Misguided criticism of the Atari 7800

[Bryan]

Software. At the end of the day, that's what it all boils down to, isn't it? The NES had the games people wanted, the 7800 didn't. Easy as that. Heck, if hardware was the only consideration, the Jaguar wouldn't have had trouble against the SNES and Genesis, right? And, the PS2 wouldn't still be outselling the xbox and Game Cube.

 

Yup. That's how the lame-o B&W GameBoy clobbered all the other handhelds. I may be wrong, but I feel Nintendo was the first to really shift development focus from arcade ports to original games.

 

-Bry

[HawgWyld]

Heck, let's face it. The old "software" game even allowed the Atari 2600 to stay ahead of the Intellivision. Well, the Intellivision controllers were rotten, of course, but that's another story.

 

And, yeah. Go pick up a Lynx, Sega Game Gear or Sega Nomad. Now, compare any of those systems to the Game Boy. Now, did the superior system win out there? Heck, no!

 

Software's everything (or, in the case of the Game Boy, software and a long battery life).

[Curt Vendel]

Well I won't doubt your research, but if the 7800 was so superior to the NES then why do even the first NES games (Clu Clu Land, Ice Climber, Balloon Fight, etc.) look better than the last 7800 games? Face it, the NES was better. Even an Atari fanboy like myself has to admit it.

 

1. The argument was whether or not the hardware was dated as many (including you) claim. I don't feel that's valid.

 

2. Point 2 is a good one. but is it an issue of hardware or;

 

* Better programmers

* Longer development cycles

* Third, fourth, fifth and sixth generation programmers pushing the system.

etc etc etc

 

The NES did have better games and it was pushed harder.

 

But my argument stands - if the 7800 was "pushed" in the same way as the NES was, what would the results be like? And no - I don't feel that the first NES games look better than the last 7800 titles.

 

Please feel free to tell me why the 7800 couldn't play Zelda on a technical level.

 

He's got your there Tempest... The NES had a LOT better development support. Also when the 7800 was released, it was released with just a few of its original 1984 games and didn't have new s/w for months afterwards. I think better development tools would've given the 7800 a fair advantage. Also NES was advertising head over heel.... meanwhile most people thought the 7800 was leftover hardware on the shelves.... Other then Activision and Froggo, the 7800 had little support. It needed Acclaim, EA and others.

 

 

 

Curt

[Clint Thompson]

Nice thread.... how I stumbled upon this old thread is beyond me (oh, I was looking for a 7800 game name

 

I think the 7800 did what it could considering.

 

It is too bad Atari was in bad shape around that time and couldn't give more to the 7800. I want to see Tetris and mostly Marble Madness... I don't need SMB for the 7800 because I may as well get the NES for that.

 

I do feel more arcade ports or originals are a must

I haven't had a 7800 since around 1996 when most of my 2600 games and 7800 titles got flushed out in a flood (wow, how many times have I said that

 

What's the issue with Galaga? I thought the 7800 version rocked! So much I played it non-stop summer of 95/96! It does tend to get slow/sluggish in later levels which can somewhat make it unenjoyable because your trying to focus and the computer is going too quickly. Yet while I hear people praise Robotron I didn't really care for the 7800 version, weird.

 

Tower Toppler was awesome and I can only hope to be able to play Midnight Mutants someday (no no no emulation).

 

Where oh where is my 7800...

[Lord Thag]

While the 7800 may not have been quite as powerful as the NES in some ways, it made up for it in others. Frankly, raw power had nothing to do with it. The reason it failed was because thr Tramiels wouldn't throw any money at it, the marketing campaign was horrible, and most of the launch games were re-releases of classic games (usually games that had SEVERAL prior releases as it was). Few consumers wanted to play asteroids and centipede again, even if they were great ports. The NES had new, original games, and much better marketing. I think the 7800 could have competed just fine if it had had piles of fun, original gmes to play. The fact that it was technically somewhat inferior is meaningless. The PS2 is technically inferior to the XBOX, yet sony holds the largest marketshare. Why? The games. Too bad. The world missed a very fun system that never got off the ground. Hopefully we can ressurect it on the homebrew scene

[Foxsolo2000]

I agree with Lord Thrag and the others. The 7800 had a lot of potential and I for one eagerly rushed out and bought it only to be be disapointed with the games that came with it. I was also a big supporter of the Jaguar and rushed out and bought that too only to again feel disapointment at some of the truly pathetic games that came with it. One or two stood out like gems but like the 7800 there was not enough decent games to keep you occupied on it.

History shows you that the winners in this market are the ones not only with the best games but the most as well. That is why the 2600 lasted as long as it did, the same with the Sega MegaDrive/Genesis, gameboy etc for there was the sheer volume of software to support it. Tramiel was too short sighted to make the 7800 work for there was very little advertising for the system and i only discovered that it had been released by accident by reading through a game magazine of the time.

Had it been released back in 84 Atari would have had that vital head start over everyone else and probably have had the time to develop enough software to keep alive the interest in it. As it stands the opportunity was forever lost and Nintendo resurrected the gaming culture all by themselves. Who knows, perhaps in a parellel universe the 7800 was the monster hit it should have been. Guess we will never know!

[Harry_Dodgson]

:-(

 

I disagree with most of this thread. The 7800 is a pain to program for. I suspect anyone that was interested in developing for it in the 80's saw the programming specs and ran away screaming. The documentation and examples are still sparse even now.

 

The Atari dev tools are pathetic. The Madmac assembler is a 68K assembler retrofitted for 6502 code. The suggested graphics tools are pretty lame too. When I found out how much easier the Lynx was to program, I gave up on the 7800. (until my recent bout of forgetfulness which caused me to try and create something for the Homebrewerpalooza contest)

 

If "they" had spent an extra few pennies on a 6532B, the ram and timer would be usable in 7800 mode. As it is, the only stable time base is the horizontal and vertical retrace timing. Depending on PAL/NTSC, "your mileage may vary". Even a simple software loop is not deterministic due to DMA.

 

If anyone thinks no-one has pushed the system - they are mistaken. It has been pushed almost as far as it can be. Unlike the 2600 and 800 which allow fancy video tricks on the fly, the 7800 buffers everything internal to the Maria chip where it can't be accessed.

 

Yes, it has lots of "sprites", but it takes more code than it should to move them vertically. It has 25 colors out of 256 in 160 mode. That drops to 9 colors in 320 mode (unless you do palette or color cycling).

 

Harry

[Lord Thag]

Interesting. I'd never heard the system described from a technical standpoint before. I'm familiar with 2600 programming, but have never touched the 7800 architecture. I didn't realize it was such a bear to program. It's another good example of how good 'ol Jack wanted to skimp on the cash during the design phase. I always thought it idiotic that a next gen system used a last gen sound chip. DK on the 7800 makes my ears bleed . Still, if Atari had pushed it earlier, I still think the 7800 could have competed. Not won, perhaps, but competed. Too bad.

[EricBall]

I suspect anyone that was interested in developing for it in the 80's saw the programming specs and ran away screaming.  The Atari dev tools are pathetic.  The Madmac assembler is a 68K assembler retrofitted for 6502 code.  The suggested graphics tools are pretty lame too. 

Hmm... I hadn't thought of things from the perspective of what was available in 1984/86. I hope that things are a little better today.

 

If "they" had spent an extra few pennies on a 6532B, the ram and timer would be usable in 7800 mode.  As it is, the only stable time base is the horizontal and vertical retrace timing.  Depending on PAL/NTSC, "your mileage may vary".  Even a simple software loop is not deterministic due to DMA.

 

If anyone thinks no-one has pushed the system - they are mistaken.  It has been pushed almost as far as it can be.  Unlike the 2600 and 800 which allow fancy video tricks on the fly, the 7800 buffers everything internal to the Maria chip where it can't be accessed.

I don't know if the absolute limits have been reached. We homebrewers are still working on our first games. That being said, I agree that the asynchronous nature of MARIA takes some getting used to and DMA can really cut into the CPU budget. In theory, software timing loops shouldn't be necessary. The game simply needs to ensure a complete DLL is available every frame.

Yes, it has lots of "sprites", but it takes more code than it should to move them vertically.  It has 25 colors out of 256 in 160 mode.  That drops to 9 colors in 320 mode (unless you do palette or color cycling).

Building DLs to start with takes too many CPU cycles. Vertical motion just means the sprite appears in multiple DLs. Vertical scrolling (with a tiled background) is even worse.

 

I'm not sure how you get your colors though. But I will agree that everything except 160A is very strange and somewhat limitting.

[Trip_Cannon]

Interesting. I'd never heard the system described from a technical standpoint before. I'm familiar with 2600 programming, but have never touched the 7800 architecture. I didn't realize it was such a bear to program. It's another good example of how good 'ol Jack wanted to skimp on the cash during the design phase. I always thought it idiotic that a next gen system used a last gen sound chip. DK on the 7800 makes my ears bleed  . Still, if Atari had pushed it earlier, I still think the 7800 could have competed. Not won, perhaps, but competed. Too bad. 

837793[/snapback]

 

Jack had nothing to do with the design of the 7800 console. It was created and ready to go before he took over Atari corp.

[Lord Thag]

Jack had nothing to do with the design of the 7800 console. It was created and ready to go before he took over Atari corp.

 

Ah, that's right. He shelved it at launch right after he took over, right? Still, he didn't want to pony up the cash to put the sound chip in for most of the carts, and he barely funded the whole operation as it was. Too bad. I bet we missed some great games that 'might have been' if more love had been shown the system.

[DracIsBack]

If anyone thinks no-one has pushed the system - they are mistaken.  It has been pushed almost as far as it can be.

 

Sorry Harry, I don't buy that.

 

For the simple reason that any system that people have spent a lot of time with learning its ins and outs always seem to surprise people.

 

Who would have thought the 2600 could do digital video? That the Genesis would do rendered graphics? There are so many examples of people surprising others with what they churn out of the technology in front of them.

 

But that requires on-going dedication. On-going interest. And on-going experinmentation.

 

I am skeptical that will ever happen on the 7800. At the same time. I just don't believe in the naysayer attitude.

[Clint Thompson]

I agree that there's certainly room for untapped potential within the 7800 wth tricks / surprises or whatever you would like to call it

 

Now ... I'd like to see it!

[Bruce Tomlin]

I disagree with most of this thread.  The 7800 is a pain to program for.  I suspect anyone that was interested in developing for it in the 80's saw the programming specs and ran away screaming.  The documentation and examples are still sparse even now.

That it is. But I think we've got it documented about as much as it can be. There may be tricks, but the design of the Maria was clearly so that you shouldn't have to rely on tricks.

 

The Atari dev tools are pathetic.  The Madmac assembler is a 68K assembler retrofitted for 6502 code.  The suggested graphics tools are pretty lame too.  When I found out how much easier the Lynx was to program, I gave up on the 7800. (until my recent bout of forgetfulness which caused me to try and create something for the Homebrewerpalooza contest)

When looking at the devcard documentation, I was wondering why they had info on a 68K assembler in there. Geez.

 

I've had my own assembler for a long time. Last year I finally converted it from Pascal to C so that I could support more CPUs without duplicating all the work in making it support macros and stuff. I've even made it accept most kinds of legacy code pseudo-ops I find so I can assemble them with minimal or no editing. With that, a CC2, a signature generator, and a decent emulator that runs on the Mac (the recent MESS upgrade), the tools issue is pretty much unimportant to me. Except for the painful lack of a 320B mode graphics editor, that is.

 

Nonetheless, I still look longingly to the Colecovision, on which I also have the advantage of a real Z-80 ICE debugger with a full 64K of overlay RAM. And nice support for character mode graphics.

 

If "they" had spent an extra few pennies on a 6532B, the ram and timer would be usable in 7800 mode.  As it is, the only stable time base is the horizontal and vertical retrace timing.  Depending on PAL/NTSC, "your mileage may vary".  Even a simple software loop is not deterministic due to DMA.

I think the RIOT timer isn't really that necessary. The only thing it got used for most of the time on the 2600 was vertical retrace timing, and the Maria frees you from that. The RAM is still usable, but it's this 128-byte chunk in the middle of nowhere. I ended up using it for a machine language monitor using the right joystick as a 2400 baud serial port, only needing 8 bytes of zero page for the pointer addressing modes.

 

The DMA on the other hand is the big problem. I couldn't even do WSYNC timing for RS-232 with DMA turned on. But the non-determinism isn't the real problem of the DMA. It's the Maria's vampire-like drain on CPU cycles which is the 7800's main failing. In my first attempt, I naively rebuilt a complete DL and DLL using 6502 code, which I found out could take more than 1/60 of a second to execute. The NES (and for that matter, the Colecovision) used a separate bus for video memory, so the CPU could run at full speed all the time.

 

And while the Maria has some kind of support for a character cell mode, it's not all that great. And it does one thing at a time, then another. Whether or not two items overlap each other on a scan line, BOTH of them get DMA fetched and stuck in a buffer. So it has to copy your background, then it overwrites its buffer with sprites. If it runs out of time, tough. That's what caused the arcade version of Gauntlet to have sprites disappear when you had lots of enemies on the screen, by the way.

 

The way Maria did DMA was a pretty good idea. Making it contend for the same bus as the 6502 was not.

 

If anyone thinks no-one has pushed the system - they are mistaken.  It has been pushed almost as far as it can be.  Unlike the 2600 and 800 which allow fancy video tricks on the fly, the 7800 buffers everything internal to the Maria chip where it can't be accessed.

Pushed in terms of tricks, yes. Pushed in terms of games which rely more on large ROM size than on high sprite count (such as RPGs), no. Still, I think there could be a few tricks there, just not ones that let you screw around within a single scan line.

 

Yes, it has lots of "sprites", but it takes more code than it should to move them vertically.  It has 25 colors out of 256 in 160 mode.  That drops to 9 colors in 320 mode (unless you do palette or color cycling).

Not to mention that when you use a lot of sprites, you pretty much have to have black background. If Robotron had a picture background, it might not have worked. If it had a picture background AND was in 320 mode, I don't think there would have been enough DMA. Someone did a test of a parallax scrolling background a few months ago. His code worked great on an emulator, but completely failed on real hardware.

 

And those 9 colors in 320 mode? You can only use three of them at a time in a given DL entry in 320B mode, and the other 320 modes have even more difficult to use color mapping. If you use character mode, the entire stripe has to use the same three colors. And then there's that funny bit mapping you have to do. Making all the graphic tiles for Tubes was a real chore.

 

Hmm... I hadn't thought of things from the perspective of what was available in 1984/86.  I hope that things are a little better today.

About the only thing that is better is that computers today have well over 1000 times the speed and disk space, so running an assembler for 8-bit code is trivial, and there is no need to break your source up into 16K chunks just because that's all the memory your editor could use, or (in a real games company) spending six figures on a VAX to store and assemble your code. The Powerbook in my lap makes a VAX-780 look like a toy.

 

Building DLs to start with takes too many CPU cycles.  Vertical motion just means the sprite appears in multiple DLs.  Vertical scrolling (with a tiled background) is even worse.

Yes, but it's still a major pain in the butt to place them in the right DL line. First of all, you almost always have to put them in at least two zones, and it can be any two zones and you have to figure out which ones, whereas a on system with "real" sprites, you just set the X and Y location at the same memory addresses. Second, it's a pain to use the 6502 to append things to the end of the lists, even if you have the CPU time.

 

I'm not sure how you get your colors though.  But I will agree that everything except 160A is very strange and somewhat limitting.

Maria 320 modes are really a pain to use, have very limited choices of palette registers they can use, and the fourth color is always either the background or transparent. 9 colors divided by 3 colors per mode means (without looking at Maria docunentation) that it lets you choose three palette registers.

 

And then there's how two bits of the graphics mode are in a register which you pretty much have to use a DLI to update properly if you want different groups of scan lines in different modes, but the other bit was in a DL entry making you have to do things in two different places to set your graphics mode.

 

And the 160 modes really look cheap. Especially on my HD-ready CRT in progressive scan conversion mode. At least the one good thing about 320 mode is that the pixels are roughly square.

[DracIsBack]

Here's fundamentally why I have a problem with people saying "the system has been pushed as far as it can" or "I know how far it can go":

 

People making claims like that are not actually TRYING to push the 7800 the way that other systems have been pushed.

 

Take Donkey Kong Country. When it first came out, it shocked the hell out of everyone. People were in awe that this rendered game was pulled off on a "lowly Super Nintendo". Nothing before Donkey Kong Country looked even remotely as good and it was sharp enough looking that the reviewers were making comments like Atari and 3DO must be shaking in their boots.

 

Now a couple of things about Donkey Kong Country:

 

First, it came out in 1994 ... well into the lifecyle of the SNES. It also took Rare and Nintendo a number of years to figure out HOW TO DO the game on the SNES. The game had a gigantic budget. New dev tools and techniques were created in order to get the game running on the SNES. Developers had already had several years of experience making code and learning how to optimize the SNES.

 

All of those things resulted in a game that completely raised the bar. People thought the SNES had reached its limit. When DKC came out, the limit changed.

 

 

On the 7800 side of the fence:

 

* Jack Tramiel did not invest in dev tools to help out (as Harry noted), but what if he had?

* Jack Tramiel did not like to pay top quality developers, such as those who created Donkey Kong Country ... but what if he had?

* Jack Tramiel did not pay for long development cycles to figure out how to achieve fantastic results, but what if he had?

* Jack Tramiel did not pay developers to make generation of game after generation of game so they could get more and more familiar with the system ... but what if he had?

* Jack did not like to pay for extra RAM or larger games often ... but what if he had?

 

By those five points alone, I do not think the 7800 has been pushed. I do not think its market position fostered what I considered to be "wow development" conditions. And likely, we'll never see it.

 

But I have trouble with "X system can't do Y" or "X system has been pushed to the max". Every time I hear someone say that, I see something that blows me away on the popular systems.

 

Donkey Kong Country on the SNES

Toy Story on the Genesis (compare this to Altered Beast!)

Solaris on the 2600 (compare to Kombat)

Mortal Kombat on the Sega Master System (compare that to Alex Kidd)

 

On the popular systems, there are too many examples of programmers pushing the bounds of what was thought possible.

 

On the 7800, there hasn't been the inclination to ever TRY TO PUSH THE BOUNDS. As such, I don't believe the world has seen the best it has to offer.

Edited April 16, 2005 by DracIsBack

[Clint Thompson]

If I ever hit the big ol' $250+ million lottery you can bet that I'd be the first person to plop down at least a few cool million to get some new game devving going onto push that sucker! Haha... that'd rock so hard and I'd love every penny being spent!

[DracIsBack]

Jack had nothing to do with the design of the 7800 console. It was created and ready to go before he took over Atari corp.

 

Ah, that's right. He shelved it at launch right after he took over, right? Still, he didn't want to pony up the cash to put the sound chip in for most of the carts, and he barely funded the whole operation as it was. Too bad. I bet we missed some great games that 'might have been' if more love had been shown the system.

837849[/snapback]

 

That's one area he could have continued. GCC wanted to have a lower cost sound chip (GUMBY) for the carts to enhance the sound but it was killed when Jack shelved the system. Had it been inexpensive to produce (compared to POKEY), perhaps more games would have had enhanced sound than the two that were released.

[Bakasama]

If I ever hit the big ol' $250+ million lottery you can bet that I'd be the first person to plop down at least a few cool million to get some new game devving going onto push that sucker! Haha... that'd rock so hard and I'd love every penny being spent!

838172[/snapback]

 

New games for the 7800 does sound nice but I wonder with that kind of could somebody could stick things like batteries, extra RAM, or better sound for the 7800. Heck, if I had that kind of money, I might try to make something like a Supercharger for the 7800.

Edited April 17, 2005 by Bakasama

[Trip_Cannon]

Jack had nothing to do with the design of the 7800 console. It was created and ready to go before he took over Atari corp.

 

Ah, that's right. He shelved it at launch right after he took over, right? Still, he didn't want to pony up the cash to put the sound chip in for most of the carts, and he barely funded the whole operation as it was. Too bad. I bet we missed some great games that 'might have been' if more love had been shown the system.

837849[/snapback]

 

That's one area he could have continued. GCC wanted to have a lower cost sound chip (GUMBY) for the carts to enhance the sound but it was killed when Jack shelved the system. Had it been inexpensive to produce (compared to POKEY), perhaps more games would have had enhanced sound than the two that were released.

838198[/snapback]

 

Heck, if Jack and the new Atari were to keep that relationship with GCC going who knows what we would have seen from them! They were pushing out amazing arcade conversions like they were nothing! Nobody knew the system better than they did. I think we would have seen some great stuff...

[Bruce Tomlin]

New games for the 7800 does sound nice but I wonder with that kind of could somebody could stick things like batteries, extra RAM, or better sound for the 7800.  Heck, if I had that kind of money, I might try to make something like a Supercharger for the 7800.

There already is one. It's called the "Cuttle Cart II". It's got the potential for a game with a total of 512K RAM/ROM, and loads games from flash cards.

[128bytes]

New games for the 7800 does sound nice but I wonder with that kind of could somebody could stick things like batteries, extra RAM, or better sound for the 7800.  Heck, if I had that kind of money, I might try to make something like a Supercharger for the 7800.

There already is one. It's called the "Cuttle Cart II". It's got the potential for a game with a total of 512K RAM/ROM, and loads games from flash cards.

839079[/snapback]

 

 

On my imaginary "If I had the free time I would" list is to investigate hacking a 7800 game (Donkey Kong, DK Jr. or Mario Bros. would be fine first choices) to take advantage of the sound in a Pokey Chip onboard a Cuttle Cart II. Hopefully someone will beat me to the punch.

[AEX]

Jack had nothing to do with the design of the 7800 console. It was created and ready to go before he took over Atari corp.

 

Ah, that's right. He shelved it at launch right after he took over, right? Still, he didn't want to pony up the cash to put the sound chip in for most of the carts, and he barely funded the whole operation as it was. Too bad. I bet we missed some great games that 'might have been' if more love had been shown the system.

837849[/snapback]

 

That's one area he could have continued. GCC wanted to have a lower cost sound chip (GUMBY) for the carts to enhance the sound but it was killed when Jack shelved the system. Had it been inexpensive to produce (compared to POKEY), perhaps more games would have had enhanced sound than the two that were released.

838198[/snapback]

 

Heck, if Jack and the new Atari were to keep that relationship with GCC going who knows what we would have seen from them! They were pushing out amazing arcade conversions like they were nothing! Nobody knew the system better than they did. I think we would have seen some great stuff...

838875[/snapback]

 

While I agree the "what if" arguments that the 7800 could have been great, the simple matter is that the Tramiels were not "of the video game" market and were out to make Atari a computer oriented company. The "new" Atari had just enough cash to "get by" for the first 2 years while they established the ST line, and to a lesser extent the XE line while selling warehouses' full of Warner inventory.

 

By the end of 1985 it was becoming clear that the memories of 1983/84 were beginning to fade and a new generation of kids wanted video games to play. It could have been the IBM NES that was launched - it didn't matter, but it was done at the right time and with a fresh attitude to the market. "If" the Tramiels had seen the 7800 as a clever cash cow in 1984 and plumped some money into the project, it could have been the NES of the time - but that wasn't the initial direction of the new Atari, and I guess they realised that a little too late.

 

Although they wanted Atari to = Computers, they realised too late that Atari already = video games, and they lost that crown to Nintendo. In the back of their minds they knew they could have had it all I'm sure, but that was the way it went for Atari and although they "sort of" tried to regain the VG market, they didn't put any real effort into it until the Jaguar, and even then it just didn't go right. But the Jag is whole different story

 

Karl

[Harry_Dodgson]

I think the RIOT timer isn't really that necessary.  The only thing it got used for most of the time on the 2600 was vertical retrace timing, and the Maria frees you from that.  The RAM is still usable, but it's this 128-byte chunk in the middle of nowhere.  I ended up using it for a machine language monitor using the right joystick as a 2400 baud serial port, only needing 8 bytes of zero page for the pointer addressing modes.

 

The DMA on the other hand is the big problem.  I couldn't even do WSYNC timing for RS-232 with DMA turned on.  But the non-determinism isn't the real problem of the DMA.  It's the Maria's vampire-like drain on CPU cycles which is the 7800's main failing.  In my first attempt, I naively rebuilt a complete DL and DLL using 6502 code, which I found out could take more than 1/60 of a second to execute.  The NES (and for that matter, the Colecovision) used a separate bus for video memory, so the CPU could run at full speed all the time.

 

And those 9 colors in 320 mode?  You can only use three of them at a time in a given DL entry in 320B mode, and the other 320 modes have even more difficult to use color mapping.  If you use character mode, the entire stripe has to use the same three colors.  And then there's that funny bit mapping you have to do.  Making all the graphic tiles for Tubes was a real chore.

 

 

If the timer was stable, I'd use it for game timing (especially sound timing) as opposed to Mstat or Wsync. It would also make timing for serial I/O much easier. The Monitor cartridge used 7 Wsync then 8 Wsync every 2 bits to get close to 2400 baud. There are a few CPU cycles between Wsync and DMA start, but not many.

 

In 320A mode you get color 2 of each palette, so that is 8 palette colors + background color = 9 colors. The Bumpertank demo for the Monitor cartridge used all of them at once.

 

If you took more than 1/60 sec to rebuild DLL and DLs then you are doing it wrong. First, don't redo the DLL each frame. I copy it to RAM and have it set with plenty of space between the start of each DL. It is a pain having to rebuild all the DLs every frame that something moves.

 

Yes, making the graphics fit into the ROM isn't easy. I converted a little C program (that I had written for the Dreamcast VMU) to change 1-bit .BMP graphics to Madmac ASM text for 320A mode.

 

One thing that can be done now that wasn't then is the use of faster ROM. With 120ns (or better) ROMs, a program can be written that doesn't have to move all the screen data into RAM. A whole screen could be replaced just by changing the DLL start address.

[Bakasama]

New games for the 7800 does sound nice but I wonder with that kind of could somebody could stick things like batteries, extra RAM, or better sound for the 7800.  Heck, if I had that kind of money, I might try to make something like a Supercharger for the 7800.

There already is one. It's called the "Cuttle Cart II". It's got the potential for a game with a total of 512K RAM/ROM, and loads games from flash cards.

839079[/snapback]

 

Is that what the Cuttle Cart II can do? Gee, I thought it just nifty multicart device. I had no idea you can make games far larger than any known 7800 games. I wonder what kind of homebrew games can you make with that.

[DamienC]

Let me first say that the 7800 is, hands down, my absolute favorite console of all time. I could sit in front of a TV and play Commando, Joust, Robotron: 2084 and Ballblazer for weeks...

 

Having read through this entire thread and heard this debate several times, I just wanted to drop in my two cents:

 

While I firmly believe that a lot of NES-to-7800 ports can be done and done well, I personally would rather see more original titles. If I want SMB I'll fire up the grey box, and I've personally played the CRAP out of my NES to the point where the games just aren't fun anymore. When you can beat games like Metroid, Bionic Commando and Blaster Master without thinking, the system kind of loses it's charm.

 

I think most of the people here are taking the wrong approach and concentrating too much on technical aspects of the system and forgetting just what the point of programming old Atari consoles is in the first place: making damn good games. I don't care so much about flashy graphics and programming techniques as much as I do about a game that I can play over and over. My own idea of the "perfect" game is one that I still want to play even after "game over" flashes across the screen.

 

I personally have a crapload of game ideas (every time I get one I jot it down in a notebook) and intend to implement them on the 7800 eventually, once I get adept at 6502 enough to do so. I realize this will take many years (and much frustration) to accomplish, but it's something I'm determined to do before I die.

 

Anyway, the bottom line is good games. I personally think games like Commando and Midnight Mutants are the pinnacle of what the 7800 can achieve technically, but what the system really lacks is good game ideas. A lot of NES games had unique game formulas and THAT's what made them special... look at Mega Man or Metroid or Zelda. I admit that side-scrolling platformers are probably the most lacking area of the 7800's library, but a sidescroller alone won't be enough to pique anyone's interest.

 

And finally:

 

If I ever hit the big ol' $250+ million lottery you can bet that I'd be the first person to plop down at least a few cool million to get some new game devving going onto push that sucker! Haha... that'd rock so hard and I'd love every penny being spent!

 

I rarely (if ever) play the lottery, but I've had similar thoughts. If I ever hit the jackpot, I would recruit a staff (probably using this message forum) to implement my game ideas. Call it laziness if you want, but hey, that's more new 7800 games for everyone else.