The Atari 2600+ is live for preorders!

[Jstick]

7 minutes ago, Atariboy said:

Most of these cores aren't cycle accurate. For instance only very recently has the Sega Genesis gotten a cycle accurate FPGA core with the release of Nuked MD for the MiSTer. Previous cores like those from Kevtris were implemented much like your typical Sega Genesis emulator rather than actually being based on replicating decapped original chips.

"Cycle accurate" doesn't imply decapped chips or netlists were used in the design of a core; software or hardware emulators can be completely cycle accurate by replicating only the observed behaviour of a system without internally implementing any of the original hardware design.

[Emehr]

Regarding the DIP switches, I think they're a neat idea. It would be super-trivial to 3D print a key where you can use a flat edge to "reset" the switches to zero and the keyed side to set the game. They can print little labels for the keys and put 'em on a key ring or something.

[roots.genoa]

2 hours ago, Atariboy said:

Most of these cores aren't cycle accurate. For instance only very recently has the Sega Genesis gotten a cycle accurate FPGA core with the release of Nuked MD for the MiSTer. Previous cores like those from Kevtris were implemented much like your typical Sega Genesis emulator rather than actually being based on replicating decapped original chips.

Uh, no, Kevtris always retro-engineers systems, he's absolutely against porting emulators.

[Recycled]

12 hours ago, Dr Karnov said:

Casuals who buy a gaming system on a whim and maybe a few carts to go with it, probably won't join a dedicated forum. You're probably safe 

Heck, I'm a casual that's been hanging out here for almost 2 decades. 😁

[Albert]

I tested a bunch of random homebrew games this afternoon on the 2600+.  Here's what I tested:

• Spies in the Night (8K - F8)
• Medieval Mayhem (32K - F4)
• Dungeon II (4K)
• Incoming! (4K)
• Dungeon (32K - F4)
• Reindeer Rescue (16K - F6)
• Thrust+ Platinum (16K - F6)
• Go Fish (NTSC, 8K - F8)
• Toyshop Trouble (8K - F8)
• A Roach in Space (32K - F4)
• Bee-Ball (4K)
• Seaweed Assault (32K - F4SC)

All worked perfectly except for Seaweed Assault, which did boot fine, but when you start a new game, the screen is full of tons of random seaweed, making the game impossible to play.  I assume this is some sort of RAM-related issue, as that game uses the extra SARA RAM, so I will investigate. 

 

 

(Forgot to put Seaweed Assault in this bin before I took the photo). 

 

 ..Al

[+jgkspsx]

So the old standard flashcarts for the Jag and the Virtual Boy only support a single game being loaded at a time. Modern Virtual Boy flashcarts modify this by actually having you select the game on an eink built in screen. Such an approach would presumably work on the 2600+… it’s just a matter of putting the data in places where the dumper can access it.

 

Was there ever a single game flash cart for the 2600? How hard would it be to make one, hardware inclined folks? It seems that with thousands of these things soon to be going into people’s homes there is a genuine reason for such a thing to exist. Could the PlusCart be configured to select its loaded game headlessly, @Al_Nafuur? Assuming the cart even gets power when the console is on…

[Thomas Jentzsch]

12 minutes ago, Albert said:

All worked perfectly except for Seaweed Assault, which did boot fine, but when you start a new game, the screen is full of tons of random seaweed, making the game impossible to play.  I assume this is some sort of RAM-related issue, as that game uses the extra SARA RAM, so I will investigate. 

You could try other 32K SC games, e.g. Aardvark. 

Edited September 18, 2023 by Thomas Jentzsch

[Emehr]

17 minutes ago, Albert said:

I tested a bunch of random homebrew games this afternoon on the 2600+.

 

Is there any noticeable input lag with this? My "go to" test game for NES emulators has always been Mike Tyson's Punch-Out and whether I can beat Soda Popinski or not. Not sure off the top of my head what an equivalent 2600 cart would be to test for fast-reflex gameplay. Maybe the BMX event in California Games? Or Dragster if you wanted to get nuts.

[Video]

Fpga perfection depends on the quality of design. Love them, but like emulation it has its own teething problems. Analog pocket is about perfect on gameboy stuff.

 

I started emulating atari back in the 90's when pcae came out, god that sucked. Later replaced with z26, a26, Stella and a few others. It was awesome seeing the evolution of this stuff and now it works damn near perfect.

 

Anyhow, if no interest is shown, it won't improve, so I don't get "I won't buy that" from people. I am hoping this will be good, or at least good enough, but more than anything I'm getting it to see it continue.

 

Its already a step above return 77, in that it has paddle support (I think retron does now, but didn't start out that way) I too wonder if the plus supports driving controls, the simultaneous left plus right thing didn't work on retron.

[Albert]

10 minutes ago, Emehr said:

Is there any noticeable input lag with this? My "go to" test game for NES emulators has always been Mike Tyson's Punch-Out and whether I can beat Soda Popinski or not. Not sure off the top of my head what an equivalent 2600 cart would be to test for fast-reflex gameplay. Maybe the BMX event in California Games? Or Dragster if you wanted to get nuts.

Not that I've experienced, and I'm pretty sensitive to that.  Initially when I hooked the 2600+ to my LG OLED television, there was noticeable lag.  I had to put the TV into a predefined "Game" mode, which presumably removes most of the image processing (which is kind of insane on some of these TVs) and that fixed it. 

 

 ..Al

[atari2atari]

I got excited and made a pre-order a few weeks ago, but I won't lie, I'm having regret reading that 'Pitfall II' might not work -- that game is like HALO was to the XBOX for me.  It's a constant, and looking across the room, I see stacked on my 2600 right now Pitfall, Pitfall II, Donkey Kong, and Mario Bros. (was having a Nintendo moment this weekend).

 

Has anyone been able to test Pitfall II ?  Confirmed not working?

[Albert]

Just now, atari2atari said:

Has anyone been able to test Pitfall II ?  Confirmed not working?

I have a copy of Pitfall II, but it doesn't work on a real 2600, so I cannot confirm one way or another yet.  However, I doubt it's going to work right off the bat due to the nature of the additional DPC chip onboard the cartridge.  That's not to say it won't work at some point, though. 

 

 ..Al

[Albert]

48 minutes ago, Thomas Jentzsch said:

You could try other 32K SC games, e.g. Aardvark. 

Yes, I will, I plan on trying all the homebrew games I can get my hands on (which, of course, includes anything I've personally been involved with).  

 

 ..Al

[John Stamos Mullet]

3 hours ago, Jstick said:

"Cycle accurate" doesn't imply decapped chips or netlists were used in the design of a core; software or hardware emulators can be completely cycle accurate by replicating only the observed behaviour of a system without internally implementing any of the original hardware design.

So basically they aren’t necessarily any better than a regular software emulator, except in the mind of the dev who “observed” a system and then wrote a core based on what they observed? 
 

lol. 

[+Al_Nafuur]

1 hour ago, jgkspsx said:

Was there ever a single game flash cart for the 2600? How hard would it be to make one, hardware inclined folks? It seems that with thousands of these things soon to be going into people’s homes there is a genuine reason for such a thing to exist. Could the PlusCart be configured to select its loaded game headlessly, @Al_Nafuur? Assuming the cart even gets power when the console is on…

The PlusCart (and AFAIK the UnoCart) cannot be used headless, only auto start is possible with the PlusCart UCA firmware. Technically headless can be done, there was just no need until now.

 

 

[Atariboy]

4 hours ago, roots.genoa said:

Uh, no, Kevtris always retro-engineers systems, he's absolutely against porting emulators.

I didn't say that he was porting emulators. Perhaps I wasn't as careful with the words I chose as I should've been, since rereading my post just now I can see why you thought I was suggesting that.

 

It's my understanding that the average FPGA core as well as the average software emulator aren't replicating things down to the transistor level. Instead of replicating each chip on 1:1 basis, they're more concerned with the overall end result that they're getting and how close it approximates real hardware, which isn't cycle accuracy. It's also why such things like Analogue products often fail tests with diagnostics carts for another example. Cycle accuracy would mean that they wouldn't since it's a perfect replica.

 

That method can still deliver amazing results when it comes to actually playing a game, as we've seen such as with the Mega Sg for instance, but also as we've seen it leaves open room for inaccuracies like we've routinely experienced in Analogue products or why fixing one thing often ends up breaking something else. But by decapping chips and replicating all the intricacies revealed by the process, it theoretically means it's truly a 1:1 replica no matter if it's a software emulator or a FPGA core.

 

That said, some successful FPGA cores have been born in this manner (creating an emulator and then porting it). I believe that the very promising looking N64 core for the MiSTer for instance started with the author coding a N64 emulator first. I don't know the rights and wrongs with that method (or why anyone would be against it in the first place), but the results and rapid progress being seen with that particular core suggest it was at least the right direction for that individual.

Edited September 19, 2023 by Atariboy

[kevtris]

9 hours ago, Atariboy said:

Most of these cores aren't cycle accurate. For instance only very recently has the Sega Genesis gotten a cycle accurate FPGA core with the release of Nuked MD for the MiSTer. Previous cores like those from Kevtris were implemented much like your typical Sega Genesis emulator rather than actually being based on replicating decapped original chips.

this isn't true.  my 68K is based on the decap, and I used extensive logic analyzer and chip testing to reproduce functionality of the rest of the system.  my 2600 and 7800 cores are based on the chip schematics.

[Atariboy]

1 hour ago, kevtris said:

this isn't true.  my 68K is based on the decap, and I used extensive logic analyzer and chip testing to reproduce functionality of the rest of the system.  my 2600 and 7800 cores are based on the chip schematics.

You should know, obviously.

 

I'll pass the buck onto someone else though for misinformation, with me guilty of just spreading it thinking that they knew what they were talking about. One of the YouTube FPGA news channels went out of their way to characterize Nuked MD as the first cycle accurate Sega Genesis FPGA core, singling out your Mega Sg core as well as the old MiSTer core as not being cycle accurate.

[roots.genoa]

2 hours ago, Atariboy said:

One of the YouTube FPGA news channels went out of their way to characterize Nuked MD as the first cycle accurate Sega Genesis FPGA core, singling out your Mega Sg core as well as the old MiSTer core as not being cycle accurate.

From what I've gathered, there are several reasons a core based on a decap could still feature inaccuracies, one being that the RAM can not always been handled by the FPGA itself (except for very basic hardware). Not everything is recreated in the FPGA anyway, especially the I/O, so that probably introduces various difficulties I guess.

Regarding claims made by YouTube channels and stuff, I would be very cautious since there's a lot of friction in the FPGA scene to begin with, so everyone likes to denigrate other initiatives unfortunately. 😩

[Atariboy]

40 minutes ago, roots.genoa said:

From what I've gathered, there are several reasons a core based on a decap could still feature inaccuracies, one being that the RAM can not always been handled by the FPGA itself (except for very basic hardware).

Plus there's always room for some mistakes. It always happens for instance on replica circuit boards for classic 8-bit computers. While not nearly as complicated as decapping and replicating something like a processor or sound chip, the first attempt sent off to a company like PCB Way inevitably always has some mistakes and needs a trace cut here, a bodge wire somewhere else, etc.

 

So while in theory it's a great sounding idea to avoid inaccuracies, I can only imagine the challenge involved in creating a cycle accurate replica of a classic game system that ramps up the complexity exponentially. So much opportunity for small mistakes to slip in.

 

I simply did a poor job of expressing myself with my initial post. I wasn't out to denigrate any project (I've spent countless happy hours with my Analogue systems, Kevtris). While I'm largely clueless in the challenges, techniques, and such involved in such projects, I was just trying to explain to Dr Karnov that FPGA hardware simulation isn't automatically better than software emulation like he seemed to think. It offers advantages like a live cartridge bus, but also comes with disadvantages.

 

There's still plenty of room for human error, incomplete technical information, and such. It's all up to the capabilities of the author or team behind a project. So just as we have top notch emulators like Stella along with poor emulators, not every FPGA core is created equal. I'd take a good software emulator over a lackluster FPGA implementation any day and vice versa.

Edited September 19, 2023 by Atariboy

[roots.genoa]

Yes, the real advantage of FPGA emulation is the lack of lag, either input lag and display lag, compared to software emulation, since the synchronicity between virtual components is usually more accurate.

[jhd]

On 9/7/2023 at 11:14 AM, AtariLeaf said:

https://www.amazon.ca/Plaion-Atari-2600-Hardware/dp/B0CG7LMFKY/ref=mp_s_a_1_1?keywords=atari+2600%2B&qid=1694095412&sr=8-1

 

And I finally placed my pre-order yesterday. Happily I was able to pay with accumulated rewards points so it was essentially "free" to me.

 

The total cost with tax was Cdn$172.49 (US$128.67).

[hizzy]

I ordered an Atari + Amazon.ca but I cancelled when I read that Atari made no donation to Stella. I'm now going to order a Raspberry pi kit, install Stella and make a donation to the Stella team. I've never done anything with a Pi but I don't think it will be worse than the Atari +.

[donjn]

The Gamestation Pro looks very nice and does some things the 2600+ does not do, at least so far, for less money.

I think having an SD card slot is a game changer, and the ability to play 2600, 7800 AND 5200.

 

It's going down tot he wire regarding which one I purchase. I am still torn.

 

[donjn]

17 minutes ago, hizzy said:

I ordered an Atari + Amazon.ca but I cancelled when I read that Atari made no donation to Stella. I'm now going to order a Raspberry pi kit, install Stella and make a donation to the Stella team. I've never done anything with a Pi but I don't think it will be worse than the Atari +.

What do you mean by that? Donation?