kool kitty89

I'm using version 1.13 if that makes any difference. (just drag and drop like any other ROM image, but I get a black screen and nothing -even if I try to use the VCS switches or joystick) They didn't own the rights, it was locked up in limbo with Warner, Atari, and GCC. It had been a work for hire contract with Warner that was then fed to Atari Inc as an actual commercial product. The issue was that GCC hadn't been paid for their R&D fees for MARIA and the 10 launch games (or 11) and Warner wanted to pass those fees on to Atari Corp even though the contract had been with them directly with the technology fed to Atari Inc as such. (a complex arrangement demonstrating some of the problems with Dual management -I don't think Atari Inc even had a direct say in the matter, ie Warner was going with the 7800 whether Atari Inc fully accepted it or not -let alone any direct influence in the hardware development or software development for the launch games) Yes, it would have been in their best interests to pay GCC and be done with it. (or at very least pass those on as additional loans to Atari Corp) However, it by far would have been even more in Warner's best interests (let alone Tramiel) to manage the split in a formally organized manner with a normal transition with every detail mapped out for what TTL would be getting as well as giving Atari Inc staff sufficient notice of layoffs ahead of time. (or better, a smooth transition to keep Morgan's reorganization going but fold that into Atari Corp -have Tramiel and Morgan formally discuss the entire thing and at very least get TTL management up to speed before leaving them to their own devices -if Tramiel still didn't want to keep Morgan on longer in any capacity, though there's decided advantages to that) Ant it was a waste of time, albeit if they had stuck to arcade centric marketing against the NES, but managed exceptional advertising/marketing/distribution/etc (including kick ass packaging/box art, more attractive cartridges, etc, etc -the lack of the FM add-on was really odd, not even released in Europe though it was in Japan and integrated with the 1987 SMS there in spite of the Mk.III/SMS doing worst in that region and being discontinued before anywhere else -pretty much dropped when the MD came out with development moved over to western releases) But really, Sega should have made note of the arcade status, but focused on general video game advertising in the US. Hell, look at Atari/Coleco/Mattel/Activision/Imagic/etc's advertizing in the early 80s even and there's a TON that's not tied to the arcade at all. (I'd say the majority -even many arcade ports had the games themselves pushed more than the fact they were arcade games -in some cases the arcade aspect wasn't even mentioned) Had Sega had the likes of Katz managing things and building up SoA, they very well could have competed very strongly in the US market and attracted enough 3rd party interest to foil Nintendo's exclusivity strategy. (ie it would have worked against Nintendo if they tried it) They definitely should have pushed the better graphics too, and the sound was weaker but still OK (less than it should have been for the 1985 Mk.III launch, but at least better than TIA, though weaker than the Intellivision/ST/Vectrex/MSX/Spectrum128k/CPC/etc), and releasing the FM add-on at a low price would have been a very nice option around that too. (let alone standardizing it internally in '87 as with Japan -for the add-on, they'd have to use external mixing as with the SG-1000 Mk.III since the cart slot and expansion port lack sound input) Arcade ports, to many, weren't arcade ports at all, but just games. (Alex Kidd was in the arcade anyway -not all the same games mind you, but Lost Stars in the arcade came before Miracle World on the SMS -sidescrolling platform games WERE arcade games so there's no point in distinguishing as such since most genres were in the arcade as well -difficulty was usually modified for the home market though, be it ports, remakes, or original games) The problem was weak marketing all around, they failed to appeal in the way Nintendo did in spite of having competitive hardware, software, and price points. (carts should actually have been cheaper to produce due to the single bus/single ROM design and lower pin count with smaller PCB) The TV ads were weak, the box art was terrible until Tonka's in '88 (Japan was at least as good as Famicom competition though), the carts had no art at all and were a bit bland/clunky overall too, and the one nice thing was the plastic clamshell cases used. (the Japanese MegaDrive adompted those as well and the US Genesis got all of that right too for the most part -especially after Katz came in) Sega even had a larger marketing budget initially if the new articles are accurate. (if Atari had had a 6-9 million marketing budget in 1986 things would have been very different -let alone if some of that went into R&D for games, though even putting 100% into advertising could have meant more and better games a la 3rd party interest and increased sales revenue following that -hype generates developer interest, just look how many devs initially signed onto the Jaguar in spite of Atari's shaky market position) See above, they did fail as such for sure. Arcade games didn't have to be marketed as arcade games alone, and there's a lot of other areas to boost things in general. (let alone direct competition with the superior hardware aspect -Katz did that later on, but the SMS had enough of an edge to push that as well, though less extreme than the Genesis) Again, focusing on the arcade was only one part of the problem, the other was the advertising overall and other aspects of marketing. (the crappy, unappealing box art, distribution network and product tie-ins, etc, etc) The very early "Arcade at Home" advertizements for the Genesis were actually pretty good too, though still sorely lacking the competitive edge that Katz added. (and Coleco, Mattel, and Atari had used to compete with years earlier -especially back to back comparisons as well as breadth claims of a better/more powerful system overall) Why? They had the software, they had the hardware, they had the money, and Nintendo was virtually a nobody on the market just as they were. (Nintendo's initial killer apps were mostly in-house games and many of the Japanese 3rd party games would: 1. be countered by Sega's in-house games and 2. may have been persuaded to Sega's side or license for Sega publishing as with several games on the Genesis in the west -including some games published by 3rd parties in Japan and 1st parties in the US like Technosoft's Thunder Force II; then you've got the US and EU 3rd party developers who wouldn't have been restricted by Nintendo's licensing if they were pulled away by Atari and/or Sega instead -thus Nintendo would be forced to loosen their policies or risk loosing 3rd party support) NEC made an even bigger mistake than Sega with the PCE later on (quite literally had the potential to be the Sony of the 4th generation), but that's another topic. Except they could have forced Nintendo to never have that stronghold (it would have done nothing but deter US devs from licensed development for the platform -let alone EU devs who already had just that dilemma). Sega did it with the Genesis and not just because of Sonic (that was the clincher after the fact, but Katz had layed the ground work and cut deep into the market with such at Nintendo's absolute peak in NES sales in '89/90 -in a FAR worse position than back in '86/87 other than Nintendo's hardware being older, but it WAS already 3 years old in '86) We don't know that for sure. No, Sega won due to those advantages and (more importantly) the marketing/management instrumental to the market as well as competition really screwing up and leaving a hole for them to poke into. NEC had the resources, the hardware, and the initial Japanese support to roar into the US rather like Sony did with the PSX (should have done very well in Europe as well), but they screwed up on some many levels it's not even funny. (they made the console bulkier but didn't even add 2 controller ports -let alone integrate the full 5 ports -which would have been awesome, even if an optional deluxe model, they didn't slash the price -very integrated hardware with vertical integration and resources to sell at a loss, they failed to push saturation marketing in the US -Europe is a bit different and could cater better to the strategy NEC used in Japan but with tighter pricing, probably keep the compact form factor as well for Europe, and they failed to push through Nintendo's licensing agreements -the biggest loophole would be to license/publish 3rd party titles under their own label and offer such to both Japanese and western developers as an alternative to licensed 3rd party publishing -offer royalty deals to treat such an NEC published game more or less as a 3rd party game, and push for a build-up of western software development in general as well as staple franchises/genres needed in the regions -like Sega did with Sports) And no, Atari Corp could have run with the success they had in the late 80s and struck while the iron was hot with a reasonably competitive 4th gen console by 1989 with a competitive price point and marketing to match taking advantage of their much stronger revenue. (given the drastically declining 7800 sales in '89 they could have even pulled back much of the resources going into it at the time -and started phasing it out in 1990) They had the potential to push into Europe quite favorably as well with the right hardware at the right price, but the trick in the US would be getting 3rd parties to develop for the system without direct commissions: the most direct loophole for Nintendo's licensing of the time would have been to offer royalty agreements to publish under the Atari label and avoid Nintendo's blocking strategy. The other thing would be to continue pushing for computer developers/publishers who still hadn't gotten roped into Nintendo's trap. (and of course, with better management the ST line could have been much healthier -at least in Europe, but perhaps maintaining niche competition much better in the US even, hell they were in a position to start building a good relationship with Tengen/AGames even by that point -they lost Katz and they lost Jack, after than things seemed to fall apart all around -any relations with AGames probably should have been delegated by Katz: hell, he did that for EA and Sega when put in a very tight situation with EA threatening to go unlicensed and SoJ being hostile towards EA's intentions and indignant that they'd reverse engineered their own tools) And, of course, this is in the context of Atari Corp being in the same position they were historically in 1988/89, no other changes prior to that. (like the night and day difference that proper management of the split could have made) OTOH there's no reason Sega couldn't have competed many, many times better than they did with the SMS in the US. (maybe not made #1, but at least come far closer than they had) Again, Nintendo had the advantage of pushing into the market for some 3 years prior to the final September 1986 launch while Sega didn't even push for heavy investment in build-up of US management/marketing or license marketing/distribution to a highly capable toy/entertainment company in the US. (they did the latter in Europe very well while Nintendo made very poor choices in many cases -other than Germany and maybe parts of northern Europe like Holland/Sweeden/etc which were small markets anyway, they screwed up even more in the UK with a very late release) Atari would have flattened Nintendo and Sega if it hadn't been for Warner's horrible management of the split. I was speaking in the context of Atari Games being an insignificant facet of the problems caused by the split. (ie the 7800 would have launched in 1984, Morgan's continued plans would have meant for very organized downsizing in the final steps in reorganization -part of that would mean retaining the most integral console programming staff, Atari Games would could have had very friendly relations -or still been part of AInc if Warner hadn't found any buyer at all or sold off AInc intact, the Rainbow Chipset may have been adapted over the ST and ATG may have been properly folded into Atari Corp R&D, the 8-bit computers would have been marketed better in '84 and the planned peripherals would have been released, etc, etc -and more revenue/funding with diminshing debt much sooner so the company would be healthier on all accounts)

Oh, I don't know...maybe? I can't get the music to play on StellaX...

They didn't have to be in ANY price war to push low prices like that (not saying that wasn't a reason, but not necessarily THE reason). It could have been pushed as such as a general marketing strategy to push for high market share. (a combination of strong advertising and low prices facilitated by high volume and vertical integration -and also cut corners on the hardware wherever they could afford including some questionable quality control) I'd thought CBM had dumped the C64's price during the "war", but in a discussion a while back I got corrected on that: CBM was making profits on every single C64 sold in the US and Europe (not taking failures/returns into account), the margins were smaller than pretty much anyone else had been pushing up to that point (and even more extreme due to the vertical integration -though TI had that as well). The C64 didn't drop below $200 until well after the end of 1983. (I don't think it dropped to $100 until late '84 or maybe '85 -not counting rebates) The C64 demonstrated that from the start with the $595 launch price for a new, relatively technically impressive 8-bit computer with 64 kB of RAM standard. (I'm not sure what the 1200XL launched at, but I think it was closer to $1000, though I think Atari was also pushing for higher profit margins and a more expensive machine in general -quality keyboard, well-built case, etc . . . except it SHOULD have been significantly cheaper to produce than a 48k 800 and that had dropped below $700 by early 1982 -the unreleased 600 should have been significantly cheaper than the 16k 400 in spite of the much better keyboard due to the consolidated single-board design with removal of the heavy aluminum castings -a big mistake to to launch that in any case, a high quality lower-end model that definitely would have undercut the C64 by a large margin at normal profit margins -for Atari- and offered a much more useful machine than the VIC or TI99 for that matter -better software lineup than either of the competition for that matter and the only one with Y/C monitor support -a shame that Atari didn't actually sell Y/C monitors for the line though, at least AFIK) CBM could afford to drop the price of the C64 due to volume production and probably opted to after the promising sales in '82. (they'd already taken a big risk with the initial production ramp-up and stockpiling -let alone the hit taken on hardware failures from factory defects) Of course, the very fact that the competition had left such an opening on the market facilitated their success hugely. (Tandy, Apple, or Atari could have cut in heavily if any of them had pushed things differently -especially Atari with their in-house software and advertizing resources, a shame they weren't pushing TV ads in '81/82 -or even earlier- like they were with the XL line briefly before -and maybe after?- the split in '84 -like the Alan Alda ones -albeit Commodore missed their own opportunities with the PET line, not to mention IBM) Even with the price advantage, a well-established A8 would have made it much tougher in the US or Europe to cut in. (plus you had the Sinclair machines closing up the low end in Europe, let alone if Apple -or Apple II clone makers for that matter- had pushed for a low-cost line of their machines by '81/82 -ie consolidated chipset with a compact/low cost design and more limited expansion -probably more like the CoCo or Atari 800/600XL- and tight profit margins -and turning around to use the same consolidation for high end models as with the Apple IIe -Acorn had a lot more potential in the UK/Europe as well, one of the biggest mistakes was pushing for the heavily cut down Electron rather than making it a fully-functional 2 MHz BBC Micro using a low cost chipset and form factor) The VIC is the one they were selling below cost and doing things that were probably illegal on the market. (though technically, they weren't doing anything that TI themselves weren't capable of matching with a similarly cheap product and selling at a loss -apparently that wasn't a game of chicken TI wanted to partake in though -that and TI didn't have a cheap machine like the VIC on the market, a Z80 based system more like the Clecovision/Sord M5/SC-3000/etc might have fit the bill more though -hell, it might have competed better against the C64 too with higher memory models -the TMS9900 was relatively expensive for the time and also used a fair chunk of board space and traces with the 64 pin DIP) And, of course, those price cuts hurt others on the market not directly tied into the price war. (plus, TI was the only major competitor on the home computer market other than CBM to have vertical integration) The CoCo probably would have been the cheapest computer on the market at the time if it hadn't been for the VIC's price dumping. (as it was, the 4k CoCo model was $299 in 1980 -at least by the holiday season iirc, so quite possibly the first really affordable hoe computer in the US -of course, the Apple II COULD have made a great low-cost computer as it aged -ie by the early 80s- given the simple design with related potential for heavy consolidation as well as the head start on the market with good software support and reputation -hell, even without vertical integration, a cost cut 4k Apple II should have been able to undercut the VIC-20 and still maintain a profit) Plus, the VIC was sold for profit initially too (and only for a loss for a short period in the heat of the price war I believe). It started at the moderate $300 price point in 1981 (at which it wasn't really a better value than the CoCo, let alone Atari 400 -more so with the software of the latter), but it was down to $200 by early '82 iirc and to a mere $100 (or 99) in 1983. (the Atari 400 was also down to $99 by the end of 1983 -though it had been officially discontinued by that point and the 600XL was placed at $200 SRP iirc -not particularly competitive given the C64 selling for the same price and given some of Atari's previous prices and the fact the 600XL should have been much more efficient to produce than the 400, you'd think they could have cut the price lower -maybe more like $150 which is what the Colecovision was at just before the crash iirc) The C64 was in a different price class than the TI99/4a, that was one of TI's problems, they limited their computer to the lower end only when they had potential for so much more. The razor and blade model with the 1st party software was especially odd for a computer though, limiting (or outright discouraging) 3rd party publishing was a very bad move. (Tandy, Apple, and I don't even think Commodore -with the PET- did that, I don't think Atari did -though they didn't provide -as standard or very cheaply/easily at least- the critically needed 3rd party development tools/kits in Europe -IBM definitely didn't either) Hell, if TI HAD wanted to aim at a low-cost computer, they probably should have opted for a cheaper CPU like the Z80 used on many other TMS9918 based systems. (perhaps even licensing the Z80 to produce in-house) That also should have meant having a memory interface slow enough to use commodity DRAM for main memory from the start. (the Z80's refresh counter would help a little with that too) What's your source on that? It's these sort of claims that gave me the wrong idea before. I need to dig though my references again (and some previous discussions tying into this), but $99 for the C64 in 1983 doesn't make sense, the VIC for sure but not the C64. After I had made similar comments a while back, I got corrected on the issue and pointed to some better references along with actual quotes on the market prices in the US and Europe at the time. The C64 didn't drop to $100 SRP until 1984 if not '85. (I'm pretty sure it wasn't $99/100 SRP until 1985 -again, short of any rebate offers) Hell, this print ad points to $215 in 1984, so maybe they hadn't even dropped to $200 in 1983. http://home.insightbb.com/~kguenther6/com64sep84ad.jpg (I know they dropped below $300 in '83, but I'm not sure what they ended with as far as SRP goes -not with rebates)

Huh? I suppose there were different pinouts but the most common SRAMs back then were the 18-pin 1Kx4, 24-pin 2Kx8, and 28-pin 8Kx8. The last two were wide DIPs and were mostly pin-compatible with ROMs of the day, so they would be the natural choice for RAMcarts, although the 8Kx8 was still expensive in 1983 ($40 each in Nov. '83 BYTE ads). I was making some assumptions on that. I'd thought SRAMs were available in flexible word sizes/pinouts like DRAMs of the time (or today for that matter), yes: higher width RAMs are going to take more pins for sure. (but the examples I gave would be those pin counts if available -ie a 4kx1-bit SRAM would be 16 pins, 16kx1-bit would be 18, etc) For some things it wouldn't matter though: like 2k SRAMs up until the dies got small enough to fit in skinny DIPs. (24 pins is the smallest wide DIP available -probably the main reason the AY8910 didn't cut below 24 pins for the sound only 8913 version -even though it includes added signals the 28 pin 8912 lacks and should have been 20 pins if not less) That wouldn't apply to the 512byte chips though sicne those were all skinny DIPs. (later 2k chips were as well -as the NES/Famicom used, though the 7800 didn't AFIK -NES needed 8-bit chips in either case too so they were 24 pin skinny DIPs vs the 7800 which could have used 2 4-bit chips and cut to 20 pins -ie 2 more address lines than 18 pin 1kx4-bit chips) I've seen a lot with 14 pin DIPs (like the Amiga and ST -and I recall seeing 12 pin DRAMs in some TI99/4s or maybe 14 pin as well -you SHOULD only need 12 pins for 16kx1-bit DRAMs though, just like you'd only need 14 for 256kx1 in many Amigas and STs though 16kx4 would be 16 pins and 32kx8 should be 20 pins -same 20 pins for the 256kx4 chips used in some later A500s and I think STs)

Same for keeping the whole system clean: less dirt and dust= better ventilation. For that matter, removing the RF shielding (where not integral) can be beneficial as well. Keeping the ICs clean can also help avoid corrosion in extreme cases. (as long as you clean only with distilled water, alcohol, or just dry cleaning -or dedicated electronics cleaning stuff -though that's mostly the same too) The region you live in will be a factor too. (and/or the conditions of your house -or wherever it's set up or stored)

Yes, and most newer stuff tends to run much hotter too (high clock speeds), there's only a handful of console/home computer ICs from the 70s and early/mid 80s that used heat spreaders, let alone dedicated fans. (totally passive ventilation in many cases at that -at least if the PSU was external) We had a similar discussion on Sega-16 a while back, and a few others on AA before too. Capacitors are the weakest link in old electronics, a few other discrete components (voltage regs, sometimes transistors, diodes, usually not resistors) tend to be more problematic than the LSI chips, but all ICs are perishable. (the silicon lasts forever, but the dopants and such that make up the logic and interconnect -I think more so the interconnect- will degrade over time) Even with very old chips that get no or practically no use, they're probably not going to last hundreds of years in the best cases. (aside from painstaking cryogenic preservation or something ) The smaller the IC process, the sooner the degradation. (take modern high-quality technology and apply it back to single layer 2-5 micron manufacturing and you'd have some really long lasting chips though ) I'm not positive on the 70s/80s tech, but it may very well be possible that a good percentage lasts close to a century. Hell, electrolytic capacitors in such systems (or TVs or other equipment for that matter) are generally LONG past their expected expiration date, yet a very high percentage of systems are still working fine without recap. (especially those that weren't abused -but even then quite a few) Though part of that is probably that the systems are rather tolerant to degrading (namely drying) capacitors vs newer systems which are far more sensitive. (a fair amount of the early/mid 90s Sega systems have notable capacitor problems -especially the game gear) The sheer size of some capacitors used on the earlier models probably plays a role too. (much greater volume to surface area ratio) In terms of hardware that gets regular use and in as far as ICs (from CPUs to ROMs to ROM to discrete logic), normal wear and tear with various risks of voltage spikes, overheating, etc will all be far more likely causes of failure than gradual degradation of the chip's circuitry alone. (like how I destroyed the chips in my light sixer ~9 years ago by using reverse polarity current -unfortunately they blew -quite literally- along with one or 2 games before the voltage regulator gave out -the motherboard is still perfect otherwise after the reg was replaced -swapped in another systems's chips- but I've still yet to find a crappy VCS to complete both systems I have with another set of TIA+RIOT+6507 -the overload blew the package right off the top of the chip for one of them -TIA or RIOT, I forget which)

What is BITD? How I hate acroynms ... http://www.abbreviations.com/BITD #1 rated answer too. (back in the day) There's also software scroling tricks on the VCS, many choppy TIA playfield pixel scrolling (ie in the ~40 pixel wide playfield sided coarse pixels -like Jungle Hunt of Vanguard), but a few cases of really smooth scaling that I don't think is being done with sprites. (Chopper Command, Stargate/Defender 2, etc) A lot more smooth vertical scrolling though. (river raid, etc -vertical scrolling is also done rather well on some colecovision games like pitfall 2) "hardware scrolling" is sort of a blurry definitition too, you could have cases of true v/h scroll registers for a bitmap or character display, or other logic that assists scrolling but isn't complete as such. (I think the C64's horizontal scrolling is a bit like that -assisted, but not fully CPU independent- not sure about ANTIC driven scrolling) You seem to keep being stuck on this SMS US sales figure thing. While I think you might be onto something, it's now evolved from "you suspect based on subjective gut feel" to a "fact". Unless you have NPD to back up the claim that the SMS sold 4 million in the US, no need to pass this off as a fact. There's already enough BS statistics on the net ... no need to add to it. Not trying to pick on you, but there are so many funky "numbers" out there right now as it is. No, I suspect based on the physical market share figures from the '86-89 period (especially '87/88) that points to Sega being well behind atari, but still high enough to be a hell of a lot more than 2 million in the US. (I'd need solid 2600 sales to be absolutely sure, but I'd gotten the impression that the 2600 was selling significantly more than the 7800 throughout that period -anecdotal reports of the SMS being more common is notable as well though) Of course, Sega has never, ever released final sales figures for any of their products, so the 2 million figure would be built up from other information in the first place. (and I haven't seen any reliable sources for that figure either -just like the 5200 and Colecovision hardware sales where it seems likely that both sold at least 2 million, but is vague beyond that) Again, without NPD data from the day, this is speculation. "Marketshare" is one of the most misinterpeted things on the internet unfortunately. Look at the friggen smartphone fanboys quoting smartphone regularly mixing up active users vs. units sold in a a quarter, units sold at retail in a quarter vs units sold overall in a quarter, US sales vs. global sales, business sales vs. consumer sales, seasonality, sales spikes for 1 week periods etc. Sadly, they seem to quote the press who mix up what they're reporting on too. I was basing it on wallstreet journal reports of market share in '87 and '88 as well as some others I need to dig through again. (all the ones I stuck with seemed to point to share in the context of hardware sales, not software -and certainly not revenue- though there's others that seem to use different metrics) The most obvious example would be Marty's own quotes from the Wallstreet Journal: http://www.atariage.com/forums/topic/143762-atari-panther/page__st__50__p__1878590#entry1878590 You're certainly right that more data is needed for a real conclusion though. (in any case it seems Sega was spinning their wheels a LOT more than Nintendo -like the 9 million 1986 marketing budget vs Nintendo's 6 million one) http://www.atariage.com/forums/topic/177019-an-interesting-little-article-from-back-in-the-day/ The most obvious loophole would be to use totally different publishers (if not developers), with bigger companies you had subsidiary labels (Ultra and LJN -though mainly used to get around the 2 game per year limit rather than multiplatform publishing) or different 3rd party publishers for others. (or first party publishing -one very nice loophole would have been for Atari -or Sega, or NEC later on- to publish the games with the 1st party console maker's label but otherwise treat the game as a 3rd party product -not commissioned, but licensed with most of the profits going to said 3rd party developers with no up front cost to Atari -etc) Activision didn't publish Double Dragon on the NES, so it wasn't in conflict. (Sega published their own port of Double Dragon -albeit a bit rushed, though still better in some ways than the NES game -mainly in terms of arcade accuracy though not necessarily overall playability) I still wonder what may have happened if Atari Corp had tapped some of the top Euro computer/console games of the time. (especially if getting those skilled programmers interested in developing for the 7800 hardware specifically -look what was milked out of the Speccy, even on the 48k beeper only models) Both systems used on cart RAM expansion among other things, the question is on the cost effectiveness of that. (and the architectural differences of MARIA needing RAM vs the NES's PPU -scanline display list architecture with other nuances vs hardware sprite+tilemap character based display with the RAM used for the tilemap data and I think sprite table as well, though maybe that's all on-chip -I know it has hardware sprite multiplexing -more RAM might have been instrumental in software hacks like raster interrupts and such -which are supported- but that's possible as it is and can be made more flexible via external expansion) I'm not sure if PaperBoy uses any RAM expansion or mapper logic though. (beyong bank switching -ie not a more advanced mapper, the MMC5 is a more extreme case that changes the 16x16 pixel color attribute cells to 8x8 -same as character cells- for much greater flexibility with the palettes -without that, raster interrupts could allow color reloading rather like A8 DLIs, but if you wanted to approximate that for a per cell basis -ie every 8 scanlines- that would still limit colors to 16x8 boxes -though unlike the mapper, you could have totally new color entries for all palettes as well as changing the additional RGB palette control for even more flexibility beyond the default bank of ~56 colors used: there's 1-1-1 RGB control over the 56 color palette for a total of ~448 possible colors to choose from as such -with the related limitations) And the 7800 has 4k shared vs 2k+2k CPU+video on the NES. (and seperate ROM -and RAM expansion- buses on the cart slot for both as well) SMB was an arcade port. Am I the only person that actually played it in the arcade? No, vs Mario bros was a conversion of the Famicom game for the arcades (a remake of sorts on different hardware and modified level set), there's also the later Playchoice game, but that's different. (just an NES with RGB PPU in an arcade cab). Vs mario bros may have been released in the US about the same time (or slightly earlier than) the release of SMB in spring of '86, but the fall '85 JP release of the Famicom game was first. There's quite a few others that have been done similarly: one REALLY obvious one would be Choplifter: there's the original Apple II game that got ported to various platforms (and a rather nicely enhanced version on the 7800), but then the Sega Licensed arcade game that went to the SMS and NES/FC later on. (Pitfall II also got an arcade game licensed under Sega -and a version of that on the SG-1000, ironically with much poorer vertical scrolling than the CV game) Atari Games had the rights to Namco's RBI Baseball, both in the arcades and on consoles via Tengen. Atari Games/Tengen had Cyberball which was huge in the arcades and did well on the other consoles. If you're going to speculate on Atari Games staying with (or being on strong licensing terms with) Atari Corp/Atari Inc (if not sold), then you can throw out all the Namco stuff. Nintendo had Namco as one of their tightest 3rd party partners on the Japanese market for better or worse. (albeit the fact that Namco continued to license games to Tengen -like Rolling Thunder- even with them going unlicensed and being sued by Nintendo, makes that a little more blurred; still ironic that Atari Games put a ton more R&D into properly reverse engineering the lockout chip vs many other unlicensed developers who opted for a cheaper and legally foolproof method of voltage spikes to freeze the lockout chip -aside from a handful of late model NESs that were glitch resistant, but that was gotten around as well with some games and the unlocked NES2 came after that as well) No, that's where Sega failed in the US. (in Europe the arcade angle had more weight -though they had other advantages like better sports games -especially soccer- better distribution/marketing, etc) Sega in the US got popular when they STOPPED pushing the Arcade at Home angle specifically. (Tonka balanced that a lot better than Sega had, but still had fairly mediocre marketing compared to Nintendo) When SoA launched the Genesis in August of 1989, they still started off with "We Bring the Arcade Experience Home", but when Michael Katz stepped in fall of '89 (cutting his vacation short), he shifted things considerably and followed up with the highly competitive "Genesis Does" marketing campaign. (the first truly successful US console marketing campaign by Sega) They didn't ignore the arcade stuff, but they didn't limit themselves to thinking on those terms either: he pushed direct competition on every level of the system, pushed for celebrity tie-ins to combat Nintendo's exclusive game licenses, pushed for more western development and a build up of SoA in general (though Kalinske pushed for all that much more), tapped the US computer development market (including EA), pushed for sports games, etc. (he'd wanted a football game out for '89, but they ended up with a horrible developer who finally defaulted in 1990 and Sega went with EA, though they actually got Madden '91 -in time for chirstmas of '90- out before Joe Montanna was released in early '91; even more importantly, Katz deftly handled EA's earlier threat to go unlicensed with a reverse engineered toolset but Katz convinced them to sign onto a favorable licensing agreement instead as well as contracting Joe Montana Football -EA's support became critical to the success of the Genesis in the US, EA sports along with Sega Sports made the Genesis the definitive sports gaming platform on the market for the entire generation -the best the SNES had was EA, and most/all SNES versions were weaker than the Genesis counterparts, like along more favorable release dates -and often price points- on the Genesis) Many of the NES's top selling hot games were arcade ports or remakes, but that the fact of them being from the arcade had less to do with it than them being good games that were well marketed. (how many people -especially kids- played NES games without ever seeing -or even knowing about- the Arcade originals? for that matter, how many Atari gamers -especially younger kids- were in a similar position at the time or in the early 80s? -probably not as extreme as the NES, but still considerable; plus you've got all the Atari/Activision/etc originals on the earlier consoles as well -some being arcade clones or enhancements thereof, but others fully original, and exceptional cases of massively successful original games like Pitfall) Atari Games' arcade games were rather insignificant in the grand scheme of things (nice additions, but way down the list). The long list of other problems from the badly managed split are worse still, and among those would be the loss of Atari Inc's consumer console game programmers. (ie the resources not only for ports, but also for original games, though they DID still have the computer game/application programmers -but limited funding on top of all that -interesting to note that at one point, a group of ACorp programmers went out of their way to port Atari arcade/console games over to the ST in a grass roots effort to push for an ST based game system -Curt mentioned that ages ago, see below) http://www.atariage.com/forums/topic/47680-the-1984-cancellation/page__p__586738#entry586738 Atari Games could have split away, but maintained a strong working relationship with Atari Corp. Yep, which is why that wasn't part of my definitive summary (bolded in my first post #130 in this thread) My comments on the other things were just musings on those 2ndary issues, the biggest ones were focused there. (hence it being one of the first things I addressed and being bolded) Yes, as I mentioned above, and the loss of that was more due to how the split was managed by Warner than the fast Tramiel bought it. mappers and RAM expansion are not necessary inclusive: mappers can be for ROM alone (bank switching among many other things -many also including sound expasion for famicom games) and RAM can be added without mappers. (in fact, I'm not sure any NES game included as much RAM as Epyx carts did with 32k SRAM chips -albeit only 16k addressed, but an expense either way) Famicom first in fall of '85, but I haven't seen definitive info on whether Vs Mario was released before, after, or simultaneously with the release of SMB in spring of 1986 for the NES. (undoubtedly, the arcade game came to some regions before the console game did -or even before the NES for that matter -which didn't launch nationwide until September of '86) It was a remake based on the Famocom game though, so SMB was NOT an arcade port in any sense. (just like Choplifer on the A8/Apple/C64/7800/etc is not an arcade port of Sega's version -though the SMS and NES games are)

Please show me a DRAM controller schematic done in standard pin through hole TTL logic chips that takes less PCB board space than 2 2K RAM chips. I wasn't saying that. I was saying that they'd have saved a LOT more in the long run once embedded DRAM interface logic (an ASIC) was used. Initial space would be more for sure, though cost would be a mixed bag. (component cost of DRAM+TTL chips vs SRAM chips in addition to number of traces and board space in either case -assuming MARIA included all necessary logic internally) And that's assuming that they couldn't push for DRAM interface logic inside MARIA from the start. (and the main reason to avoid that would be the time constraints -as with some other things like no built-in MARIA sound logic -or enabling RIOT interrupts in 7800 mode, toggling between MARIA and TIA analog video rather than shorting them together -more or less, etc) You are assuming that it would be possible to get tracks to said POKEY located on the PCB without altering the dimensions of the PCB. I'm assuming they could have pushed for POKEY from the beginning of the PCB design (or ar least the beginning of the 2 sided PCB design). They'd arrange the board to cater to POKEY as such, or plan to work around a riser board instead. (I believe they were already pushing a riser board for the 1984 production models -seated atop of the RF modulator casing) And again, make the RF modulator external from the start as well. (also gives some nice potential for a 5V powered auto switchbox) Likewise, an 8k SRAM chip should have been considered from the beginning with MARIA's address decoding and the production PCB design catering to that. (unless 8k SRAMs were really expensive back in '83 and GCC didn't want to bet on prices dropping) Or, if they recognized early on that POKEY wouldn't fit, and considered a low-cost off the shelf option. (namely the SN76489) Another option could have been having a dedicated expansion port mapped into the main memory address space that could displace the main RAM and add some other expansion lines as well. (a much more useful port than the video overlay one, but in this context mainly intended for adding RAM and POKEY in a compact and convenient module later on -when it was also cheap enough to totally merge those features with the base unit as well, be it 8k+POKEY or 16k+POKEY, or even 32k+POKEY -which would have been practical by '87 at the very least) Hell, you could have that same module tie-in with a computer add-on using POKEY's I/O for key scanning and SIO. Granted, that could all also be done via the cart slot (in an XM like module). That method could be used for PWM (or PCM) effects in-game reasonably? (any idea why no games pushed that?) Is MARIA just a lot more bus hungry than ANTIC+GTIA or does the A8 have a more organized set up for bus sharing? (interrupts are obviously favored heavily in either case) Does MARIA grab any graphics data from ROM or is it working all in RAM? (and the CPU is doing all the ROM fetches to RAM -in that case, putting MARIA on a separate bus like the MS9918 would have no disadvantages other than design complexity of the dual buses -unlike other cases where you'd have the advantage of direct access to graphics data in ROM, though there's also the RAM expansion issue) There is already an IRQ line on the cart slot. Ah, thinks, pin 31, missed that. In your opinion . No, definitely, in terms of games in general. Just 1 game released early on with POKEY would make omitting it onboard moot and wasteful (as long as said game or games were popular), RAM would definitely pay off due to the savings in board space. However, that's in contrast to "chipped" carts, not the other option: an add-on. If they'd pushed an XM like device before releasing any POKEY/RAM chipped carts, they could have made that the definitive upgrade route and concurrently standardized it on new 7800+ models. (be it displacing 4k with 8k or 32k would depend on the timing) Hell, in that case, they could have dropped to just 2 kB initially and planned for such expansion from the start (and not in the on-cart fashion). Dropping to a single 2k chip should have allowed a SN76489 at the very least, if not a full POKEY -though moving RF to external and 2k RAM should have freed up the space needed for POKEY. (again, talking from a fairly early stage in PCB design, not a last minute hack) You're right, from the start, smart programming could work around things, but not the sound limitations to the extent of being able to approximate stuff already being pushed in '82/83/84. (namely games using more complex music -SFX were generally fine with TIA; Gyruss is a prime example, though the VCS port got a pretty valiant attempt at pushing the music, it's full of sour notes due to TIA's pitch resolution -maybe they could have stylized it more and stayed in the lower frequency range as much as possible while possibly dropping some notes in general -I think a few other tracks got simplified for TIA to avoid sour notes as such, but at best it was going to sound a lot weaker short of higher res via CPU modulation) Ballblazer does 8 way scrolling as well. (plus animation on top of that for the 3D perspective checkerboard pattern -a bit like space harrier) Please show your formula for proving that interleaved DMA is faster than burst DMA on the 7800. If burst DMA in the 7800 is using much higher bandwidth for MARIA than the CPU's bus speed, then yes, you'd have a big disadvantage there. (let alone the potential for FASTER burst DMA with MARIA with faster RAM -until you hit memory faster than MARIA could go, and then you could push for interleaving with no performance hits)

Coleco had the "Home Video Game" rights, and Atari had the "Home Computer" rights. Almost makes one wish they would have created the XE Game system instead, because Atari's Donkey Kong would have BLOWN away the CV. On the bad side:games like 8-bit Dig Dug were an embarrassment in some ways. However, I like the music and multi-colored sprites better on the 8-bits. The 5200 had nicer dirt ... lol..(Why couldn't they combine the best of both worlds on that one?) The 5200 Centipede may have never surfaced, since people could already have bought the lame 8-bit version. I personally would have missed two fire button and the keypad. I like the analog sticks too for the most part. Why only almost? A 1982 incarnation of the A8-game-system concept (ie XLGS, etc) could have been far better in every way than the historical 5200 as executed. (they 5200 COULD have had other advantages -namely lower cost due to the removal of PIA- but that didn't happen due to many, many, many areas of inefficient design in the system) Basically, it's what the A400 was inteded to be back in '79/80, a high-end game system, but in '82 it could finally be at a competitive price for a game console. (as it was, the 400 was in a similar price range to the 5200, less with the $50 rebates offered in late '82 iirc) They could have based it on the quite compact Atari 600 computer design, though that SHOULD have been launched in '82 in its own right as well. (they could have pushed a game system version in parallel though without PBI or BASIC ROM built-in -perhaps a built in game instead- and a low cost membrane keyboard -since it would be a game console first and foremost) Whether they used 2 or 4 controller ports is another issue, but given only 1 game (Asteroids) needed 4 joytports for simultaneous play, that wasn't THAT big of a loss (and with the XL's standard of 2 ports, newer games would cater to that anyway -and you'd still have 8 player paddle games). OTOH, they could have kept cost down and still included 4 ports as an option via a simple expansion port on the PCB, or as standard. (probably not too expensive, but would mean a heavier redesign of the board vs the 600 -that and the market would be a bit split with the 8-bits not having any support for 4 controllers, though technically they could have if there was a switching mechanism for 48k RAM mode with 4 ports and 62k mode with 4 ports) So you could have the high end/mid-range 1200XL (not sure what they were charging in '82), the lower end 600 (probably around $250-300), and the entry level/console XLGS launching at $200. (would have been neat if they kept the cool styling of the 600 prototype and had new cartridge casings in a similar style -and black, like the 5200 but the size of the A8 carts, granted they wouldn't look as cool/impressive as the 5200s, but they could have gotten end labels at least ) Hell, they probably should have offered the 600 in both 16 and 32k versions (like the 400), but the XLGS fixed to 16k. Oh, and they should have updated the joysticks too (and offered the new ones for A8s in general). The 5200 layout/design is pretty comfortable, they just needed slightly larger (maybe) solid plastic side buttons (7800 style might be nice, but the 2 button per side layout has advantages too -less likely to squeeze by accident), all PCB contacts (no flex circuits) with rubber or metal dome switches for reliability and consistency, and finally, a simpler (and cheaper) 4 switch digital joystick of similar size (I think a CX-40 type rubberized stick that doesn't spin might be better -same for the 7800). Analog joysticks could have been optional accessories, paddles would be normal VCS/A8 paddles, and the stocks with added keypad would also probably be optional. (or nonexistent, requiring the console's onboard keys instead) The digital sticks could derive added buttons from the 2 POT inputs. (many A8 games already use software pause via the space bar, so that's a non issue as well )

I'd gotten the impression that production had been halted (or heavily curtailed) due to the extensive stockpiles left over from overproduction (mainly due to faulty market growth/demand figures derived from the flawed distribution system) on top of the reduced demand in the heat of the crash. (Morgan was planning on launching the Jr in late 1984 though, so that would imply they were at least continuing production to some degree up to that point) Katz mentioned that they were selling beyond capacity by late 1985; that is: selling faster than they could produce more consoles (so they were in production by about mid '85 at least) as they hadn't boosted their output yet. (something addressed by early 1986 -Katz commented that they probably could have sold several hundred thousand more 2600s in 1985 with optimal capacity) Wan't the lauwsuit largely generated by Warner's horrible management of the split and utter confusion over it: no notification of staff (upper management or otherwise), no normal transition, etc, etc. Tramiel didn't fire anyone during the split: Warner laid off 100% of ALL Atari Inc staff (save the coin guys who got rolled into Atari Games). What Tramiel had to do was start interviewing and hiring people to TTL (which then became Atari Corp by name). Every single Atari consumer employee became unemployed with the signing of the contract. (that's one of the many issues of Warner's horrible management and the compiled issue of it happening over a holiday weekend -4th f July- etc, etc) Jack wasn't given a remotely adequate inventory of what he was taking on with Atari consumer, let alone the necessary preparations for any realistic transition from the ongoing plans and operations. Nope, the Amiga deal was 100% dead a couple days before Warner made the final offer to Jack and entered negotiations for the split, all that would have changed is how well/quickly any litigation over breech of contract would go. (to stop that from happening, Atari Inc needed to stop whatever employee from cashing the return/cancellation check from Amiga) Yes, the 6501 suit was something that made Atari's decision for immediate 2nd sourcing all the more important. The price war wasn't with the C64, I got corrected on that assumption too. The price war (the one peaking in '83 that led to TI dropping out of the market) was mainly with the VIC-20, not the C64. (though the C64 closing up the upper end of the market was significant too -sandwiching the TI99/4a between the 2 -except TI never pushed for a proper C64 class version, or higher end even with a ton more main RAM -a shame really since DRAM by '82/83 should have been getting fast enough to support the TMS9900's requirements -though they should have been able to add wait states much earlier for use with slower DRAM, though they must have already had a slower memory interface for ROM, so it's even stranger that they didn't offer more RAM even back in '81 . . . for that matter the VIC was all SRAM including the external 8k/16k/etc expansion carts -more RAM and a desk top form factor might have pushed it into the "serious" computer role especially with the rather powerful CPU, which had been one of the major issues jacking up the price as it was vs related Z80 based derivatives of the same chipset, though that's also an argument for TI releasing a lower-cost Z80 based system as well -OTOH they also made a massive mistake with the 3rd party blocking strategy and emphasis on 1st party software, it was a computer, not a game console, so I have no idea why they thought that was a good idea in the competitive mass market) Again, I don't think Warner was necessarily in a position to outbid and out-haggle Tramiel on the MOS merger: CBM should have had some fairly solid funds in '76, but more so, they still had the added leverage to pull MOS towards them.

No, the 7800 has hardware scrolling, not special tricks. (the A8 and 5200 have V/H scroll for the playfield as well) Several games used it too, though I'm not sure of any doing diagonal scrolling on the 7800 BITD. (more of a limit on the games put on the system than anything else) Double Dragon, Scrapyard Dog, Ninja Golf, and a few others had horizontal smooth scrolling as good as the A8, NES, C64, etc. Atari 3200, in development in 1980 and 1982, canned in favor of the "fast" option with the 5200 in '82. (and the 5200 ad advantages, but all were wasted as mentioned in my first post above -cost effectiveness and provisions for a cheap/simple/efficient VCS adapter were among the major problems) Yep, 3200, basically a 7800 with MARIA replaced with STIA+FRANTIC (STIA includes TIA compatibility, FRANTIC is ANTIC, STIA was to either add GTIA-like support or expand that even further), though they could have ANTIC and GTIA on the board like MARIA instead for a "fast track" option to cut R&D time. (CGIA would make it even more like the 7800 -and the cheapest sound expansion could haeve been enabling RIOT interrupts for CPU modulation of TIA and GTIA's 1-bit toggle channel) No, the Japanese Arcades is where Nintendo had locked Atari Corp out already in mid 1985 (when Katz had tried to get licenses), Nintendo had real clout in Japan by that point, rather like Atari in 1979 in the US. (the NES didn't really become significant in the US until after the Septembr 1986 launch and the successful holiday sales season -the 1985 test market failed and the early '86 stuff was promising but nothing to really put Nintendo on the map) Yes, that and releasing the 600 in 1982 alongside the 1200XL would have made a lto of sense. (a lower end game console and a low-cost computer in the price range of higher end consoles and with graphics and sound capabilities also on part or ahead of those consoles ) The 7800 sold 3.7 million to Nintendo's 38m US / 61m World. Atari may have been second, but it was not much of a competitor. You're looking at it wrong. You need to compare 7800 and 2600 sales in 1985-1988 with NES sales. (they were still a distant 2nd from '87 onward, but MUCH closer than the NES with it's massive 1989, 1990, and 1991 sales -and respectable 1992/93/94 sales) The SMS almost certainly outsold the 7800 by a small margin (or came close too it -the 2 million figure is wrong, it's closer to 4 million in the US), but the 2600 sales pushed Atari to more than double Sega's market share up through '89. (up through '88, Atari averaged above 20% of the US market share I believe, though in '88 specifically they were down to around 15% or a bit less) In 1989, the 7800's sales dropped dramatically (below 700k) and in 1990 they were below 100k units. 1987 and 1988 saw about 75% of those 3.77 million units in the US market. (as such, 1989 is when they needed to push for a successor -like a Lynx or STe derivative, etc) Then again, I believe Ms. Pac-Man was the top-selling title on the 7800 and was apparently the biggest selling third-party cartridge on the Genesis. Or maybe people just REALLY liked Ms. Pac-Man... I think people REALLY liked Ms. Pac-Man. The Pac-Man series are still big sellers today, 3 decades after they were first relevant. As for it being the top seller on the 7800, I think that says more about the lack of software than anything else. Ms. Pac Man probably should have been a pack-in game for the 7800. It's older, but also timeless and addictive. (and one of the best ports of pac man -especially at the time) If they added a 2 player vs/co-op mode to it, that would have been perfect! As for arcade games in general, a TON of the NES's library of hot games were arcade ports or remixes, but they were newer arcade games and in many cases, games the Nintendo users never even bothered playing in the arcade. (so they only knew the NES games) Indeed. Or a POKEY, for that matter. What's really sad is that 28-pin SRAMs are only available in 8K or 32K (someone correct me if I'm wrong about this), but Atari's cartridges could only support up to 16K. So if a game needed more than 8K, they had to put a 32K chip in there and then use only half of it. Very wasteful. They probably could have fit more RAM inside the console if they had used larger RAM chips than the pair of 2K SRAMs that they went with. Perhaps two chips were used instead of one because they needed to be scanned independently by the hardware, but even so, larger ones would have been nice. The 7800 should allow up to 48k of flat mapped ROM or RAM via the cart slot, I think Epyx only used 16k to avoid bank switching (or is the game larger than 32k? -maybe they wanted 32k rather than 16k banks) or possibly to cut on the logic needed for address decoding. However, the fact they were willing to cram 32k SRAM onto carts in 1987 and that Ballblazer (the first -of 2 POKEY games) was released in '87 as well points to the missed opportunity of a 32k+POKEY XM type module instead. (probably have one of those games pack-in with it too, maybe embed it onto the motherboard of a revised "7800 plus" model too, hell, maybe intentionally map only 28 of the 32 kB -leave 20 kB for ROM- and use a single 32k SRAM chip to displace the 2 2k chips on the motherboard mapped in the same fashion as the add-on -with a software or hardware switching mechanism to enable or disable the added 28k vs full 48k flat ROM space -that's another advantage of keeping it SRAM as DRAM might not be able to displace the 4k internal memory) POKEY's I/O could have been used to provide an interface foe a keyboard as part of a full computer expansion as well. (a better idea than the XEGS for '87 if they wanted a good entry level computer -albeit 32k vs 62k of the XEGS, or nominally less with the OS loaded-) I was going to start a thread on this. That would have been proactive of Atari Corp: they couldn't redesign the initial production run (5000+ units already built), but they could correct the shortcomings after the fact to make it more competitive late gen. (just a shame MARIA doesn't allow interleaved DMA -even in fast SRAM- and that there's no IRQ on the cart slot or RIOT interrupts -so no interrupt driven audio) Even without any interleave for ROM (ie for faster ROMs later on), interleaving in RAM would have been a boost even in 4k, let alone 32-36k. (where you could afford to load much into RAM -or even compress data in ROM and load into RAM before a level -or mid level in some cases) Oh, and 8k SRAMs should also be in 24-pin DIPs, 16kx4-bit should be well within 24 pins. (still a lot of space to use on cart though) A lot of people seem to be missing this. Atari coin was only useful for licensing/borrowing for ports on the console market, the loss of the console programmers from Atari Inc to Atari Corp was a MUCH bigger issue on top of the funding problems. That almost certainly wouldn't have happened (among other things as mentioned in my first post above) if Warner had managed the split with a smooth transition to allow Morgan's management to hold things together as they were with a formal, well organized shift to Atari Corp. (maybe even keeping the entire Atari Inc staff initially and let Morgan complete reorganization -perhaps with some tweaks at Tramiel's request) The split and resulting mess also ruined any potential for a good working relationship with Atari Games. (lawsuits over who owned the rights to what, a fair amount of animosity, etc)

Actually, the NES has advantages over SID and POKEY, and (for games specifically at least) I'd say is more well rounded overall, though POKEY is probably ahead overall for games too due to the sheer number of channels along with the various CPU+interrupt assisted tricks for more flexible sound. (except maybe with the SID's 4-bit hack/bug channel used) The NES has 2 pulse wave channels (4 duty cycles, so not as wide a range as the SID or CPU modulated POKEY, but pretty nice in any case), 1 fixed volume triangle wave channel, 1 dedicated noise channel, and 1 1-bit delta modulation channel decompressing 1-bit samples in hardware with DMA (8 or 16k banks with interrupt support for bank switching) with 16 sample rates from ~4 kHz to 33.5 kHz decompressing 1-bit constant delta samples to 6-bit PCM. (quality of the output depended on the encoding to a good degree and the sound being compressed, but it was a nice feature for the time -dynamic range compression would be one preprocessing option to improve slope error and SNR-) The DMC port could also be written to as a 7-bit DAC using the CPU for software PCM playback. (ie any resolution up to 7 bits, usually using the hblank interrupt feature -for up to 15.7 kHz or any division thereof) Also note that 33.5 kHz 1-bit DM is only 33.5 kbps or ~4.19 kB/s. (equivalent to a 4.19 kHz 8-bit sample or 8.38 kHz 4-bit sample) A lot of developers used DMS for music samples and a few for voices (some use CPU driven PCM as well), usually used for slap bass or percussion among a few other things. (Konami used it in almost every game the made, Sunsoft used it a lot too, Nintendo used it in a few games including the steel drum in SMB3, etc, etc) One of the most amazing examples on the console: http://www.youtube.com/watch?v=PBM-m82Za_0 And if you want the NES in classic 80s European chiptune style (sort of a cross between SID and AY/YM PSG tunes), you've got codemasters: http://www.youtube.com/watch?v=xhk5pA0tLgY (gah! arpeggios ) (no use of DMC AFIK -maybe use of the sawtooth hack of the DMC channel) And the closest you'll get to classic US 8-bit gaming music, straight from David Crane (it's relatively simpler, has sort of an early/mid 80s "atari" feel to it, not exceptional but interesting for contrast, uses the DMC channel a lot too) 5200's is more comfortable to use, but it's more finicky for some things. (aging issues aside -for soem games the analogistic on a good 5200 controller isn't too bad, a shame they didn't keep the form factor moving to the 7800, but dropped to 7800 like buttons and a simpler digital joystick mechanism -maybe a rubberized stick with a tougher boot more like a scaled down CX-40 stick) After writing three scrolling homebrew games (2 of them scroll in all 4 directions) I can't agree with this statement. The 7800 is easily capable of doing scrolling in all directions. As are the 1979 Atari 400 and 800, and the 5200. (the VCS managed some very nice vertical smooth scrolling though, horizontal was rather clunky though -Jungle Hunt managed it OK though given the limitations -vertical software scrolling seems a lot easier to manage in most systems -generally tied to the scanline based nature -a shame the MSX2 added V scroll only rather than H scroll ) Yeah, if he were able to pull off miracles like some kind of business guru, he might have been able to take a near bankrupt company and battle Commodore and others in the computer market. Heck, if he were really good he might even unload all that 2600 stuff they had lying around and make tens of millions of dollars in sales off it. And if he were truely amazing, he might even take a near bankrupt company and make it into a Fortune 500 company and get it to last from 1986 all the way to the mid 90s. Too bad he couldn't manage any of that. Oh wait. And if he was REALY good, he also launch a highly successful 16-bit home computer that dominated Europe in the mid/late 80s (and only really declined after he left Atari). Or he'd not only "unload" old hardware, but sell so well that they start up production again and continue it for over 1/2 a decade. Oh, and he'd do that in spite of Warner botching the sale and making a chaotic mess out of the division he'd just purchased and also ruined most of the hard work that the preceding president had spent the last 8 months on. Yeah, no big deal at all. I think it was GCC's decision, I don't think Atari Inc even had any direct say in the design of the 7800: it was with Warner who was to "force feed" it to Atari at their own discretion. (yeah, not characteristic of dual management or a bureaucratic mess at all, right?) Doesn't it have a dost cover like the VCS? (embedded in the cart slot itself) Expansion port was pointless, B&W lables are legitimate though. (at least the boxes -the important/flashy marketing side of things- were full color though, and they switched to color labels later on -may have been done due to increased funds or other issue, not necessarily related to Jack leaving and may not have even corresponded to that -ie may have been planned long before it reached market) I was asking for some specific titles of arcade games that were ported to the NES in its early years so that they were unable to be released on the 7800. You could get around the exclusivity requirement for arcade games by licensing it to a separate developer. Yep, and atari corp could have offered such a loophole with special royalty agreements for developers to publish under the Atari label. (and by "special" I mean near 100% profits go to the developers and developers have control over the ROM size, features, etc just as with a true 3rd party game -but Atari makes no investment for commissioning the game, maybe even have the 3rd party manufacture and distribute the game, but with Atari's lable and the related license) If were a parent in the late 80's, and my choice was between an XEGS 8 bit computer/console, NES, SMS, and the 7800, my kids would get the XEGS so they could also use it as a computer. They could learn BASIC programming, use it to word process, and play that great catalog of 8 bit Atari video games. Yeah, because that obviously worked Yeah, it "worked" just like that in Europe. http://custurd.b3ta.com/heyhey16k/ Yes, but computers give so much more excuses to annoy their parents with. But as above, XEGS or XLGS should have been in 1984/85. (or instead of the 5200 based on the 600 proto -which should have been out in '82 as well) Or fast enough RAM to allow interleaved DMA (with halts for ROM sharing -maybe options for slow and fast/interleave ROM select as well), that's one good thing about SRAM. (though with Amiga class DRAM, they should have been able to interleave with DAM even -and maybe even afford 32k)

Not necessarily. The 2600 was the cheap option. If that cheaper option was not available, they would either have to do without a video game console or choose between the more expensive options at which point they could just as likely have chosen to buy the NES instead. Yes, and the PS1 and PS2 case is the best example, even as the PS2 was getting cheaper (and was fully backwards compatible), the PS1 was still on the market in the budget option. Those were GCC's decisions to make though, possibly with some influences by Warner. They'd have been made on the grounds of cost, time, and backwards compatibility constraints. There's a TON more they could have done with the system if they were to open up any of those restrictions, let alone all 3. (the design was frozen for production in early 1984, so it was limited by that) The decision to go with SRAM over DRAM was probably mainly a time constraint (DRAM is more complex to interface, let alone integrating that interface logic for a more efficient design -albeit that could be left for later -I think using discrete logic for DRAM interfacing initially with much more consolidation later on might have been the right move: 2 16kx4-bit DRAM chips -that's 2 8kB 16 pin DIP chips costing much less and taking less space than the 2 2kx8-bit SRAM chips used though with added interface logic needed -more than worth it in the long run though). Short of DRAM, they at very least should have aimed at 8 kB of SRAM from the start, a single 28-pin 8kx8-bit SRAM would take about 1/2 the board space and be cheaper in terms of component costs in the long run. (that and putting the RF modulator externally -using an A8 type monitor port internally- could have freed up the board space to include a full POKEY without a riser board) GCC had initially wanted to add sound to MARIA, but that was dropped due to time constraints. I'm not sure if they ever requested the schematics for POKEY 9or engineering assistance) from Warner/Atari, but that might have accelerated things. (ie re-use the sound block from POKEY using a smaller die in a smaller package -like 24 pins or less) Though Atari Inc had a huge supply of plain POKEYs at the time, so it may have been more cost effective to aim at using the full 40 pin DIP as such. Enabling RIOT interrupts in 7800 mode would also have been a nice feature. (in some situations useful for driving CPU modulation of the TIA channels -for pulse waves, PCM samples, or just higher resolution square waves, though you'd also need to avoid saturating the bus with MARIA in active display -or you'd miss interrupts or delay them and throw the sound off) Other options like dual buses with dedicated video and/or CPU RAM would probably have been up to development time (more complex design), and interleaved DMA would have needed RAM and ROM at 2x the speed originally used. (RAM wasn't so much an issue since they were using SRAM -relatively fast- and maybe they could have had both interleaved and holey DMA modes for ROM based on the cart telling MARIA if it was fast or slot ROM, but that's more complexity still) Otherwise I addressed the key problem of the 7800 (and Atari Inc to Corp in general) above alogn with a lot of other stuff. Also, short of POKEY, they could have opted for a small/cheap, off the shelf SN76489 PSG (the Colecovision,SMS,TI99/4, etc sound chip, a small 16 pin DIP). If GCC had been developing a MARIA console on their own, they probably would have opted for an AY-3-8910 (use for the controller I/O as well as sound, not as flexible as POKEY overall, but better than the SN76489 and pretty decent -the best off the shelf chip on the market), but that's another issue entirely. Yes, if they were planning on putting POKEY on even 1 popular game (early on), they should have put it inside the console even if it meant using a riser board. (but again, focusing on a single 8k SRAM and external RF modulator would be very simple ways to free up board space) Adding RIOT interrupts would be another option for better hacking of TIA. (as long as you had enough CPU time for even intervals in active display you could manage some nice software modulations -still only 2 channels max though, aside from software mixed/interleaved 4-bit PCM ) Adding RIOT interrupts would be a complexity issue (a more comprehensive split from the 7800 and VCS modes -they also should have fully separated MARIA and TIA analog video too though and allowed much better color/video quality -a switch in hardware between MARIA and TIA's analog composite video output) On the sound and RAM issues though, they COULD have opted for an add-on (like a mid 80s flavor of the XM -like a POKEY with 32k of SRAM or DRAM -in 1987 you say the first games to use POKEY or RAM those Epyx games had full 32kx8-bit SRAM chips inside), and then add those features standard to later 7800 models. (ie 7800 Plus or something) Hell, they could have introduced that module along with a keyboard add-on instead of the XEGS. They should have had IRQ input on the cart slot though, or a dedicated expansion port for RAM+sound+etc with more than the cart slot offered. (not the video expansion port...) It would have been cheaper and cleaner in the long run to push more RAM and POKEY from the start though, and possibly put POKEY's keyboard+SIO lines on an expasnion port (for the computer add-on ). Given the low-cost emphasis of Tramiel, an add-on module would make a lot more sense. (the added RAM could even allow compression in ROM to load into RAM -there were some decent lossless compression schemes at the time, and you wouldn't be doing it on the fly, so CPU resource was a non issue -just some relatively short load times) Hmm, support for interleaved DMA in fast RAM (but waits in shared slow ROM) would have really paid off with a nce chunk of RAM to work with. (ie load almost everything into RAM as needed for full parallel 1.79 MHz CPU+MARIA bus access ) It seems like GCC engineered the 7800 without really any direct contact with Atari Corp (all through Warner), but I still don't see why they wouldn't have pushed for POKEY from the start. Everything else you mention is mainly tied to Warner's poor transition: even some outright mistakes made by Tramiel may have been avoided with a proper transition supported by Warner. (ie Morgan's plans continue in a transitional period and maybe even have him retained under Atari Corp -his plans for downsizing shouldn't have conflicted with Tramiels ones -the problem was that Tramiel was forced to do so under extremely adverse chaotic circumstances due to Warner) The takeover was totally averted before Warner took any actions to sell off Atari Inc (not considering a split until later still, after being rejected by all potential buyers). The attempted takeover did influence the sale of Atari though, but in the context of Warner hiring a firm to assess their finances who then recommended that they part with Atari Inc. A major factor was Warner's shareholder as well, Warner needed to get the debt off the books. I do wonder if, technically speaking, Warner couldn't have *simply* spun-off Atari Inc as a separate company to offload the debt (maybe under Morgan's ownership 0if he was willing to make that sort of commitment) and allow reorganization to complete normally as such. (in that case, Morgan would also be more free of Warner's management conflicts, and if he was confident enough in his own capabilities, that may have been a favorable prospect -I wonder if that was even a consideration at all) In 1987 Atari Corp published 2 games (Epyx's Summer and Winter games) with wach had a 32kx8-bit (32 kB) SRAM chip onboard, though only 16k was actually used (I think they could have mapped more -48k directly addressed to the cart slot- but that would have meant bank switching the ROM on top of using SRAM -you'd only have 16k left to address ROM). But that's 1987, not 1984 when the system was intended to be released. (at that time, 8 kB -a single 8k chip- probably had more advantages than disadvantages -about 1/2 the board space, more flexible use of MARIA, cheaper in the long run with fewer board revisions, and more room on the board to potentially add POKEY -there's a line where feature creep exceeds practicality, but I don't think 8k SRAM and POKEY had reached that point, especially since POKEY could have made for an easier keyboard/computer add-on -key scanning, SIO, etc) Unless you did a multiplexing sheme with a multitap module (a la genesis), or added I/O logic to the cartridge (a la Codemaster's J-Cart on the Genesis or added I/O in an add-on). Not even close, and you're way off with the Sega CD comment too. (32x and Saturn is pretty much spot on though, but Sega CD was released 3 years after the MD and 3 years before the Saturn/32x -in Japan- or 3 years after the Genesis and 2 years before the 32x in the US) The 5200 was released a full 5 years after the VCS (not super long, but about average for later game consoles), the 7800 was intended to launch 2 full years after the 5200, but the 5200 was formally discontinued in favor of the 7800, no overlap as such. (they were also planning on a 7800 module to play 2600 and 7800 games on the 5200 to bridge the gap and promote better PR) The problem was that the 7800 got held up due to conflicts over Warner's sloppy split of Atari Inc consumer and didn't launch until mid 1986 (limited releases early that year -test markets back in mid '84 if you count that), so that was almost a full generation (4 years) after the release of the 5200 and almost 9 years after the VCS's 1977 release. (and 6 years after the VCS really hit massively in 1980 with Space Invaders) So nothing like Sega at all. (though the 5200, like the 32x, often gets scapegoated as "the problem" when both were symptoms of much bigger problems inside the companies at the time) The XEGS was a genuine mistake though and one that could/should have been avoided. (Now, in 1984/1985, a special low-cost gaming version of the 600XL would have been great to push, and if it was popular enough, maybe it would have made them think twice about even launching the 7800 -have the VCS in the low end, the mid range gaming computer, and the mid range 8-bits along with the ST higher up still; or in either case, position the 600GS -or whatever- in such a way to allow the 7800 to come in later if they chose to -a computer could push huge loopholes into Nintendo's licensing policies -and Nintendo couldn't afford to block out computer publishing as 3rd parties would be extremely reluctant to even consider that, let alone with Atari's own computer game system getting popular) Of course, you could argue that a 600 (original 1982 600) based game console back in '82 would have been much more fool proof than the 5200. (and could have Donkey Kong -and a better version than the CV at that with all the levels ) Most 5200 games are character based (more RAM is also needed for more flexible character graphics effects -more complex DLI tricks, etc, etc), the 16k of DRAM was also useful for bitmap modes, but not limited to that by far. It was also cheaper than the 4k SRAM in the 7800 (but needed more interface logic -still a huge savings -and bonus- in the long run with embedded interface logic). More RAM also means more options like decompressing graphics into RAM, framebuffer/bitmap modes, and other things. (using fast enough RAM, they potentially could have used interleaved DMA rather than holey DMA -at least in RAM, maybe optionally in ROM too) 8k SRAM is also attractive, but I'm repeating myself. A repackaging of the 65XE in a gaming bundle should have been cheaper, and could have been positioned better as a "games and educational computer" and marketed as an entertainment/educational/entry level computer not overlapping quite as heavily with the 7800 as the XEGS did. (the one change should have been moving the cart slot to the top of the machine -composite outputs is a moot issue, the monitor port is no more difficult to use than the Master System AV port -and there's S-video ) As above, a gaming version of the XL/XE was much more needed in '84/85 (due to the delay of the 7800), though also a foolproof alternative to the 5200. (like a cheap version of the 600 launched in '82 alongside the full 600, etc, etc) An even better reason to do so in '84/85 and have more of an impact before it was too late. (though, again, in '84/85, probably a 16k system, not 64k -32k would be pushing it, though switching to a single 32kx8-bit DRAM chip later on would have been great)

No, no, and no. Others have addressed this but: same thing as if Nintendo had canned the NES in 1991 (when it had 3-4 years of high profitability left and more in Japan -the Famicom was still in production in Japan almost as long as the 2600 iirc). (or SNES after the N64 launched, or Master System after the Genesis launched -in Europe or Brazil where it was really popular, or PS1, PS2, etc, etc -Sega made that mistake with the Genesis in the US, not supporting it properly in the highly profitable late gen market into the budget years -so much so that they sold off all stock in '97 and Majesco offered to start new distribution under license, so the Genesis came back -albeit weaker than Sega could have had it- through '97, the Genesis 3 in '98, and into 2000 before discontinuation) It was a budget console place in the market, separate from the 7800 (as it would have been if the 5200 had stayed on the market), keeping the 2600 on the market wasn't the problem. (especially since they were selling it for a profit, probably a considerably greater margin than the 7800 for that matter) Dropping the 2600 to "force" buyers towards the 7800 wouldn't be any good, it might help boost 7800 popularity a little, but would more likely hurt Atari more in general. (more people would go for used stuff rather than new, or wait for an NES rather than a budget system in the short run) Also note that in today's $, the 7800 was launched at an SRP of $160 and the 2600 Jr at $100. (1986 is almost perfect 2:1 for current inflation) Supporting and marketing the 7800 better was a separate issue from keeping the 2600. I will say that back in 1982, they could have positioned the 2600 and 5200 better as such and pushed the 2600 into the full budget marker sooner, but there's so many more problems from that period as well from mismarketing the computers, to management issues, the hugely problematic distribution system inflating the marketign and causing the crash, etc, etc. (1982 was the point Warner/Atari Inc needed to act to avoid disaster, by the time Morgan finally came in to start fixing things, disaster was already set -more or less- and he had to clean up the resulting mess after the fact -he wasn't perfect either, but he was a huge help that was needed ayear earlier -a shame warner ruined his efforts with the sloppy management of the split/sale in mid 1984 -hell, under a smooth transition, Tramiel may have even been a positive impact on Morgan's efforts ) And that's where the definitive problem of the 7800 comes in: Warner's sloppy split resulting in the conflict over GCC that delayed the 7800, also caused a huge mess in Atari Consumer with Morgan's promising reorganization completely ruined, any real chance of an efficient and organized downsizing process destroyed, etc, etc. Atari's 16 bit computer designed became moot (more or less), they lost the console game programmers (in addition to the Coin-op division spun off as Atari Games -with a horrible relationship with Atari Corp due to the sloppy split and conflicts), the A8 line had delays and misses in marketing, the 2600 Jr was delayed, etc, etc. Warner didn't do anyone any good with that split: a prudent, well organized transition would have benefitted them, Tramiel, and Atari personnel in general, but they shot themselves and everyone else (aside from Nintendo and Commodore) in the foot by doing things like that. Tramiel was lucky to pull things together as well as he did. By that point, Atari Corp was at a strategic disadvantage, they didn't have the in-house resources to compete like Sega and didn't have any 3rd party support (aside from 1 Activision release -just Double Dragon iirc, the rest were commissioned and published by Atari Corp). In spite of that though, the 7800 managed to sell well over 1 million units in 1988, and similar in '87, those were the peak years when roughly 75% of 7800s were sold, but it declined significantly in '89 as Nintendo pushed near their absolute peak. ('89 should have been a good time for a successor to the 7800 in that regard, be it agreeing to the MegaDrive distribution deal in '88 or going with one of their in-house console projects -Jack and Dave Rosen couldn't agree on terms over the MD and at the time, it would have been a major trade-off for Atari given their growing in-house resources and position of the ST in Europe) The real problem was that both Jack and Katz left towards the end of '88 (Katz in early '89 -several months later he cut his vacation short to join Sega of America just after the Genesis's launch and create Sega's first truly successful US marketing campaign with "Genesis Does"), so Atari Corp lost both Jack's shrewd business sense and Katz's knowledge of the entertainment/games market and marketing skills. (Sam Tramiel proved to be far less capable than his father in pretty much every respect as president and CEO of Atari Corp and led to a downward spiral taking the company from a multi-division, multi-product, multinational fortune 500 company down to a single product company -just the Jaguar- barely staying afloat in 1996 when Jack stepped back in -albeit with enough assets to liquidate profitably, though that wouldn't have been true if things had gone on much longer -Sam got lucky with the short term Jag hype and winnings from Sega's law suit in '94, but they were bleeding money heavily in '95) The computers were extremely important and Warner/Atari Inc missed out by not marketing the A8 line well in the early 80s (they could and should have kicked both Apple and Commodore to the curb in the US and done much better in Europe). The XEGS was a bit of a mess though and wasn't marketed as the entry level computer it was truly intended to be. Honestly, the XEGS would have been awesome in 1984/85 given the difficulties Tramiel had with the GCC contract, and with the right marketing they could have placed it as a low-end game-oriented computer but leaving room for the 7800 to come in if/when they smoothed things out there. (though you could argue cutting out the 7800 and focusing on A8 software might have been preferable -if the A8 game system had caught on well at least) Basically take the 600XL from 1984 and offer a gaming bundle, then probably reduce an even more cost cut version that removed the built-in BASIC (perhaps with a built-in game instead), and probably a low cost membrane keyboard. Heh, one really cool thing about that is that Nintendo's licensing agreements would have been screwed: Atari's game system would be a computer (technically), NOT a console. (don't bother with a separate keyboard -other than maybe an add-on- and keep an all-in one form factor with a top mounted cart slot like the 600XL, but shrink it down a bit -something closer to the XE styling, but in all black might have been really cool) It would lack lockout, but given Atari Corp's position in the late 80s, that would be more of a blessing than a curse by far IMO. (as it was they probably should have opened up free licensing on the 7800 as it was, just the cost of the dev tools/system, and maybe offer a route for a developer to publish through Atari Corp to get around Nintendo's problems -but still have normal profits/royalties from the games) It's hindsight, but a game computer would really break Nintendo's licensing schemes as they were. The 5200, like ET and Pac Man on the VCS, more of a symptom of Warner/Atari Inc's problems than a problem outright. (their biggest problem was the distribution system that returned false, bloated demand/sales figures and led to oversaturation and overproduction, followed by the crash) However, on the 5200 specifically: it lacked compatibility with the VCS or A8 computers, but lack of compatibility is OK if it takes advantage of that fact: indeed they stripped out PIA and used POKEY+GTIA alone for I/O, no RIOT or TIA on the board, and potential for consolidation (ANTIC and GTIA were merged into CGIA in 1983, but not pushed into production). The architecture was close enough to the A8 to allow fairly simple and unadulterated ports, and it used cheap DRAM as the computers. (which also meant added logic for interfacing -but that would pay off with later consolidation) The problem was it was just so poorly executed: the lack of internal compatibility could have been partially mitigated by an efficient expansion interface (via cart or an elegantly placed expansion port -like the CV did) that allowed as little cost or hassle as possible for a VCS module. (say enough connections to allow only TIA and RIOT to be on the module with a clock input to put the 5200's CPU into 6507 compatibility mode, etc, etc) Controller ports could have been VCS compatible, but flexible (perhaps even remapping I/O) to allow more buttons and possible accessories (maybe more I/O lines on the expansion port and optional keyboard add-on), and they could be low-cost "digital" controllers similar in form factor to the 5200 but with no keypad, better side buttons, and a digital/pseudo-digital (pull up resistor analog) joystick with PCB inside rather than flex circuitry for higher reliability. Then one of the biggest issues: the motherboard needed to be as compact as possible as well as the case, not just for convenience but also for cost of production, packaging, distribution, and shelf space for retailers. (they should have used 8k DRAM chips from the start to further cut component costs -the C64 and 1200XL were already using them too) They didn't need to bother the the storage tray either, just a nice sleek design based on the 2700's styling but no tray, more compact, and controller ports on the front. (I like the idea of keeping to a purer 2700-like wedge over the 5200's more slab and semi-curved appearance) The cart slot should have supported audio input and RAM expansion too. But seriously, the thing should have been smaller and cheaper than the 1982 600 prototype, but it was not only larger and heavier than the 1979 Atari 400 with heavy castings and multiple boards, but it had a larger footprint than the Atari 800 even and a larger motherboard than the 1200 XL. (which had to include SIO, 4x the RAM, PIA, and some added logic for the memory mapping) The 5200's specs and components used should have allowed Atari to undercut Coleco as well as have a quickly boosted library converted from the A8, but it was more expensive and less definitively supported. I still think the 3200 may have been more foolproof. (or a hybrid hack of the 3200 concept and 5200 for that matter -basically a 7800 with GTIA and ANTIC instead of MARIA, though using DRAM would have been nice) That, or you could argue they should have just pushed out the Atari 600 as a game system (they should have launched it either way in '82 though), or maybe a full 600 along with the gaming version with a derivative of the cheap 400 membrane keyboard and Pac Man built-in rather than BASIC. (plus they could get around Coleco's Donkey Kong license for game consoles -since it would be a game Computers, not a dedicated game system ) Hell, as it was, the less cost effective 16k Atari 400 was cheaper than the 5200 in 1982. (under $200 with the $50 rebate) I rambled on a lot of this even more extensively here: http://www.atariage.com/forums/topic/121484-why-did-atari-ditch-the-5200/page__st__375__p__2206800#entry2206800

Yes, I'm certain Synertek second sourced MOS parts (let alone Atari's custom versions like Sally), not just 6502/6507, but various support chips (RIOT, PIA, etc), and supplied Apple and others as well, but Atari was their biggest customer by far iirc. Yes, Synertek was set up as Atari's secondary source for 6507's from the beginning. They already had a relationship with them via their PONG IC's, which is how they got Jay Miner. You can read a bit more about the process in an older article of mine - http://classicgaming.gamespy.com/View.php?view=Articles.Detail&id=401 Yes, that's the article I was looking for! And they seem to have been set up as a second source for the whole VCS chipset (my heavy sixer has a Synertek TIA -or maybe RIOT, need to double check, but a lot of other motherboard shots showing a mix of different manufacturers in different combinations -with Synertek seeming to do all 3 chips, not really surprising though but good to note). On that note: do you know if Atari/Warner management ever considered negotiations for a merger with Synertek? (it seems like they had a pretty strong workign partnership with them -at least as far as normal business relationships go- and Synertek does seem to have been in the smaller -and younger- category that would have favored a merger/buyout -probably more so than a company like MOS even, which was an idea tossed around at Atari Inc at one point apparently, and Synertek wasn't snapped up by Honeywell until after '79, so that's a pretty fair window to work with when Atari was doing well on the market -vs MOS which got snapped up by CBM back in '76)

I decided to take some screens from 7800 and NES versions of Donkey Kong, and correct the aspect ratio using the pixel aspect ratios found here While the 7800 version is still a little fatter, it's nowhere as obvious as before. Yep, and those are graphics not intended for the 7800's resolution. (granted, the only other option is modifying the art design more -probably a stylized version of the arcade original, Mario Bros could have been closer to the arcade than the NES game even since that one isn't all that faithful -like the tiny shellcreepers with missing animation- the Atari XE Mario Bros is the closest port to the Arcade at the time iirc, though the colors are limited -a shame the 7800 version wasn't a more colorful version of that game; the A8 DK also has all the levels and looks pretty good given the limitations -better than the CV game in some areas on top of the added content- though the C64 version is arguably better still -made by Atarisoft and very similar to the A8 game but with a bit more color, though weaker sound IMO) Of course, PAL users would have it look pretty much like the double wide emulator screen shots, but also extra wide for the NES (and CV), but proportionally similar. Edit: there's a later DK conversion on the C64 by Ocean that's a good bit improved over the Atarisoft game, and they seem to have used coarser graphics to cater better to the resolution and aspect ratio at the expense of detail. (seems to be especially be converted for PAL, so even more extreme) The sound is definitely better than the C64 atarisoft game. (the A8 sound is arguably better though) http://www.youtube.com/watch?v=vrmR9k8k9vk Edit: Hmm, seems like the A8 version does stylize things too: (DK's aspect ratio is kept better, btu he looks rather different -also caters more to the lower color) So not as close to the C64 as I though.

Edit: Rockwell was definitely one of Atari Inc'c (and Corp's) prime vendors for 6507s and also a fair amount of 6532s. I've had a hard time finding any Atari motherboard shots with MOS ICs on them at all. This one's got a Rockwell CPU, and a Synertek chip plus a Motorola one. (not sure which is TIA and which is RIOT) Also interesting that they were still using sockets for the earlier 4 port models.

Wow, $9 millon for Sega in '86 vs 6 million for Nintendo, really shows how much better Nintendo was organized for the US market. (not too surprising given they'd been attempting to enter the consumer electronics sector for about 3 years by that point) Kind of a shame they screwed up like that though. (ie didn't invest more in building up a US subsidiary infrastructure or licensing to/partnering with a prominent US distributor -Tonka came in '88, a big improvement but not enough to make a real difference by that point) I wonder what Atari Corp's budget was. (presumably in the sub million dollar range) I do wonder if they priced too low though, I think others have brought this up too. $10 seems pretty low, and all they probably needed to do was maintain a significant gap below the competition for that tactic to work well. (it's a good strategy, but this makes me wonder if they took it a little too far, more so with the games than the hardware -software is where the money is, so the $80 price was probably fine -especially if it was for a slim profit margin, maybe they could afford to drop it a bit more even, maybe a "core system" with no game pack-in- but they probably could have pushed for $15-20 initially if not a fair bit more and still had a big price gap from the competition -but more profits to invest back into other things) I wonder what the 2600 games were priced at. Here's another article from early '86 with a similar topic: http://tnca.myrmid.com/art9.htm (that also has a quote from Katz mentioning the strong sales of the 2600 in mid/late '85 and how they had shortages) The reason why Atari is cranking up production of a refurbished 2600 and the new 7800 system is that "last fall, with no advertising or promotional effort on our part, we sold plenty of 2600s. We could have sold hundreds of thousands more if we had the production capacity. It proved to us that the industry is alive and well."

Just imagine a kick-ass custom IBM based machine built around ROMP (or similar), custom OS, custom graphics chipset, custom sound hardware, etc. Nothing posted yet, not sure if there are any pics of it in action. (or code for that matter) He hasn't elaborated more AFIK. Is it confirmed Atari bought 6502s and 6507s from Synertek? The way the press has told it is Atari bought those CPUs from Commodore's MOS. Maybe Honeywell wanted too much money for Synertek in 85 because Tramiel was talking about the need for vertical integration practically right after Atari Corp. was christened. Yes, I'm certain Synertek second sourced MOS parts (let alone Atari's custom versions like Sally), not just 6502/6507, but various support chips (RIOT, PIA, etc), and supplied Apple and others as well, but Atari was their biggest customer by far iirc. Atari had several other vendors as well, I think Rockwell was second sourcing MOS chips for them as well, maybe WDC, not sure. (basically they pushed for the lowest bidder AFIK, and I don't think that MOS parts were super common in later models -my 1977 VCS only has one MOS part, and it's only stamped on the bottom of the chip -need to check if it's RIOT or the 6507, there's 1 Synertek chip and 1 chip from somewhere else -I need to check again) MOS had extremely favorable licensing agreements for their chips (at least up to the CBM merger), so you had 6502s, 6507s, other 6502 derivatives (some licensed and others modified from licensed 6502 cores), RIOT, PIA, VIA, CIA, RRIOT, etc, etc all being produced by a number of vendors as common off the shelf parts. What gave CBM the edge, is cutting out the middle man: ie producing chips at cost (no profit for the vendor) as well as hainv much faster (or more predictable) turn around times for chip deliveries. (as well as profits of being a 3rd party vendor) They wouldn't have to gouge prices at all, in fact they couldn't do that and expect any sales as a 3rd party vendor (due to competition), they just cut out the overhead for their in-house machines. (which could be a small to very substantial advantage depending on the profit margins pushed by vendors on the market at the time -Atari tended to push competitive bids with different vendors to get the best deal iirc -as most reasonably smart companies in the business would have done- and also pushed pretty high volume orders, so they could get pretty good deals on chips -Nintendo never had an in-house vendor, but made some very tight deals for massive orders with Ricoh -one of the ways they managed a highly competitive price with the Famicom at launch against competition using off the shelf parts -and of course also having the afvantage of new, highly integrated LSI chips -CPU+IO+sound and video all on 2 40 pin DIPs- to push for low cost as well -the motherboard is about the size of a ZX81 for god's sake! -imagine if Atair had pushed that for the 5200 -merge POKEY with Sally and use CGIA, plus a DRAM interface ASIC, 2 small 8k DRAM chips, and a very compact motherboard closer in size to the VCS -OTOH the 5200's 1982 chipset should have had a motherboard no larger than the 7800 as it was...) It also shows how much Warner botched the transition . . . all those issues should have been dealt with before the sale went though and (ideally) much of Morgan's plans for NATCO and Atari consumer in general should have continued under TTL/Atari Corp. (they wouldn't have conflicted with Tramiel's plans either, and in fact would mean downsizing in a much more orderly manner than Tramiel was more or less forced to do -it was so much of a mess with the entire staff laid off by Warner, total chaos in the company, noone knowing what the hell was going on, lots of blame and animosity, frustration, lawsuits, etc, etc) Tramiel may have shifted Morgan's plans a little (Morgan wasn't pushing the 16-bit computers very hard), but that's not necessarily a bad thing. (especially since MICKY was dead -honestly I'm not sure how marketable such a high end game system would have been in 1984, maybe they'd use it in the arcade more so early on -there's no way to set up the amiga graphics chipset -let alone sound- with much less than 128k and still have it useful, 64k would be pushing it and you'd have to deal with single buffered graphics and 32k or less of work RAM). Honestly, I don't see the Amiga (or rainbow) chipsets being attractive consoles in any configuration until 1987 at the very earliest. (and even then more dependent on consolidation and features cut out to customize for a console only chipset -you'd really want at least 128k DRAM -plus ROM, of course)

Except for a handful of people who still use their Amigas (or ST/Falcon based machines even) for regular computer tasks. Far less common than PC users sticking to Win98SE or 2000 (both of which have some pretty nice patches and expansions after offcial support was dropped). I know a couple people running win98SE as their main OS (or "supercharged win9x") though more that use win2k. (and obviously a ton who have XP -which is still more efficient and less bloated any 32 or 64-bit vista/7 and more compatible, not sure about 64-bit XP though, but otherwise the main reason to switch to 7 is for 64-bit use or because it's pre-installed and you don't want to deal with transferring a license -the case with my 32-bit vista laptop, though in hindsight I wish I'd transferred by XP license... at least vista/7 32-bit supports DOS shell windows still) Heh, it seems like MS's OSs are most stable and useful just as they're getting close to being discontinued. (in the cases of better OSs at least . . . I don't know anyone who prefers using Win95 -or older- to 98SE, or ME for that matter over 98SE or 2000 -or XP for that matter, unless it's a machine that's really nto well suited to XP's bloat, or at least the post SP1 bloat)

I think it's HIGHLY dependent on the monitor in question. I've seen many cases where the native res is pixel perfect (including the VGA LCD units we've been using for the last ~6-7 years). Maybe newer monitors auto-scale like that, but all the ones I've tried default to square pixels if you've got an unscaled 640x400 source. (some can't scale below that though, including our old GEM monitors, it shown a blank screen if you try 320x240 or 320x200, I don't remember trying 320x400/480 though) My current laptop does the same thing, though I had to disable the stupid "aspect fit" (or whatever) mode the video driver was forcing initially, now it's always square as it should be for most such options. (scaled on a program by program basis, but otherwise with boarders for anything that's not 1.6:1) But back to VGA vs DVI, I've never seen quite what you describe. All VGA inputs to dedicated VGA (and even some mixed VGA/DVI) monitors look great via VGA or via component video adapted to VGA (have a converter). Hell, analog is the ONLY way to get a pixel-perfect image on some displays, especially HDTVs. WAY too many TVs won'y let you use HDMI/DVI at the real native resolution, especially 1365x768 ones (they tell the device that it's 1280x720 native)... the ONLY pixel perfect DVI res we could set up on our Sanyo set was 1024x768 (with pillarboxing), but VGA looks perfect in 1365x768. (totally pixel perfect, you can look right at it and make out all the pixels -which aren't all that small given it's a 32" display at 1365x768 -actually might be 1360x768, I forget) But bottom line is that I haven't seen that issue of forced AA via a monitor (TONs of video driver related problems, but not on the monitor end). I've seen plenty with super blurry AA at nonnative resolutions, though even then it tends to be less as you get closer and closer to the native res. I still prefer CRTs though . . . every time I use one (with proper contrast/color/scan calibration) I get spoiled when I go back to LCDs. (even the high-end LCDs on the ~2 year old Mac Pros at school can't compete with good CRT monitors -though they're getting closer at least, save for variable resolutions which are still a long way off compared to multi-sync CRTs) A shame no HDTVs ever pushed VGA monitor quality CRT displays. (there might even be a market for high-end CRT HDTVs to this day if that were the case . . . albeit interlaced stuff would be really flickery on VGA quality low-persistence phosphor, so mainly good for progressive displays and deinterlaced stuff -a lot of bad deinterlacing out there too, the proer way to do it should be to combing each possible pair of fields for a 50/60 Hz progressive output -2 possible pairs for each field, just like on a true interlaced display: it's really hit and mis to find HD sets that do that even though a lot of lower end LCD SDTVs from a few years ago did just that and even perfectly detected 240p -which a fair amount of HD sets do, but far too many that don't; even stranger that it's often the cheaper HD sets that have much more flexible support -like Sanyo, which is one of the lower end brands now, though I think some Sony stuff is OK, Samsung is sometimes good for SD support, and I think Phillips is as well -a few even support 15 kHz sync through the VGA port )

Huh, I thought the C64 used a colorburst based clock like most other TV-oriented computers. (or others due to the cheap nature of 3.58 MHz oscillators) Yes, if the boarder is like that, that's right. But WOW PAL has a lot of vertical overscan! Just 256 lines of 288/576i for the standard, that's WAY more than average (ie too much overscan) NTSC sets show about 224 lines (better sets show very close to 240 lines (480i) with the same PAR), so that's a huge gap compared to NTSC and means a less dramatic PAR difference than I expected. (there's a MUCH higher percentage of unused lines in PAL then with 256/312 vs 224/262, and again a much bigger difference from the "standard" 288 vs 240 lines) From talking with some other PAL users in the past, I'd actually gotten the opposite impression, that average newer PAL sets showed about 272 lines or in some cases closer to 266 lines. (so 256 lines would leave a boarder . . . if that was the case) But if it's really ~256 lines to the edge (as much as 224 lines is for NTSC), that's really close to perfect with 7.16/7.14 MHz. (the ST in PAL would still be closer than the Amiga on NTSC TVs though) This info also changes my perception of MD/Genesis H40 (6.7 MHz dot clock) in PAL, it's actually closer to square in PAL than NTSC then vs ~7-9% off in both cases. (6.25 MHz should be pretty close to perfect for NTSC, at which you'd get ~298 visible pixels horizontally and 224 lines -or close to 320x240 with perfect overscan calibration, I think the N64 runs at that rate for 320x240 mode, not positive though -the Neo Geo is 6 MHz iirc, so also very close, various 5.37 MHz examples are a fair bit wide though)

Here's how my modern ($5/Goodwill) VGA monitor displays the ST monochrome, with one of those eBay cables: No, I meant LCDs, not CRTs... CRTs you can make any aspect ratio you want (and easily on newer ones too -just go into the menu). For LCDs it's square pixels by default with a boarder, you need a video driver set to scale otherwise. For a CRT (especially a modern CRT), simply go into the menu and adjust the overscan yourself. (that's how I get 1920x1080p in 16:9 on my 20" 4:3 CRT monitor ) So the question is: what's the default calibration and how easy is it to change it. (ie external POTs, menu, or opening the case to adjust the pots internally) Oh, and I looked a bit closer on this: http://webspace.webring.com/people/tg/geosteve_99/ST-Mac.jpg and both are about 11% off from square in terms of aspect ratio. (ST very slightly more, closer to 12%, but both are about as far from square as the Genesis is on NTSC or PAL displays -NTSC is tall, PAL is wide)

Can you explain how? ??? An 8 MHz dot clock (320 wide mode) is very close to what PAL (50 Hz vsync 15.6 kHz Hsync) needs for square pixels with standard calibration, they'll be a little bit tall normally, but barely. Amiga (and A8/C64/etc in 7.16 MHz dot mode) is the opposite, very slightly wide pixels in PAL. (ideal square pixels is ~7.5 MHz) In NTSC, they're all way, way off from square without custom calibration: you'd need to ajust the V/H overscan to get the PAR (pixel aspect ratio) you want, and some monitors only did that internally (ie opening the case), others didn't. With an SD (15 kHz) monitor at NTSC spec, you'll have tall pixels for the Amiga and ST (taller for ST), and also a considerable boarder on all sides (it should be a 320x200 window while the screen is actually showing 224 lines -or closer to 240 in some cases- and either ~382 pixels wide or ~342 pixels wide for Amiga with NTSC calibration) CGA is identical to Amiga in that respect. (320 wide is 7.16 MHz, 640 is 14.3 MHz) With custom calibration you could have square pixels, or a screen that is stretched exactly to the edge with almost 0 overscan. You can't get 720x200/400i on the Amiga withotu custom calibration since normal NTSC (or PAL) calibration would have much of those 720 pixels in overscan. (at 14.3 MHz, you'd have a max of ~684 pixels visible with normal calibration) Some older TVs had external pots (on the back usually) to adjust the overscan (talking 60s and 70s TVs here, maybe early 80s), and many monitors do, most newer TVs offer it in the service menu. I'm not sure what the Amiga and ST had to offer on average, but I know some had the POTs internally. A lot of PC monitors had external knobs (and later digital menus) to adjust the overscan manually. (especially multi-sync monitors where it's a necessity) You see a ton of ST and Amiga monitors calibrated way off from NTSC or PAL spec, more like manually adjusted PC monitors where you fill to the edge of the screen. (a normally calibrated monitor should have a pretty massive boarder on the ST or Amiga -like on a TV) There's also a lot of Amiga software that assumes something roughly like 320x200 stretched to 320x240 (.83:1 PAR), which would obviously look very wrong for PAL users, rather like this image with square pixels: http://toastytech.com/guis/wb_10.gif The high-res monitors are totally separate from the SD calibration though, but in any case, this random example seems pretty close to square: http://www.jeffn.com/vintage/sm124.jpg That's also how modern VGA monitors will display it by default: eBay Auction -- Item Number: 320526274539 (albeit in this case, it has the boarder all at the bottom) Hmm, actually, if this is anything to go by, average ST calbibration is closer to square than the Mac monitor. http://webspace.webring.com/people/tg/geosteve_99/ST-Mac.jpg So in the end, it's really up to the monitor you've got and/or if you're willing to open it up if it doesn't have scan adjustment pots externally. You could even have square pixels with 640x200 (amiga, ST, CGA, EGA, etc), but that would mean a pretty short display (huge letterboxing), something like this: http://upload.wikimedia.org/wikipedia/commons/8/83/Arachne_CGA_Mode.png Hell, the Apple II and IIGS could easily have square pixels if calibrated as such, at least in the 280x192/320x200 mode (in PAL, they'd be almost perfectly square as-is, like the A6/C64/Amiga in 320 wide res mode).

I think they got it working on a production level: it was fully prototyped back in '83 iirc and there's some quotes from '84 and '85 listing deadlines for full production. Atari Corp continued to work with some of the ATG staff from the project (namely Sam Nicolino) to finish the chip design (Nicolino left after completing AMY-B, I believe), so the major snags many have been down to implementing workable development tools for the chip. It wasn't going to be really cheap either, it was a full 40 pin DIP that also required an external 16-bit DAC (or lower res if wired properly), so an internal DAC would probably have been one of the first cost reducing revisions. (not only saving the cost/space of an external DAC, but also cutting the package down to a 24 pin DIP due to the removal of the parallel lines and assuming the 2 "reserved" pins were removed. (you'd still have enough pins for all functionality of the 40 pin version plus 2 analog lines out for stereo, or limit it to mono) Adding logic for a direct write mode on the DAC would be nice too. (and for early models with an external DAC, you'd need a bit of logic to provide a parallel interface to the external DAC -or just modulate the volume/amlitude of one of AMY's oscillators set to a very high frequency -assuming the amplitude response is fast enough) I have no idea why they wanted to put it in the XE, dual POKEYs would have made more sense there (with potential to cut down the 2nd POKEY as well -remove all I/O stuff and cut to 24 pins or less). AMY should have been aimed at the ST if anything, though it's not like PAULA at all, and not like FM synth either, it's pure additive synthesis (though only 8 frequency rams with the remaining 56 oscillators slaved to those 8 freqency settings and only amplitude modulation control for those slaved oscillators). It would probably be less impressive overall than Yamaha's YM2151 released in '85 (32 operator/oscillator sine wave device, but configured into 4 op FM synth channels with a variety of algorithms -including simple additive modes- so much more capable in spite of 1/2 the oscillators -and also 8 voices, unless you added channels for added effects/complexity, similar to AMY's 8 frequency ramps effectively limiting it to 8 voices -more only if you used some for simpler fixed pitch sounds, sort of like how Yamaha's OPL series allowed 3 2-op channels to be slaved for 5 percussion channels -plus other trade offs like Yamaha's envelope control and other effects). Given the YM2151 was new, fairly expenive, and in limited quantity, the in-house AMY would have been far more attractive. (even in the long run once further consolitation took place) Amy was a synth chip, not a PCM device, but an interesting one, and probably would have been a good addition to the ST (especially in the consolidated 24 pin package), let alone the intended Rainbow based machines. (and true synth also means lack of memory hogging samples -aside from a few for software based PCM -prior to adding DMA sound) I'm really not sure what the problems were, but maybe Curt and Marty have more details. (the synth mechanism doesn't seem all that complex compared to contemporary synthesizers, and even if their in-house engineers couldn't put good tools together, I don't see why they couldn't have done a bit of outsourcing -shouldn't have required anything extravagant, especially if they could contract some of the original developers from the project) Hell, in the practical sense, they could have saved cost in other areas in the ST if they dropped the YM2149: they could have replaced it with a CIA or VIA and also cut out one of the ACIAs (and still had extra parallel I/O lines over the YM). OTOH, without the YM and if AMY was delayed beyond the ST's release date, the ST would have no sound hardware or suffer a later release. (maybe they could have tossed in a plain DAC to be driven by the CPU and/or interval timers -or multiple DAC channels for that matter, short of actual DMA sound -would be weak for games, but AMY should have come pretty soon and they could have kept provisions for such an add-on or a socket to drop it into the motherboard at a service center) Hell, bare DACs would have been worse for games, but better for music applications and such: and 4 bare 8-bit DAC ports (be it separate DACs or a simple IC mixing 4 ports to a single internal DAC) you could emulate PAULA much more easily than software mixing (still a ton of overhead for having to write every sample, but no mixing or software pitch bending needed -or simple multiplex mixing for fixed pitch samples on 1 channel and variable sample rate on the others), and again, interval timers may have been able to do square wave tone generation in hardware as well. (a la PC speaker, but possibly multiple channels) Let alone with 12 or 16 MHz models driving that. And given the higher-end nature the ST initially had, general game performance wasn't as critical as long as it was a temporary early issue that got corrected, especially in Europe, and especially due to the built-in sampling capabilities it would add. (hardware scrolling and added sound should have been upgraded across the board early on though, and probably a socket left for AMY -or a DMA chip to drive the DACs- as well as have SHIFTER socketed to allow a scrolling model to be dropped in when available -sort of like GTIA) 12 MHz (maybe even 16 MHz) model should have been there by day 1 though, regardless of other hardware compromises. (the only modification needed would be for wait states to be used in place of the interleaved DMA -and you'd still get accelerated internal operations beyond that for the many cases of longer periods between bus accesses -vs the 650x) Sound, hardware scrolling, and faster CPUs were the 3 things in hardware of the ST that should have been there from the start (CPU optionally), and if delays were unacceptable, at least provisions for simple upgrades along with standardization on all production across the board for scrolling SHIFTERs and better sound. (be it YM2149 to YM2203 or bare DACs to DMA, AMY, etc, etc) Other upgrades like a blitter, 32-bit models, 12-bit RGB (or higher), more bitplanes, packed pixel modes, etc, etc could come later. (not only would those things compete more with the Amiga, but it would put a larger gap against PCs and other competition as well as makign the machines an even better value -scrollign should have added negligible production cost, mainly R&D, and the sound system had the aformentioned options for cost cutting anyway -ie a CIA/VIA displacing YM+ACIA, faster CPUs would be an optional thing initially for higher end models) Desk top form factor definitely too, from the start. (alongside low end models) Aside from any potential for the ATG hardware and OS. But we're nowhere near the 5200 topic anymore.