Remote H2H Atari product: why does it not exist?

[edburns]

Hello masters of Atari hardware, I have had a product idea that's been kicking around in my head for ages, and I wonder why no-one has built this product.  Perhaps you can offer some insight?  Scenario: a product that enables me to play h2h Atari games on real hardware on a real CRT where the other player doing the same thing, but over the Internet.  I was imagining a bluetooth device you could plug into the DE9 on the back of the Atari that syncs with an app to send/receive control inputs from the other player, who also has one of the same on the other side.  Video/audio is outside the scope of the product and is well handled by other solutions: Point the camera at the screen and use Teams/Zoom/Jitsi/Facetime/etc.  Why does this not exist?

 

Ed

[+Stephen Moss]

If by H2H you mean Head to Head and your reference the DE2 on the rear mean you are referring to the 2600 then I am no expert when it comes to the software side of the 2600 but I would think the reasons it does not exist are...

1) The Controller connectors are designed to be input not output (assuming the device plugged into the controller port is both sending and receiving player data which is the most logical approach).

2) The processing speed of the 2600 is probably too slow to have meaningful communications with Bluetooth dongle and even if it were not such communication may take up so much time that frame rate suffers badly.

3) The overhead in the driver for communications with the Bluetooth dongle would probably require a lot of cartridge space, thereby compromising what can be done for the game itself in the remaining space, resulting in poor games nobody would want to play.

4) It could not be used with existing games, only new games specifically written to communicate with the Bluetooth dongle would work with it, thereby limiting it usefulness to any potential customer.

 

I am not entirely sure what the camera is for, presumably to capture the position of whatever it is you are controlling for transmission by the app to the other console. That means you have to not only developed the Bluetooth dongle but capture the camera data, get it into your app, get the app analyse the image to identify player position, transmit said data and write a new game that supports the Bluetooth dongle otherwise you cannot test it all works.

That is a massive amount of work for what is likely to be one person and requires technical expertise in many fields (electronics, 2600 programming, App programming, Image processing).

 

When ever someone asks why has X not been done yet you can usually guarantee two things...

1) The person asking has no technical idea of what is actually involved to make it work, otherwise they would probably know why it has not been done.

2) It is not the great idea someone thinks it is. If it was a good idea then someone would either have already done it, be working on it or has already considered and rejected it as the time and expensive of development is prohibitive when compared to the size of the potential market and any income that could derived from it.

 

Edited July 9, 2020 by Stephen Moss

[edburns]

Thanks @Stephen Moss that explains it!

[+DrVenkman]

For what it’s worth, I believe the Atari Collection on Steam has some head-to-head options for 2600 titles. I have never tried them though.

[mr_me]

It's not going to happen with real hardware but there have been emulators with this feature.  A few years ago I used MameHub to do this with intellivision.  Would have worked with Atari 2600 for sure.  Not sure what the state of MameHub is today.  Retroarch offers online multiplayer, see if their atari 2600 emulator supports it.

 

Edit:

https://docs.libretro.com/library/stella/

Looks like the retroarch 2600 emulator supports Netplay.  So if the Mamehub servers aren't running the libretro servers should.

Edited July 9, 2020 by mr_me

[mr_me]

From a technical point of view it is certainly theoretically possible to do it with real hardware and existing cartridges.  One atari 2600 would be running with it's video/audio captured and streamed to the second player.  There would also be a network device plugged in to the player two controller port.  This device would take controller input from the remote player over the internet.  Atari controller ports can communicate input or output but only input is needed here.  In practical terms there would be far too much latency for the remote player both visually and with controller input for this to be acceptable. 

 

Emulators with online multiplayer play gets around this by each player having a local session of the game running.  Inputs for each player are local so response is pretty good.  A server communicates with each emulator and synchronizes the game state on each players emulator.

 

There is a third option and that is developing new Atari cartridge hardware and games specifically with online network play and network interfaces.   Two Atari 2600s and two copies of the networked cartridges and games would be needed.

 

What Atari 2600 games would you be interested in playing online with another player?

 

 

[flickertail]

First, I just want to say that I have no intention of doing the following, mostly because I don't see it being used very much - a lot of work for not very much return.

 

That said, if I was to do it, this is how I would do it.

 

First, the MrBoehm controller that David and I currently have in development could do most of what Mr_me has in mind. For more info on MrBoehm see my AtargiAge blog. We are very near to manufacturing our first physical prototypes.

 

Because MrBoehm runs off a RPi Zero W, it can connect to the internet. Though it is not my intention to enable this as it would bring with it a host of problems... though it's possible.

 

Anyway, through the MrBoehm/PI Zero two or more 2600's could communicate through WebRTC.

 

Of course, doing it this way, you'd have to give up using original Atari controllers, and default to using the USB/Bluetooth controllers that MrBoehm will support... for example... the modern VCS controllers that might someday get released.

 

You'd also need some kind of server-side management for this, which I suppose could be done through a website like AtariAge. I'm just speaking hypothetically. I'm just an AA user. I don't have any connection to AA.

 

Because AA already knows when you're logged in, and can send you forum/message updates in real time, I'm guessing an enterprising person could implement code on the AA server to handle WebRTC communication along with some sort of Web/Phone App that could put two or more players together for playtime.

 

But only if each user had a MrBoehm and the same Cart. So I don't see why it couldn't be done were someone be willing to put in the work. But such a person is not me.

 

That said, I see a big problem with Frame synchronization, and you would very likely see me destroying your player on my local Atari, but on your Atari the bullet missing your player completely. So there is that.

 

Also, there is the lag. My guess is the whole setup would suffer from lag issues.

 

Potentially, some of the synchronization and lag issues could be mitigated on Homebrews if they were specifically designed to pause between frames to wait for WebRTC input over the Pi/MrBoehm.

 

That's my two cents.

Edited July 16, 2020 by flickertail
Added a word.