2600 Homebrew Quality Dropping:Proposed Solution

[+SpiceWare]

Battery backed RAM has been around for some time now. I wonder what the challenges would be making a Starpath Supercharger compatible cart - in that the game resides in 6k RAM. Except, if it was battery backed you could have self modifying code and sandbox style gameplay.

 

Harmony and Melody support supercharger games. It's also possible for Melody games to save information in the cart, Chetiry uses that ability to save high scores.

[Albert]

Harmony and Melody support supercharger games. It's also possible for Melody games to save information in the cart, Chetiry uses that ability to save high scores.

 

The Aria board, which is less sophisticated than the Melody, can also be used to save data to the cart. Star Castle Arcade uses the Aria and saves high scores as well as your current graphics setting (NTSC/PAL/B&W color set).

 

..Al

[+SpiceWare]

I always thought if we were going to rely on hardware assist anyway then go full Hellbound Hellraiser for the NES. Use the display kernel like a framebuffer and just use the 2600 for i/o. We're already using a separate processor in the carts for these boards anyway.

That's pretty much what I've been doing in my games. The 6507 reads the state of the Atari, sends it over to the ARM processor. The ARM then creates the data to be displayed and hands control back to the 6507. The 6507 then draws the screen using the pre-calculated data. Main difference between BUS and DPC+ is we've made it even faster to access the pre-calculated data so the display kernel can modify the display even more than before. As an example of how much more, the 96 pixel kernel in that RPG demo can show 12 colors per scanline, the DPC+ 96 pixel kernel can only show 2.

 

 

I guess a similar example would be that new Vectrex board:

http://vectrex32.com/

Thats slick!

[+SpiceWare]

It's really, REALLY tough to find anyone outside of AtariAge making 256k boards or boards with SARA RAM. If you want RAM other vendors scavenge used parts which are getting pretty old. I'd like to see more options and I always make the pitch to those getting into the cart making business. Sadly, I've yet to get a positive response.

Seen the cartridge ZackAttack's designing?

[+SpiceWare]

Melody cartridges have DPC+ chip that help with processing data. Starpath bankswitch between the 3 2KB bank of RAM.

There's actually no such thing as a DPC+ chip. Melody carts have an ARM processor in them which is just fast enough to emulate the various cartridge bankswitching formats, including the DPC chip in Pitfall 2. Pitfall 2 only uses 8K of ROM and 2K of display data, whereas the Harmony/Melody can handle 32K of ROM and 8K of RAM. So I asked if it'd be possible to expand DPC to utilize the additional resources and we came up with DPC+. You may find this blog entry interesting.

 

 

One thing I don't know if the lower memory of the atari 2600 mapped at $0100-0FFF is could be use for something like adding, soldering, RAM into that region or is it unusable. The cartridge I know is mapped at $1000-1fff.

Interesting question - TIA and RIOT are mapped multiple times throughout memory, known as mirroring, which would probably make it difficult to add something else in there.

[+KaeruYojimbo]

True about the second thing said, but adding more memory DOES change the way games are made. Think about it. The game industry would have been dead, or alive and dull without memory updates. New genres have spawned because of memory upgrades.

 

You could just as easily argue that the games industry has become bloated and uninspired. How many big releases these days are just the next iteration of a franchise with a few updates? Many of the most exciting and interesting new releases are from smaller developers who are putting interesting ideas ahead of making huge games. And there are still plenty of very good 4K homebrew 2600 games being written. Sometimes forcing yourself to fit into certain limitations can lead to amazing discoveries.

 

I'm not saying don't push the 2600 technology-wise. I'm saying there are still plenty of ideas out there. Your argument seems to be that the only way for 2600 development to survive is to make bigger games, but there are creative and clever people who will continue making good games, whatever the size.

Edited November 3, 2016 by KaeruYojimbo

[enthusi]

I actually propose the complete opposite.

As mentioned lowering the bar to start on a system naturally brings on more content. Since the minimal effort is much lower, there of course a increasingly large fraction of lower quality.

Still just a fraction, though.

Games that require DPC or even more sophisticated features are (in my view) no longer Atari2600 games, but Harmony-Games (which as such does not mean that they are bad).

And please dont bring up Pitfall here

Allowing for even more capabilities at an even lower entry level will result in two things:

- people caring even less about optimization (of either gameplay nor code)

- people being confused about what the Atari 2600 actually is/was ("HERO was such a lame game! Check out this fancy new blabla for the same system")

- LESS creativity since you can invest all your drive for exploration in the sheer amount of data that you now HAVE and need to handle (large gfx, levels, audio).

 

I like to bring up the C64 for comparison. The scene that limited itself very strictly to the plain vanilla system of ~ 1983. Audio differs slightly but hardware capabilites are 100% identical.

Ram Expansions and alike basically play no role at all.

 

Yet MANY games (and demos) are produced for this system.

Among the most successful sources of new games is a competition of 16KB (C64 has 64KB ram) Games.

Restrictions are what trigger creativity and innovation.

Maxing out a system, not extending it at first signs of challenge.

 

You can code games in python on PC, you know.

They will suck badly unless you have a super great game design (which is veeery hard to do).

We all know games on PC that LOOK and SOUND great but feel pretty lousy.

The less restrictions you have technically the more you can hide a bad game play or even design.

 

This is of course just my own opinion.

To each his own and I do not judge people for what they enjoy in their spare time.

[Trebor]

...Games that require DPC or even more sophisticated features are (in my view) no longer Atari2600 games, but Harmony-Games (which as such does not mean that they are bad)...

Applying the same principle, that added hardware chips/features not a part of the main/base console makes a game no longer part of a particular system's platform, eliminates a considerable number of games from several different platforms that are considered games of that specific platform.

 

As examples, no-one states it's not a 'SNES game', but a 'DSP-1' (Super Mario Kart), 'Cx4' (Mega Man X2), or/and 'Super FX' (Super Mario World 2: Yoshi's Island), game. Nor do people state it's not a 'Sega Genesis' game, it's a 'Virtua Processor' (Virtua Racing), game. Claiming MMC3 hardware based games (SMB2, SMB3), or MMC5 hardware based games (Castlevania III), are no longer viewed as NES games would seem odd as well.

...And please dont bring up Pitfall here ...

I won't...But the sequel with the DPC chip is still considered an Atari 2600 game.

 

Allowing for even more capabilities at an even lower entry level will result in two things:

- people caring even less about optimization (of either gameplay nor code)

- people being confused about what the Atari 2600 actually is/was ("HERO was such a lame game! Check out this fancy new blabla for the same system")

- LESS creativity since you can invest all your drive for exploration in the sheer amount of data that you now HAVE and need to handle (large gfx, levels, audio).

Checking some of the aforementioned games across the various platforms, many are considered some of the best titles to experience and play on the system.

 

...This is of course just my own opinion...

Respected; and agree to disagree.

[Thomas Jentzsch]

There is a difference between hardware that was available back then and hardware which only became affordable now.

[enthusi]

Also the mentioned (AAA) Titles are professionally made. Such is Pitfall.

Not homebrew. I find it amazing i.e. what can be reached with say bus-stuffing.

This is certainly challenging and creative.

Exploiting something new or for the first time is very interessting per se.

 

Making use of maybe 20% of its capabilities to 'improve' all future atari 2600 games (exaggerating here) is something I consider boring and lazy.

(I am well aware that this is no rule to the world but my personal view only)

[alex_79]

Everyone has his own personal idea about what's the limit between an authentic 2600 game and what should be considered a different platform, and I don't see the point in trying to estabilish a "standard". Anyone would just keep programming/playing the games he likes and just ignore such a standard.

That said, here is my personal idea (and I reserve the right to change it anytime without notice ):

Any game that you can plug in an unmodded Atari 2600 console, and that uses the TIA output is an Atari 2600 game.

In consoles using mass storage media (tapes, floppies, CDs, etc) you have a full working microprocessor based system with it's own memory and I/O. The software is read from the media and copied into the system memory to be executed. Hardware add-ons are well distinct from software in this case. The storage media only contains the software to be executed in the machine.

On the other hand, a 2600 without a cartridge plugged in is not a complete system. It doesn't work at all, actually. The cartridge is by definition "hardware" and by plugging it you complete the system architecture by connecting components (rom, ram, i/o ports, coprocessors and whatever) to the system bus.

A 2600 with Combat cart is different hardware-wise than a 2600 with Pac-Man cart, which in turn is different than a 2600 with a Supercharger module or a 2600 with Harmony Cart. What's a "vanilla" 2600 system here?

It's just up to the programmer to decide what are the goals and limitations.

You can take advantage of modern hardware like it would have happened if the 2600 was still produced today, or only use hardware which was avialable while the console was in production, or only use what was available in 1977 when the VCS was launched.

The same can be applied to the development tools.
You can use modern cross-assemblers, emulators and debuggers or decide to only use dev systems of th e'70s and '80s.

 

All these approaches are completely arbitrary. And I appreciate them all!

 

 

As for the topic starter, I think that producing games that require a separate hardware expansion is just a very bad idea. The effect is to further reduce an already niche market. All the hardware must be included in the cartridge itself, like it has been done so far. Add-ons (like the AtariVox or custom controllers) are ok if they're optional (that is, the game works also without them albeit with reduced features).

Also I don't think programmers will ever run out of ideas. More probably we'll run out of programmers interested in the console at some point...

[Mr SQL]

Everyone has his own personal idea about what's the limit between an authentic 2600 game and what should be considered a different platform, and I don't see the point in trying to estabilish a "standard". Anyone would just keep programming/playing the games he likes and just ignore such a standard.

 

That said, here is my personal idea (and I reserve the right to change it anytime without notice ):

 

Any game that you can plug in an unmodded Atari 2600 console, and that uses the TIA output is an Atari 2600 game.

 

In consoles using mass storage media (tapes, floppies, CDs, etc) you have a full working microprocessor based system with it's own memory and I/O. The software is read from the media and copied into the system memory to be executed. Hardware add-ons are well distinct from software in this case. The storage media only contains the software to be executed in the machine.

 

On the other hand, a 2600 without a cartridge plugged in is not a complete system. It doesn't work at all, actually. The cartridge is by definition "hardware" and by plugging it you complete the system architecture by connecting components (rom, ram, i/o ports, coprocessors and whatever) to the system bus.

 

A 2600 with Combat cart is different hardware-wise than a 2600 with Pac-Man cart, which in turn is different than a 2600 with a Supercharger module or a 2600 with Harmony Cart. What's a "vanilla" 2600 system here?

 

 

A vanilla system is one that uses legacy Atari technology like the SuperCharger, Sara or CBS RAM chips!

 

There are some fantastic games where the gameloop is written in C and runs on a 32-bit processor - these are clearly 32-bit games written in C as opposed to 8-bit gameloops running at 1 MHZ on the Atari; I agree with enthusi that it is fun to push the legacy hardware using legacy equipment and methodology and won't even use the debugger in Stella for that reason; I also refuse to look at Rubics cube solutions, and it's still a (as yet unsolved) challenging puzzle for me.

 

The modern games are awesome too and lots of fun to play, but only the bB DPC+ games can qualify as retro Atari games like Pitfall II because their gameloops still runs on the 6502 and the ARM is used as a coprocessor. Games with 32-bit gameloops that run on the ARM are awesome fun but clearly a modern category to programmers at least - imo having another video out port on the cart like we see on some enhanced Nintendo carts with additional video processors are still more retro than the actual game code being 32-bit.

 

Legacy technology games, weather in asm or BASIC don't need a distinguishing label but I think it's fairer to developers writing challenging legacy code to label modern 32-bit ARM gameloop games as absolutely not anything like Pitfall II instead of making believe they are just using a coprocessor - when the gameloop runs on the ARM, the 6502 becomes the coprocessor.

[iesposta]

It's really, REALLY tough to find anyone outside of AtariAge making 256k boards or boards with SARA RAM. If you want RAM other vendors scavenge used parts which are getting pretty old. I'd like to see more options and I always make the pitch to those getting into the cart making business. Sadly, I've yet to get a positive response.

 

So what's wrong with just using AtariAge's hardware?

AtariAge can put your game made with any bankswitch type and any amount of extra RAM in a cartridge at around $20 and up.

If it isn't shovelware, it will be accepted to sell in the store if you wish.

[+SpiceWare]

You can take advantage of modern hardware like it would have happened if the 2600 was still produced today, or only use hardware which was avialable while the console was in production, or only use what was available in 1977 when the VCS was launched.

 

Yep - what we're doing with Bus Stuffing is not any different than what they were doing back in 1982 with the Graduate.

 

I find it puzzling that so many seem to be of the opinion "software hombrew good, hardware homebrew bad". Oh well.

[Thomas Jentzsch]

IMO its more like software homebrew is considered very different to hardware homebrew.

 

BTW: Has any software from the Graduate ever appeared? The 128 bytes program executed in RAM sounds interesting.

[+Nathan Strum]

Yep - what we're doing with Bus Stuffing is not any different than what they were doing back in 1982 with the Graduate.

 

I find it puzzling that so many seem to be of the opinion "software hombrew good, hardware homebrew bad". Oh well.

 

No, no. It's "software homebrew good, fire bad".

 

Personally, I find it to be really exciting seeing new games that push the 2600 beyond anything anyone imagined it could do. But it depends on the game.

 

I felt that way when I bought the SuperCharger back-in-the-day. Not all of the games for it were good, but a few of them were standouts, and it really felt like they were giving the 2600 some special, new capabilities. The best of the lot were Phaser Patrol, Suicide Mission, and Escape from the Mindmaster (I hadn't bought Frogger, but I'd have included that, too). But there were some duds in there like Communist Mutants from Space, Killer Satellites, and Fireball (really? We needed another version of Breakout?). The potential was there, and reading about what the SuperCharger was supposed to bring to the system was exciting. It ultimately fell short, because the software just wasn't there.

 

CBS made sure to brand some of their games with the RAM Plus moniker, so you knew you were getting something more advanced than a typical 2600 game. Again, those games weren't necessarily better, although I thought Tunnel Runner was pretty amazing for its time. Omega Race, on the other hand, was terrible (I was a big fan of the arcade game, and I thought CBS really missed the mark).

 

It was pretty widely touted that Pitfall II used a special chip when it was being sold. Activision already had a reputation for quality, and now you had an even more advanced Activision game that was a sequel to their most popular title. How could you go wrong? (Well... the fact that the game became an agonizing grind, for one. But it was still impressive for how vast it was.)

 

Advancing game technologies were all part of the marketing at the time. Even if the public didn't really understand why extra RAM or the DPC chip enhanced a game, they knew it was still more advanced than what had come before. This was the video game industry of the time. For that matter, it still is. Just on a different scale.

 

None of those advancements put me off from buying those 2600 games in the early 80's because "they didn't have that way back in '77" . The 2600 library was always advancing, always improving. I remember reading about Asteroids using more ROM and it somehow being a big deal (I had no idea what it meant at the time). It was simply the progression of technology. The programmers needed to make more advanced games, and they figured out ways to do it. Everything built upon itself, and it had to because the 2600 lasted so much longer than anyone expected (if you want to find out why, there's some video you can watch about that).

 

When DPC+ was worked out for the Harmony cart, I had mixed feelings about it (as I'd had about the canceled Chimera which preceded it). I thought, "Would the games still feel like 2600 games? If I wanted to play a game that looked or felt like something from a different system, I'd buy that system." But the games are still 2600 games. They still look and feel like 2600 games. They have those unique characteristics - the sounds, the graphics, the colors, that blocky playfield. It doesn't fundamentally change how the games feel. DPC+ (or Bus Stuffing) basically gives you extra time to throw more instructions at the 2600, so the 2600 can do a little bit more of what it already does. When I get to plug Scramble into my 2600, I'm going to think, "Awesome! I get to play Scramble on my 2600!" Not, "Gee... I wish they hadn't 'cheated' to make this."

 

I think there's a place for every approach to 2600 programming. 2k or 4k mini-games can be incredibly fun. Seemo (for one) excels at making very fun, unique, and distinctly 2600 games. Omicron, Lead and Sync, all started out as smaller mini-games. And he's programmed a bunch of others which have yet to make it onto a cart. He uses the unique characteristics of the 2600 to his advantage. He designs original games that leverage what the system was designed to do.

 

That said, one of the projects I'm most proud of being a part of was Stay Frosty 2. That's a huge game that kept improving and growing year after year. I don't know how much of what was there in the end would have been left out, had Darrell not moved it over to using DPC+. But it certainly wouldn't have been the game it is now. He used the tools necessary to make the game the way he wanted to make it. I think that's the bottom line: What game do you want to make, and what do you need to get there? From my standpoint, if the game ships on a cartridge that you can plug into any 2600, it's a 2600 game. Regardless of how it was programmed, what year it was made, or what's in the cart.

 

Ultimately, what matters to me is the end result. You can throw a ton of extra RAM or processing power at a game, but if it's a poor idea or badly executed, it's going to be a bad game. You can take the simplest of ideas, and if it's fun and executed well, it can be a great game. However you program it, for whatever hardware you use, if you want to make a truly excellent game, you have to be willing to put the time into it to test it, debug it, polish it, and listen to feedback from the community (that's not to say you have to acquiesce to everyone's ideas - it still has to be your game). If you just want to make a game that's "good enough", that's probably all it will ever be.

[+SpiceWare]

No, no. It's "software homebrew good, fire bad".

- I originally had that wrapped in <caveman> </caveman> tags

[Thomas Jentzsch]

Ultimately, what matters to me is the end result. You can throw a ton of extra RAM or processing power at a game, but if it's a poor idea or badly executed, it's going to be a bad game. You can take the simplest of ideas, and if it's fun and executed well, it can be a great game. However you program it, for whatever hardware you use, if you want to make a truly excellent game, you have to be willing to put the time into it to test it, debug it, polish it, and listen to feedback from the community (that's not to say you have to acquiesce to everyone's ideas - it still has to be your game). If you just want to make a game that's "good enough", that's probably all it will ever be.

/signed

 

Everything else is a matter of role, perspective and individual preferences.

[Keatah]

Oh, sorry to disappoint you. Yes, just like we did with DPC+ back in 2010, we'll be releasing the Bus Stuffing driver along with documentation and examples on how to use it. Stella's also being updated to support it, we'll work with stephena to get that released at the same time.

 

Remarkable!

[+SpiceWare]

 

Remarkable!

 

 

- that's due to experience. While Stella supported DPC+, it didn't support ARM code for most of the time that I worked on Stay Frosty 2. That was a major hassle at times.

 

While watching the Stella at 20 tapes one of things mentioned was using a VCR to record the Atari, then freeze-framing it to track down bugs. I basically did the same thing to figure out what these dashes were that I'd occasionally see on in Stay Frosty 2.

 

 

 

And again to figure out the source of screen jitter:

[+thegoldenband]

I think pushing the VCS hardware (and the stuff seen in this thread) is terrific! I'm all for skilled programmers and hardware designers upping the ante in every way possible.

 

The only thing that irritates me is when I see half-assed efforts that use extra hardware (especially an on-cart CPU) to achieve an effect that a skilled programmer can achieve routinely without an additional CPU. That annoys my thrifty side -- it seems like a waste of resources and an unnecessary additional expense -- and also bothers me in the same way it bothers me when a composer writes a piece for "violin" but doesn't actually understand how the instrument is played.

[Supergun]

I'm probably alone on this, and I've said it before, but I rarely play or buy homebrew software for any system.

 

In fact, the original ideas are the ones that interest me the least. The ones that actually do interest me are the tech challenges and dream games.

 

For example: "PAC-Man 4K" drew my interest big time. As did those last two 32K Donkey Kongs.

 

I would love to see someone now do the reverse of the above. Give us a better DK, limiting yourself to the original, and give us the most amazing arcade PacMan possible.

 

I also like the dream games, like say a Donkey Kong 3, or even the recent Super Mario Bros.

 

Forgot to mention:

My personal dream game:

2600 R-Type

Edited November 4, 2016 by Supergun

[gliptitude]

Wow what an awesome and dense conversation. I can't tell where it's going in relation to where it started, or what actually matters.

 

I think i disagee with some of the original post, that memory and hardware upgrades are necessary. .. I definitely do DISAGREE that running out of ideas is a legitimate or immediate reality or that hardware is the reason for this. .. I think there is a laziness that coincides with an overblown conception of what actually constitutes a new idea. Most substantial "new ideas" derive predominantly from existing ideas and content. You don't have to abandon the past or fundamentally change the ingredients in order to make progress. .. In fact doing so usually results in a shallow and dismissible result.

 

But...

 

New memory inspires greater games. An Atari 2600 programmer from 1978 would have never even THOUGHT to make an open world game like the 2016 game Ark:Survival Evolved.

I totally agree that new capabilities provide for a free flowing inspired mode of creation. In a way that's a crutch because it is totally unnecessary, and it really betrays the fact that there is so much more that could be done with what already exists. But whatever enables somebody to make something is productive. .. I think there is a power trip involved with breaking new ground, which empowers the author(s) and provides some creative safety and confidence. But it's all just a contrivance.

[Keatah]

Of all the VCS homebrews out there, I have about 10-12 on my frequent playlist.

[enthusi]

For all systems that became home of homebrews, there exists alot of crap, some okish games and some true gems.

I seem to note that systems with a high dev-entry-threshold naturally dont have super-lame entries, because whoever managed to set up the tool-chain etc. did that for a purpose and according to a plan.

But that is not necessarily the solution

I wonder if there is a working ranking/voting system for Atari2600 homebrew?

For C64 you get a pretty good idea when checking

http://csdb.dk/toplist.php?type=release&subtype=%2811%29&submit=Check%A0chart

Of course the CSDb site is not fully representative and you notice that big names have a big impact but I am not aware of something similar for the 2600 homebrew.

The store-reviews hardly count The titles for sale are somewhat pre-selected of course and who's review a game bad?