the A8 when it came to visuals like the HUD panel on the lower portion of the screen. A8 was capable of far better. At least it had a good soundtrack.  Then come the likes of "Castle Wolfenstein", where the C64 version looked and sounded great but the A8 version is almost as bad as the Apple version!  But that depended on the games, for which those generally were the best on the systems directly developed for - any port on another system always seemed to be less polished.

 

All in all, the Atari systems generally got more polished versions of games. Few new ones.