116 replies to this topic

[Retro Rogue]

wood_jl, on Sat May 16, 2009 12:09 PM, said:

Interesting comments, if somewhat slanted.

The Tramiels never knew - or cared - anything about the videogame market.

That's not true, they were counting on Atari's home console portion of consumer to keep Atari Corp. afloat during the interim period before the ST was able to be on the market. Likewise they needed them to help work off the mass amounts of Atari Inc. debt they took on as part of the deal from Warner.

Quote

That is why they foolishly shelved the 7800 for years - and only revived the game consoles (7800, 2600jr, XEgs, and ultimately - Jaguar) to make a go of it.

Again, that's not true. They started working on the jr program just after taking over and evaluating all the projects, etc. they had inherited, and in fact test marketed it during the fall of '85. They were also in on again and off again negotiations with GCC for the 7800 during that time as well, with Warner pushing them to be on. When the Jr,'s test marketing was a success, they decided to move ahead with the 7800. Remember, the NES was only test marketed that Winter of '85 as well in a small local market. It wasn't a success yet, and in fact didn't become the market juggernaut success that everyone remembers until several years later.

Quote

If they had released the 7800 years earlier, there is a chance they would have met with greater success. The NES was, obviously, a fierce competitor. I think the videogame market was just foreign to them.

Who's to say? There's also a chance it could have been yet another market casualty of the time. There's a reason NOA had to go through so many well documented hoops to get their console on the market, and why store chains and distributors put up those hoops (and walls). Part of the NES's success wasn't just as a console, but in Nintendo's reforging of the practices surrounding the video games business as well. Licensing, distribution, etc.

Quote

As any Atarian worth his salt knows, cheap was a Tramiel attribute, rather than an Atari one.

I wouldn't say that's entirely true either. Warner was pretty good at canning Atari projects in favor of the bottom line as well, some that would have been great products. That was the problem during the Warner years, there was a lot of dual management going on.

[philipj]

kevincal, on Sat May 16, 2009 11:58 AM, said:

It was the right move to go with the Jaguar. Atari simply should just NOT have pushed the whole 64-bit thing... They should have just called it 32-bit and called it a day... Even if, super technically, it IS 64-bit, in some ways...

I agree... I think SNK labeled their Neo Geo as a 16bit even though it was capable doing 24bit task... Or so I read somewhere.

[kskunk]

If you want to be a little more cynical about it, you could say that the Jaguar did exactly what it was intended to.

Before the Jaguar announcement, Atari was on its last legs, in debt, and the stock price was in the dumps. Then the Jaguar hype began. The stock price shot up (for a while), they were able to partner up with businesses that normally wouldn't have worked with a failing company (IBM, TWI), and they survived long enough to beat Sega in court and collect $90M in cash.

So if you believe Atari was already doomed by 1991 (I do), then the Jaguar was a great example of taking lemons and making lemonade.

By the time the Tramiels shut down Atari in 1996, the financial position of the company was healthy enough to allow them to completely liquidate their holdings and walk away with millions of dollars. Without the Jaguar, they would have been bankrupt.

It's easy to armchair quarterback, but I think it's remarkable Atari shipped hundreds of thousands of Jaguars and millions of game cartridges, considering their business reputation and financial situation. They were a skeleton of a company with under 100 people for most of the Jaguar's life. Comparing them to Sony or Sega is comparing David to Goliath.

From that perspective, I don't think the Panther could have worked. Its specs were inferior in almost every way to its 16-bit competitors. The Jaguar on the other hand, had amazing specs for its day. Those specs helped prop up Atari stock long enough to engineer a somewhat happier ending for Atari.

- KS

[kool kitty89]

You guys may wan to check out this thread: http://www.atariage.com/forums/index.php?s...119048&st=0

I think the circumstances with Sega were quite different (at least for the Dreamcast), there had been serious communication issues between SoJ and SoA (plus all other westen subsidiaries) from arround late 1993 onwards, this had a huge contribution to the massive problems in the 5th gen for sega. It seems like SoJ stopped paying attention to Sega of America's suggestions (whihc had led them to such success in the NA market with the Genesis despit the failure in the Japanese market), SoA knew the Saturn was a mess and had suggested Silicon Graphics' MIPS chipset instead (N64...) and SoJ engeneers rejected it, SoJ later rejected a pretty fair offer set op by SoJ for a deal to partner with Sony on the PlayStation...

Too add to this, while the 32x was largely an SoA project, the originally planned SoJ proposal for a (very moderately) improved Genesis in a non-add-on form was even worse; and in SoA's defence they were supposed to have at least a solid Year with the 32x in the US prior to the Saturn's launch (with continued overlapping support of the Genesis/CD/32x through the transition to the Saturn), had they known SoJ planned to make an early release to one-up Sony they probably would have scrapped the 32x.

To make things worse SoJ discontinued support on the Genesis and it's add-ons in 1995 while the market was still quite strong, and thus hurt the transition further and alienated many loyal customers. (and may very well have been the move that allowd the SNES to outsell them)

In many ways the Saturn was similar to the Jag, albeit not as severe. With complex, difficult to rpgram for hardware, a rushed release, poor development tools, and a host of other issues including Bernie Stolar's decisions of "quality control" and the decision to discontinue the console prematurely. (and not only give no overlap with the Dreamcast, but also leave a huge gap in time before the New console was available)

And throught this period there were a number of Sega executives that were pusing to drop hardware and go to software, whihc had a significant impact on the Dreamcast's early dismissal. (in particular Okawa)


No there were certainly still issues with Sega durring the Dreamcast, but for the most part it was screaming in the western markets, there was the issue of some 3rd party support concerns due to the (perceived) piracy problems, but it was still selling extremely well in it's sadly short life. It cured all the technical problems with the Saturn, being economical, simple to program for, and very powerful. Had they followed through with the Dreamcast they may have come out ahead of Microsoft and Nintendo in sales, though they still would likely have financial issues. Given all the losses they took from cutting the console short (dropping stock value, further loss of customer loyalty, and lack of direction as a 3rd part software developer) they may very well have been better off that way.

I read a lot of discussion on this (and particupated) over at sega-16.
BTW here's a good related article: http://retro.ign.com...4/974695p1.html (not perfect, but a good article, and by far the best thing written on IGN retro in a while)


I'm not sure what the actual Panther offered as I haven't seen the official specs, and what I've read on it is is very mixed.

Edit: at first glance it may look attractive (fast 68000 CPU, impressive sounding Panther object processor, 32 channel PCM audio etc), but in its existing for it was crap, most importantly due to only 32 kB :shock: of RAM, to be shared for both main and video!

You could perhaps create a cut-down version of the Jaguar (with backwards compatibility planned), to launch a system sooner, but I don't know.


On the issue of the Jaguar, the biggest issues were the hardware, it needed another year before a market release to fix the bugs, finetune the hardware and get the RISC's up to the full 40 MHz, take care of the 68000 issue (most simply replace it with a 32-bit processor, another Jag RISC, a MIPS, ARM, or a 68020/030), add more ram (which would cost less than 1/3 a year later) and offer more cartridge address space. (which won't be a problem with 32-bit addressing) Also very importantly, make some kick ass development kits and offer them to any developers interested. (I'd still keep the CD as an add-on, at least initially, to keep the price point low)

There are other issues like marketing, the controllers, and asthetics of the Jag CD and the cartridges, adding a RAM expansion port, but these are far less important than curing the aformentioned issues which both limited the system's capabilities and (even more importantly) greatly limited the software library of the console.

Edited by kool kitty89, Sat May 16, 2009 11:39 PM.

[kool kitty89]

philipj, on Sat May 16, 2009 7:58 PM, said:

I agree... I think SNK labeled their Neo Geo as a 16bit even though it was capable doing 24bit task... Or so I read somewhere.

Neo Geo labeled their console as "24-bit" often quoting 16+8 = 24 (68k + Z80), which is just wrong and silly tbh. (and is the kind of crap many people assume is going on with the Jaguar... 32+32, or the highly erronious "four 16-bit processors")

The Neo Geo's got the same two main porcessors as the Genesis (and several arcade boards), a 68000 and a Z80 coprocessor. (normally the sound CPU) Technically speaking the 68000 is a 32-bit processor (it's internal registers are 32 bits wide), but its on a 16-bit data bus and a 24-bit address bus. (just like Intel's 386SX) The 68k was capable of performing 32-bit operations. Perhaps the "24-bit" thing you remember is the issue with foreward incompatible (with the 68020 etc) "24-bit" software which used the upper byte for something other than addressing, thus failing on an otherwise compatible 68020 with a 32-bit address bus. (though the cut down 68EC020 with only 24-bit addressing avoided this problem)

By that kind of reasoning of the 68k being a 16-bit processor, the SNES's 65816 would be an 8-bit (8/16-bit) processor due to the 16-bit internals and 8-bit data bus, the 32x would be 16-bit (16/32-bit) with its data bus, the N64's CPU would be 32-bit (32/64-bit) due to the 32-bit data bus.

So the only thing 24-bit about the Neo is the 24-bit addressing, limiting it to 16 MB of addressable memory. Actually this leads to an interesting question, how did the the Neo Geo address the massive amount of ROM with the limited address space? Integral bankswitching? (which would be odd due to bankswitching supposedly increasing the normal 330 Mbit/41.25 MB limit to 716 Mbit/89.5 MB)

Edited by kool kitty89, Sun May 17, 2009 3:01 AM.

[kskunk]

I hope you don't mind me playing devil's advocate. I'm an armchair video game console designer (like everybody else here ), so this is a fun discussion for me.

kool kitty89, on Sat May 16, 2009 10:02 PM, said:

On the issue of the Jaguar, the biggest issues were the hardware, it needed another year before a market release to fix the bugs

I think everyone agrees about a bug fixed Jaguar being a better Jaguar, but it's not clear that Atari could have survived that long. They were already in financial straits and the fact that they were releasing in 1993 (although they barely did), hypothetically getting a head start on Sega and Sony, helped them secure the additional investment required to launch. The Jaguar looked good next to a SNES, but put it next to a Playstation in 1994 and it's a lot harder sell for investors.

Quote

finetune the hardware and get the RISC's up to the full 40 MHz, take care of the 68000 issue (most simply replace it with a 32-bit processor, another Jag RISC, a MIPS, ARM, or a 68020/030)

Another Jag RISC would have killed the overloaded memory bus, unless you add cache to the RISC, like they did in Jag 2. That's a significant design change, not a bug fix. MIPS and ARM are not cheap, and below you mention that it needs to stay cheap.

The Jag is what it is mainly because it was cheap to make that way. Atari considered a 68020/30 -- the Jaguar chipset has support for either. They also considered NO 68000 at all, because even a 68K was considered pretty expensive for such a cheap console.

Quote

add more ram (which would cost less than 1/3 a year later)

That's not how RAM prices worked back then. RAM prices held steady from 1992-1995 at about $3/megabit, or $48 for the 2MB in the Jaguar. The RAM in the Jaguar was its most expensive single component.

(Source: http://phe.rockefell...b/DRAM/dram.htm )

Quote

and offer more cartridge address space. (which won't be a problem with 32-bit addressing)

This would have been nice, but note that there were no bank switched games for the Jaguar. One reason is that large ROMs were quite expensive, especially compared to a CD. Atari even reduced Cybermorph to 1MB to save money.

Put yourself in the shoes of a game developer. 3DO says, 'Hey, we'll charge you $3 per disc.' Atari says, 'Hey, we'll charge you $12 for a 2MB cartridge or $16 for a 4MB cartridge. Um, you want 16MB? Sure, that'll be $40.' The stores say, 'We'll pay you $36 for each game we sell at $59'. Not gonna happen inside the Jaguar's life span...

Quote

Also very importantly, make some kick ass development kits and offer them to any developers interested. (I'd still keep the CD as an add-on, at least initially, to keep the price point low)

Dev tools were key.

Quote

There are other issues like marketing, the controllers, and asthetics of the Jag CD and the cartridges, adding a RAM expansion port, but these are far less important than curing the aformentioned issues which both limited the system's capabilities and (even more importantly) greatly limited the software library of the console.

A RAM expansion port only makes sense if you're using the CD, right? The N64 could expand its main memory, but only because it used a unique 9-bit RAMBUS architecture. In the Jaguar, your best hope for memory expansion would be like the Saturn's RAM expansion, which puts additional memory on a slow cart bus, not the fast main memory bus. And if you are already a cart, you don't need more slow cart memory -- just use your ROMs for that!

Quote

Actually this leads to an interesting question, how did the the Neo Geo address the massive amount of ROM with the limited address space? Integral bankswitching? (which would be odd due to bankswitching supposedly increasing the normal 330 Mbit/41.25 MB limit to 716 Mbit/89.5 MB)

The Neo Geo has 5 buses on the cartridge. The graphics address bus is 25-bit (256 Mbit) and the 68K address bus is 24-bit (128 Mbit). Bank switching was also used for graphics in a few later games.

The Neo Geo is a good example of what happens when you let your engineers go wild.

- KS

Edited by kskunk, Sun May 17, 2009 1:03 PM.

[kool kitty89]

kskunk, on Sun May 17, 2009 11:59 AM, said:

I think everyone agrees about a bug fixed Jaguar being a better Jaguar, but it's not clear that Atari could have survived that long. They were already in financial straits and the fact that they were releasing in 1993 (although they barely did), hypothetically getting a head start on Sega and Sony, helped them secure the additional investment required to launch. The Jaguar looked good next to a SNES, but put it next to a Playstation in 1994 and it's a lot harder sell for investors.

What about those lawsuits that resulted in Atari getting a bunch of money in the mid-90's, and allowing the Tramiels to sell and be quite well off. Did these hinge on the Jaguar? (I can't remember what the one with Sega was apecifically, but did it have to do with the Jag? and would they not have had the funds to win those cases without the small profits from the Jag?)

Quote

Another Jag RISC would have killed the overloaded memory bus, unless you add cache to the RISC, like they did in Jag 2. That's a significant design change, not a bug fix. MIPS and ARM are not cheap, and below you mention that it needs to stay cheap.

The Playstation wasn't released in the West until mid 1995 (the saturn slightly earlier), so definitely get the Jag out before the 1994 christmas season, probably somewhere arround the Saturn or PSX releases in Japan. They only direct competition in the western markets at the time would still be the SNES, Genesis+CD+32x, and the 3DO. (and the CD-32, though very minor)

Tom's got a 4kB cache, and Jerry has 8kB, and one os Gorf's suggestions was to add another Tom, but replace the Object processor and blitter simicon with 16 kB of cache, using it as a "GLP" like the Jag II. (and the 020/030 or EC030 have caches, thogh small) Though he also mentioned using the 68k for game logic and AI on it's own bus and 64 kB, and just trhowing out a CPU entirely and including a small unified cache. (which would be the most cost effective)

See this thread, I mentioned previously: http://www.atariage.com/forums/index.php?s...119048&st=0
Specifically these posts:

Gorf, on Fri Jan 4, 2008 8:13 PM, said:

link

Gorf, on Wed Jan 23, 2008 7:27 AM, said:

Gorf, on Tue Jan 8, 2008 12:11 PM, said:

Quote

The Jag is what it is mainly because it was cheap to make that way. Atari considered a 68020/30 -- the Jaguar chipset has support for either. They also considered NO 68000 at all, because even a 68K was considered pretty expensive for such a cheap console.

Quote

add more ram (which would cost less than 1/3 a year later)

That's not how RAM prices worked back then. RAM prices held steady from 1992-1995 at about $3/megabit, or $48 for the 2MB in the Jaguar. The RAM in the Jaguar was its most expensive single component.

(Source: http://phe.rockefell...b/DRAM/dram.htm )

Sorry, I was basing that on this:

downix, on Wed Jan 9, 2008 1:18 PM, said:

Intel style CPU would never have worked. BE vs LE here. The logic to do the adaptation would add an extra $75 to the bottom line.

The 68000 in 1993 was $21, the 68020 in 1994 was $25, not a dramatic difference in price. Also, 2MB of RAM in 1993 cost $15, 4MB in '94 $8.

I still have my Digikey catalog collection

-from that same above thread, and note someone responded that the Hitachi 68000 should be significantly cheaper.

So in that case, keep the ram the same, or only slightly increased (2.5-3 MB), and add a RAM expansion slot to allow later expansion, if this was at all possible.

Quote

This would have been nice, but note that there were no bank switched games for the Jaguar. One reason is that large ROMs were quite expensive, especially compared to a CD. Atari even reduced Cybermorph to 1MB to save money.

Put yourself in the shoes of a game developer. 3DO says, 'Hey, we'll charge you $3 per disc.' Atari says, 'Hey, we'll charge you $12 for a 2MB cartridge or $16 for a 4MB cartridge. Um, you want 16MB? Sure, that'll be $40.' The stores say, 'We'll pay you $36 for each game we sell at $59'. Not gonna happen inside the Jaguar's life span...

I really don't think they could afford to add a CD drive, even with the later release though, and offering the add-on might not even be a good idea to bother with at all (might want to just wait for the Jag II for discs). 6MB is really limiting though, and assuming th system is more sucessful and makes it past 1996, games of 8-12 MB would be reasonable and necesary to stay compeditive. However, you could still probably manage that with 24-bit addressing (which you'd have with the 68EC020), you mentioned 8 MB being reserved for RAM addressing, so couldn't you cut that to 4 MB, and bring ROM up to 12 MB (though you'd have 10 MB for carts with the 2 MB reserved for boot ROM)
And I don't know what the addressing would be with the unifide cache (no 68k) layout, that could have changed things though.

Quote

Quote

Also very importantly, make some kick ass development kits and offer them to any developers interested. (I'd still keep the CD as an add-on, at least initially, to keep the price point low)

Dev tools were key.

Yes, certainly, even with the current bugs and limitations, a good SDK would have made a huge difference. THe problem is that would take more time too, so if you're stuck with that, you might as well fix some harware problems (at least the most significant ones, the bugs, and the CPU issue -cheapest solutions are remove it completely, or use a 68EC020). Making a good kit with all the problems would likely be quite a task and have to include a lot of workarrounds to address this. With the (major) issues corrected, creating a decent SDK should be easier as well. (there would be a decided advantage of having something familiar like the 020 in there too, opposed to just using Tom and Jerry alone, or having another Jag RISC)

Quote

A RAM expansion port only makes sense if you're using the CD, right? The N64 could expand its main memory, but only because it used a unique 9-bit RAMBUS architecture. In the Jaguar, your best hope for memory expansion would be like the Saturn's RAM expansion, which puts additional memory on a slow cart bus, not the fast main memory bus. And if you are already a cart, you don't need more slow cart memory -- just use your ROMs for that!

There's no way to address an external RAM expansion to the main bus? Could there have been any modifications to the system to allow this, whithout excess added cost.

But running out or RAM and instead storing the data decompressed into ROM forces much more ROM to be used (which is expensive as you mention), so a little RAM can go a long way in that respect, as I understand it.
I had a conversation with Chilly Willy (currently programming Wolf32x) on this at sega-16, and he mentioned that a big reason for the 32x (and SNES) versions were lacking so much (including the textures on the SNES) was due to RAM limitations, and you could have included it all for the 32x (only 256 k ram), but with all that decompressed data in ROM, it would take a cart around 32 MB in size, but only around 3 MB with 2 MB of ram available. So that would allow games that would be far too expensive to put on cart, and impossible on CD. (wihout cutting the game down a lot)
and a note: the SNES Doom was even worse off as adding more ROM was simply not an option, the Super FX 2 chip only supported 2 MB of ROM address space.

Actually some SNES offered a solution to this, there were several on-cart chips used on the console to allow data to be decompressed "on the fly" and streamed. (this was used on many large games, already 32Mbit/4 MB in size, and I think all 48Mbit/6 MB games)
So possibly including something like this onboard the Jaguar would be something to considder too.

Edited by kool kitty89, Sun May 17, 2009 9:02 PM.

[kool kitty89]

Ok, I've already corrected myself on a couple things. Going by those DRAM figures, the price of 1 MB chips went up in 94, and 4 MB in 95 (though not as sharply). SO in 1994 it would actually be almost as expensive to use 3x 1 MB chips as 1x 4MB chip. )though the difference was particularly high that year in favor of the 4MB cost, and all ram cost fell sharply in '96)
So we're practically stuck with the 2 MB. (and I'm still unsure on the expansion possibilities)

And thinking it over, having a 68k series there really is an advantage, particularly with the programming language issues you mentioned in the SFIII thread. So the 68EC020 is probably the best option for this in terms of cost, and removng many of the limitation imposed by the 68k. (faster, 2x the bandwidth, and a cache) The addressing issue was still there, but you mentioned bankswaitching could be used is need be, and keeping games memory sizes small and cheap is certainly important. (there is that decompression thing to considder though)

One problem though, is that I think the 68020 only went up to 33 MHz, so if Tom and Jerry are up to the full 40 MHz, it'll need a seperate clock, a master clock that supports both speeds (obviously varying depending on the 020 model used), or be forced to run the 020 at 1/2 the speed of the RISC's, 20 MHz. (which wouldn't necessarily be that big a deal)
Otherwise you could move up the the 030, which was offered in 40 MHz, more specifically the cost reduced 68EC030 lacking the MMU (though featuring a 32-bit address bus) could be used, but that may cause cost issues.

Edited by kool kitty89, Mon May 18, 2009 12:52 AM.

[mcjakeqcool]

Atari should have releaced the Panther as a 32 bit console, and marketed it as a 16 bit console, that way it'll be a ahead of it's time 16 bit console, rather then a before it's time 32 bit console. That would have given Atari the advantage in the 16 bit wars. Strange stuff, I know, but it works, or at least would have worked.

Edited by mcjakeqcool, Mon May 18, 2009 9:25 AM.

[kool kitty89]

As mentioned the existing panther was fundementally flawed in its design.

However, as a note to the "bits" thing, that is really a very qague comparison and also with no definitive satandard with marketing.

Technically speaking the Motorola 68000 (of the Genesis, Jaguar, Amiga, ST, Neo Geo, X68000 etc) is a 16/32-bit processor, that is a true 32-bit core (with 32-bit wide registers) that's connected to a 16-bit external data bus (and was quite capable of performing 32-bit operations). (plus the address bus is 24-bit but that's a seperate issue).
However, by this logic the SNES's (and Apple IIgs etc) 65816 would be 8/16-bit (8-bit in the way the 68k is 16-bit) due to it's 8-bit data bus, and the N64's CPU is 32/64-bit with 64-bit internals connected to a 32-bit data bus. Even the SH-2's in the 32x would be "16-bit" due to their data bus.
The Dreamcast has a 32-bit CPU on a 64-bit bus (so this would be 64/32, "64-bit" by the previous reasoning); likewise the Jaguar's "Tom" RISC processor is 32-bit connected to the 64-bit data bus so also 64/32 and as "64-bit" as the Genesis is "16-bit."

However all this talk has nothing directly to do with processor power, there's a large number of other characteristics that determines that. (the simpler of which include clock speed and cycle efficiency) The 68k may be 16/32-bit but that doesn't make it more powerful than the 65816, the latter of which is more cycle efficient, and thus actually a bit faster at the same clock speed. (though not nearly efficient enough to make up for the SNES's compared to the Genesis which was more than 2x the clock speed) Though with RAM of similar speeds the 816 is at a disadvantage with 1/2 the bus width. (also both feature 24-bit addressing -16 MB max addressible memory)

Hell, the Atari 2600 is 8-bit (it's 6507 is a low cost version of the popular 6502, a derivative of which was also used in the NES among others), but noone would put the 2600 in the same class as the NES or Master System of the "8-bit" generation. (which is why refrencing the 1st, 2ns, 3rd, etc generations is more accurate)


And until the last couple years having a full 64-bit CPU was rare for common applications (ie home computers), and 32-bit ones are still very common (and pretty much all standard pack-in OS's are 32-bit)

In fact, up through current gen, the only consoles to feature 64-bit CPU's would be the N64 (internal, on a 32-bit bus), the PS2, and I think maybe the PS3. (the cell processor may be built with 32-bit cores)
The Dreamcast, Xbox, GameCube/Wii, and 360 all have 32-bit CPU cores. Generally speaking there' very little use for these systems' CPUs to perform 64-bit precision operations. In fact the N64 uses practically none (and the major emulators actually trap/discard the few 64-bit subroutines that are used), had there been a 32-bit MIPS processor available with the same performance as the R-4300 at a lower cost, Nintendo probably would have taken it, but the next step down is the PSX's R-3000 series which topped at less than 1/2 the speed.

Edited by kool kitty89, Tue May 19, 2009 1:28 AM.

[kool kitty89]

Ok, Atari did win that big settlement with Sega in 1994, but I you really think they were in a dire position financially (I really don't know specifics on this) and Atari really needed a new product to support them, but still not rush the Jag, we need some kind of intrime product. (for '91-94) -as mentioned the Jag's important effect was chaning their immage and I beleive improving company stock value (I don't think too much profit was directly made through the Jag)

The Panther could have been that, if it wasn't terrible.

Atari's homw computer market seems to have been struggling (I don't know as much about this aspect of the company, though along with commodore they were being choked out if the market -especially in the US- by PC's with Apple filling in most of the rest). So the Falcon may not be the answer, I know support was dropped early (supposedly to focus on the Jag), but I don't know how it could have done, and Gorf (and some others) have mentioned its limitations as well. (forcing the 68030 onto a 16-bit bus) Along with the ST line (and STE) and the TT030 (which lacked the Falcon'sbus limitations) are something to considder, but it seems like they were getting driven out of the market, and had thus decided to leave home computers.
Still, cutting support so quickly probably wasn't a very good idea, even if there weren't to be new developments thay could have at least supported the current machines (and making minor improvements +cost reductions) as long as it was profitable. (not to mention there were things like the Falcon 040 in the works, but who knows how those would turn out, though many were cancelled, at least in part, to focus on the Jaguar)

As the Panther really doesn't look like a viable option, and designing an intrim console based on cu-down Jag hardware may be impractical due to the same hardware issues as the Jag. (and including backward compatibility with the Jag may have limited it)

Another possible option would be to develop a console out of their existing computer hardware, though I'm not sure how that would work out. (maybe a parallel development of the STe or Falcon, or a combination of them, though the lower cost of the 68000 based ST/STe/MegaSTE line would probably be preferable, perhaps cut it down in certain areas and make some modifications in others like color capabilities -without having to use software tricks, or -possibly more important, the sound hardware, maybe add something like the Apple IIgs had, or maybe an FM synth chip -which would probably be pretty cheap)
The STe/MegaSTE/Falcon line could (and did) use controllers almost identical to the Jag, using the same DE-15 port, so the controllers may have gotten modified (ie Pro controller) prior to the Jag's release.

Had they gone with the 030 based Falcon, they could have marked it as a "32-bit" system (though technically the Falcon's CPU arranemment made it no more 32-bit than the 68000), the Falcon seems to have pretty good specs for a game console too, so long as it was stripped down to console necessities to make it cost compeditive. (probably using a more integrated, compact board too) The audio seems pretty good too, something to put the Genesis and SNES to shame. (of course, given the Falcon's development and release timeline, you'd have to be doing this console a a parallel design if you wanted it released in the same timeline as the Panther was to be -late '91/early-mid '92) As long as you could strike while the 16-bit war was still hot I think they'd have been good, to make an impact with a "superior/next-gen" image they'd need to beat the 3DO to the market. (which you'd want to do anyway to give their system enough time before bring the Jag out, preferably at least 2 years)

I don't think a TT030 derivative would be a very good choice, it would seem to expensive and seems to have a somewhat odd graphics set-up (no blitter, using CPU power instead, possibly like the Panther and 7800), though it would be a "true" 32-bit system. (honestly they probably could have gotten away with that on a Falcon derived console though)


Cancelling the Lynx was also probably a bad choice. Other than the lynx II, they could hve developed an even lower cost, more compact version ("lynx Lite" or something), without backlighting (and of course, a reflective coating on the screen as all un-lit LCD's have) and using only 4 AA's. (honestly, Sega probably should have done this too, even w/out lighting they had a big leap over the Gameboy, and this would have narrowed the gap of the GB's advantages in compactness, battery life, power efficiency, and cost)

Edited by kool kitty89, Tue May 19, 2009 4:46 PM.

[kool kitty89]

Sorry for so many posts in a row, but I was thinking, the 16-bit data bus on the Falcon was included for backwards compatibility with the ST series, but you wouldn't have to wory about that on a dedicated game console, so had a Falcon derived gaming machine been developed, it could very well have featured a 32-bit data bus.

But was the Falcon too late a development to make a vaiable late 4th gen console? (had it been a parallel development it may have been possible, but it would have to be released at or prior to the time of the Falcon) I suppose this is rather like the CD32 concept (particularly as the A1200 was the Falcon's direct competitor) Granted it would probably be more cut down from the Falcon than the CD32 was to the 1200. (and no CD drive)

Assuming they could get it out in time, to keep costs down, you'd probably want to downgrade the CPU to a 68EC020 (you wouldn't need the address space, though I'm not sure how having a seperate MMU would effect costs), cut ram down to 1 MB (or less), and possibly modify the sound and video chips to reduce cost (if practical with time constarints), and strip out the OS ROM (which is unnecessary) use a small amount of boot ROM, and consolidate the board as much as possible, use a minium of external connections, just the A/V ( a common mini DIN of some kind, or board edge like the Jag if it's cheaper, RGB, Composite, S-Video, include RF with an external modulator like the Genesis II), DC power, and controller ports. (maybe an expansion port, but only if cost allows)

If you were to

[mcjakeqcool]

I think that if Atari releaced the Konix Multisystem like they were planning to,it would have been a huge sucsess,would have slipped in perfectly between the 7800/XE and Jag,1990 would have probally been a good releace date.The ammount of bad decinions Atari made post crash of '83. Damn!

[Bill_Loguidice]

Atari was never planning to release the Konix Multisystem. Different companies entirely. The Konix Multisystem was an intriguing concept, but focused too much on peripherals and gimmicks rather than compelling games to ever had been a success in my opinion if it were ever actually close to release, which it really wasn't.

Edited by Bill_Loguidice, Thu May 28, 2009 2:53 PM.

[carmel_andrews]

only konix was going to release the multi system, only connection with atari was that if was designed by the same peeps that did the panther and jaguar

Later on a company called MSU and a taiwanese firm called TXC tried relaunching the konix games system in an updated form

And they all used to work for sir clive (who gave them their big break in the micro industry, along with miles gordon, of MGT/sam coupe fame)

[atarian63]

Bill_Loguidice, on Mon May 11, 2009 10:11 AM, said:

Poopopyo, on Sun May 10, 2009 6:07 PM, said:

If Atari did release the panther they could have raised enough money to not make the Jaguar a complete failure.

Another pre-Jaguar failure might have damaged Atari to the point where they could have never released the Jaguar.

Also, the reality was it ultimately wasn't the bandwagon bashing that hurt the Jaguar (that didn't help of course), it was the lack of software. There were huge gaps between releases and while Atari got decent third party support to be promised, few of those publishers actually delivered games, and the ones that did actually deliver a title or two didn't follow through with additional titles. If the Jaguar had stronger and more frequent releases in its first year, it might have stood a chance as a profitable niche system, but the reality was with the release of the Sony PlayStation in 1995, it was the beginning of the end for all non-3D systems. Sony correctly gambled on the importance of polygons and where gaming was headed.

We were an Atari dealer and when the Jag was release the customers were very excited. What killed it as has already been said is lack of software. A crying shame! Made me feel guilty a bit as we really hyped it.

[mcjakeqcool]

The Konix Multisystem however was the spiritual predessor to flare 1 and 2, the Panther and Jaguar.

[kool kitty89]

mcjakeqcool, on Thu Jun 4, 2009 10:12 AM, said:

The Konix Multisystem however was the spiritual predessor to flare 1 and 2, the Panther and Jaguar.

No the Multisystem was a developemtn of the Flare One computer design, it preceeded it. The Multisystem is fairly dated for the period and has some peculiarities as well, not to mention it used a fairly weak 8086 as the CPU, but again, this is late 80's tech, still probably not really compeditive over the Genesis or SNES though.

By the specs (on wiki) it otherwise may have been better than the convoluted Panther (which only has a CPU and OP and a tiny 32 kB of SRAM on a shared bus) as it's got a blitter, a lot more RAM, and possibly some decent sound hardware, the x86 CPU would still seem a bit of an odd choice though, particularly for Atari who was mainly dealing with the 68k with the ST line. (and of course the Panther featured this as well, as did the Jag, unfortunately)

And overall, I don't see much of an advantage to building a console from the ST/STE, or later, the Falcon. (though that would be getting a bit close to the Jag, depending on how much extra development time they did to the Jag in this scenario)

A cut down ST in the late '80s or maybe a MEGA STE derivative in 90-91 would be interesting. Assuming the 80's one it could be stripped of the OS, taken down to maybe 256 kB of RAM, and had other unnecessary features removed (possibly the low color 640x video modes), the sound would be weak though, the simplest improvement would probably a Yamaha FM syntheses chip, like that of the PC's Adlib/Soundblaster card. (or Sega Genesis, various arcade boards etc). The STe line had the added digital sound chip which would help too. (and they added the blitter which would open up other graphical capabilities)

To directly compete with the Genesis/SNES I think a cut down MEGA STE would probably be the most competitive, 16 MHz 68000, blitter, decent color (good master palette, but somewhat limited on-screen), and a decent amount of RAM, for the console you could probably cut it down to 512 MB, maybe all the way to 256. (though that would be pretty constricting) Of course adding an added sound chip, like the Yamaha mentioned above. (it would have ben nice if the original ST had used POKEY's though, I understand the I/O issues and the utility of the YM2149, but they could still have added a POKEY or two for some nice audio capabilities, and then I'd probably wouldn't suggest an FM chip either)

On that line of thinking, the AMY sound chip probably would have been good for the ST line, particularly in competition with the Amiga, of course that development team was fired when Tramiel bought Atari, so he couldn't use the desing properly in his projects, and the A8-bit idea was a bit misguided to begin with. (odd that AMY was applied to the 8-bitters but deosn't seem to have been tried with the ST, which would be more along the lines of AMY's original intended product, at lease in respects to the 68k CPU)

atarian63, on Thu May 28, 2009 9:29 PM, said:

We were an Atari dealer and when the Jag was release the customers were very excited. What killed it as has already been said is lack of software. A crying shame! Made me feel guilty a bit as we really hyped it.

Interesting, I've heard opposite stories about retailer with the Jag as well, going as far to warn customers about it...

Do you think an Atari console earlier on (in say '90 or '91), succeeding the 7800, would have had that kind of interest? (as in ST based console, which supposedly they came close to doing in the late 80's arround the time the XEGS was launched iirc)

Edited by kool kitty89, Thu Jun 4, 2009 6:43 PM.

[atarian63]

kool kitty89, on Thu Jun 4, 2009 7:35 PM, said:

mcjakeqcool, on Thu Jun 4, 2009 10:12 AM, said:

The Konix Multisystem however was the spiritual predessor to flare 1 and 2, the Panther and Jaguar.

No the Multisystem was a developemtn of the Flare One computer design, it preceeded it. The Multisystem is fairly dated for the period and has some peculiarities as well, not to mention it used a fairly weak 8086 as the CPU, but again, this is late 80's tech, still probably not really compeditive over the Genesis or SNES though.

By the specs (on wiki) it otherwise may have been better than the convoluted Panther (which only has a CPU and OP and a tiny 32 kB of SRAM on a shared bus) as it's got a blitter, a lot more RAM, and possibly some decent sound hardware, the x86 CPU would still seem a bit of an odd choice though, particularly for Atari who was mainly dealing with the 68k with the ST line. (and of course the Panther featured this as well, as did the Jag, unfortunately)

And overall, I don't see much of an advantage to building a console from the ST/STE, or later, the Falcon. (though that would be getting a bit close to the Jag, depending on how much extra development time they did to the Jag in this scenario)

A cut down ST in the late '80s or maybe a MEGA STE derivative in 90-91 would be interesting. Assuming the 80's one it could be stripped of the OS, taken down to maybe 256 kB of RAM, and had other unnecessary features removed (possibly the low color 640x video modes), the sound would be weak though, the simplest improvement would probably a Yamaha FM syntheses chip, like that of the PC's Adlib/Soundblaster card. (or Sega Genesis, various arcade boards etc). The STe line had the added digital sound chip which would help too. (and they added the blitter which would open up other graphical capabilities)

To directly compete with the Genesis/SNES I think a cut down MEGA STE would probably be the most competitive, 16 MHz 68000, blitter, decent color (good master palette, but somewhat limited on-screen), and a decent amount of RAM, for the console you could probably cut it down to 512 MB, maybe all the way to 256. (though that would be pretty constricting) Of course adding an added sound chip, like the Yamaha mentioned above. (it would have ben nice if the original ST had used POKEY's though, I understand the I/O issues and the utility of the YM2149, but they could still have added a POKEY or two for some nice audio capabilities, and then I'd probably wouldn't suggest an FM chip either)

On that line of thinking, the AMY sound chip probably would have been good for the ST line, particularly in competition with the Amiga, of course that development team was fired when Tramiel bought Atari, so he couldn't use the desing properly in his projects, and the A8-bit idea was a bit misguided to begin with. (odd that AMY was applied to the 8-bitters but deosn't seem to have been tried with the ST, which would be more along the lines of AMY's original intended product, at lease in respects to the 68k CPU)

atarian63, on Thu May 28, 2009 9:29 PM, said:

We were an Atari dealer and when the Jag was release the customers were very excited. What killed it as has already been said is lack of software. A crying shame! Made me feel guilty a bit as we really hyped it.

Interesting, I've heard opposite stories about retailer with the Jag as well, going as far to warn customers about it...

Do you think an Atari console earlier on (in say '90 or '91), succeeding the 7800, would have had that kind of interest? (as in ST based console, which supposedly they came close to doing in the late 80's arround the time the XEGS was launched iirc)

We were a mostly Atari dealer, probably the most pro Atari dealer in town. I would imagine other retailers were less excited.We chose to sell it,we could have ignored it.
In the late 80's.. well I think so, just a steep hill to climb with genesis out. Possibly an STE based system. Cart based and a better sound chip than a stock st. No gem or hidden. I just don't know. Neo Geo was big for us too, The Jag seemed the right system, just needed to be sooner with more software.

[kool kitty89]

Curt Vendel mentioned a "project Robin" a while back, supposedly an effort to produce a gaming console based on the ST. (or rather to persuade Tramiel to) And quite a bit earlier than my proposition for a 1990 release.

Curt Vendel, on Sat Apr 17, 2004 12:32 PM, said:

The industry as a whole was moving to computer-gaming, even magazines were changing their names to either have computing added or changed from Video to Computer Gaming... Computers like the C64, Apple //C and ][GS as well as the Atari 800XL and XE's were all very inexpensive (well maybe not the GS) and it appeared at the time video gaming was moving in that direction.... the moment the ST's and Amiga hit the shelves, companies saw them as great gaming platforms...

In fact there was a grass roots effort within Atari by Rob Zydbel and several other programmers who actually went out of their way to port games like Star Raiders, Moon Patrol and many other tried and true Atari titles and licenses to the ST with hopes of getting the Tramiels to make the ST technology into a game platform and their efforts almost convinced the Tramiels... There was project "Robin" which was an ST in an XE case that was in the works...


Curt

He's made some similar posts recently proposing such a system in ~87, basicly launched in place of the XEGS, which seems to have been what Tramiel defaulted to after the "Robin" idea died. (which was worse as it was more in direct competition with the 7800 than the high end ST console would be)

In 1987, you could have the 2600 as the aging budget system (slowly being phased out), the 7800 as the mid-range current mudgets system, and the ST gaming system as a high end console. (maybe allow it to be expandable to a full ST computer as well) The XEGS had the problem of being to close to the 7800 market (among other issues) in terms of capabilities and price range, and used aging hardware. (that was available at a similarly low price as the compuer counterpart)
Something like the XEGS might have been a good idea in place of the 5200 in 1982, but not in 1987...

[Crazyace]

kool kitty89, on Fri Jun 5, 2009 1:35 AM, said:

mcjakeqcool, on Thu Jun 4, 2009 10:12 AM, said:

The Konix Multisystem however was the spiritual predessor to flare 1 and 2, the Panther and Jaguar.

To directly compete with the Genesis/SNES I think a cut down MEGA STE would probably be the most competitive, 16 MHz 68000, blitter, decent color (good master palette, but somewhat limited on-screen), and a decent amount of RAM, for the console you could probably cut it down to 512 MB, maybe all the way to 256. (though that would be pretty constricting) Of course adding an added sound chip, like the Yamaha mentioned above. (it would have ben nice if the original ST had used POKEY's though, I understand the I/O issues and the utility of the YM2149, but they could still have added a POKEY or two for some nice audio capabilities, and then I'd probably wouldn't suggest an FM chip either)

On that line of thinking, the AMY sound chip probably would have been good for the ST line, particularly in competition with the Amiga, of course that development team was fired when Tramiel bought Atari, so he couldn't use the desing properly in his projects, and the A8-bit idea was a bit misguided to begin with. (odd that AMY was applied to the 8-bitters but deosn't seem to have been tried with the ST, which would be more along the lines of AMY's original intended product, at lease in respects to the 68k CPU)

I dont think the MegaSTe video would stand up against either the Genesis or the SNES. The both supported multiple scrolling layers, and more colours ( 64/512 and 256/32768 without tricks ) - I think the TT video would be needed at least ( 320x480 256 colours on falcon , only 320x200 256 needed for TV )
( Maybe another hack would be to 'double' the STe graphics - so you have two independant 320x200x16 colour layers )

kool kitty89, on Fri Jun 5, 2009 5:58 AM, said:

Curt Vendel mentioned a "project Robin" a while back, supposedly an effort to produce a gaming console based on the ST. (or rather to persuade Tramiel to) And quite a bit earlier than my proposition for a 1990 release.

He's made some similar posts recently proposing such a system in ~87, basicly launched in place of the XEGS, which seems to have been what Tramiel defaulted to after the "Robin" idea died. (which was worse as it was more in direct competition with the 7800 than the high end ST console would be)

In 1987, you could have the 2600 as the aging budget system (slowly being phased out), the 7800 as the mid-range current mudgets system, and the ST gaming system as a high end console. (maybe allow it to be expandable to a full ST computer as well) The XEGS had the problem of being to close to the 7800 market (among other issues) in terms of capabilities and price range, and used aging hardware. (that was available at a similarly low price as the compuer counterpart)
Something like the XEGS might have been a good idea in place of the 5200 in 1982, but not in 1987...

A console version of the ST ( with 128K - maybe? ) may have worked competing with the SNES and Master system - but it's a real pity Atari didn't have the Amiga , that could have competed more with the Megadrive as a console.

[Goochman]

Atari's fate was sealed the day it was seperated from the Arcade division. Atari was never going to be a 'business' computer company. If Jack had changed the name to something else - maybe, but very few were going to use Atari as a corporate system. Looking back it was a matter of time before Atari/Commodore/Everyone else was pushed out of the computer business.

How does the arcade divison play into this? Video Game hits - When Sega and Nintendo started producing their own consoles the cool games in the arcades became the hot video games to play at home. Even the 'Atari' arcade division would not release some of their hot titles on the Atari Corp consoles!

Look at the 7800 - it rehased all the 'classic' arcade titles Atari could get their hands on and then some goofy made up names that didnt resonate with the buying public (good game or not, Ninja Golf wasnt going to sell a million).

Here comes the Jaguar - Atari didnt have the $$$ to pay/market new games form small companies. EA and others had no incentive to build stuff for Atari - Sega and Nintendo were going to sell 10x more consoles, and then Sony entered the game.

Atari had no library of titles to produce interest in their home console unit - Even with T2K and AvP were were excellent titles, it wasnt going to sell millions.

Now even if the Arcade division was never seperated - it prob still wouldve failed as it seems that group failed to deliever alot of good titles over the years and finally got squashed as well.

[Crazyace]

I think Atari actually did pretty well with the ST's They sold well, and had a lot of good software written for them. Maybe they could have succeeded with a console version as well. ( I just wish that h/w scrolling had been part of the original ST - rather than the STe, where it was just too late )
Even the Jaguar had quite a few good titles - they just arrived too late to save it, as the PSX and Saturn had raised the bar.

[S1500]

Atari's Coin-Op Division was like that college-educated older brother with a 6-figure income that still won't talk to you at family reunions.

[mcjakeqcool]

Anything would be a better competitor to the SNES and Mega Drive then the 7800 and XEGS, except the 2600 jr (which was a competitor aswell), Atari needed a big next gen console around '89, '90 why they waited until '93 is beyond me.