Btcc22

Don't push your luck.
 

Don't push your luck.
 

Don't push your luck.
 

Don't push your luck.
 

Or no hope. Depends on the angle you view it from.

Awesome.

I never claimed that you said that it was simple, just that you were and still are oversimplifying what'd it take to make it work.

I meant to make this post a while back, so excuse the delay.
 
A.I. syncing isn't feasible unless some individual is willing to spend an inordinate amount of time on it.
 
Without wishing to state the obvious, the netcode just wasn't designed for it. Co-op, RTS and other A.I. heavy games often use an entirely different networking model (lock-step) to make it work. Lock-step is essentially where all clients advance a frame at a time and the simulation can only advance once every client has received updates from the rest of the clients.
 
The main benefit of a lock-step model is that assuming the A.I. is determinstic (behaves the same way given the same inputs), there's no need to use bandwidth syncing A.I between the players. It's also easier to ensure that the game remains consistent between all of the clients.
 
As we've all seen in videos of modders attempting co-op, A.I. quickly fall out of sync because each client is running its own simulation and the game-state on two clients is almost never identical, leading the A.I. to behave differently. It just doesn't work.
 
The obvious solution is to have the host simulate the A.I. and then pass their actions and other key bits of data to the clients. While it could work in theory, it'd still be a huge effort to extend the netcode and the A.I. system to support it and without crunching the numbers for the bandwidth, potentially pretty laggy (at best it'd warp like multiplayer and need leading).
 
Ultimately, I don't view it as the the holy grail of modding as others do but rather as something that'd be cool to try once and then forget about. The community has never been particularly strong when it comes to creating original single player content, with even the best effort failing to achieve parity with a stock map (in my opinion) and I doubt many players are going to stick around just to play the campaign over and over. I doubt it'd do anything for player numbers.
 
TL;DR: Technically possible but with compromises. Not worth the effort though.

I meant to make this post a while back, so excuse the delay.
 
A.I. syncing isn't feasible unless some individual is willing to spend an inordinate amount of time on it.
 
Without wishing to state the obvious, the netcode just wasn't designed for it. Co-op, RTS and other A.I. heavy games often use an entirely different networking model (lock-step) to make it work. Lock-step is essentially where all clients advance a frame at a time and the simulation can only advance once every client has received updates from the rest of the clients.
 
The main benefit of a lock-step model is that assuming the A.I. is determinstic (behaves the same way given the same inputs), there's no need to use bandwidth syncing A.I between the players. It's also easier to ensure that the game remains consistent between all of the clients.
 
As we've all seen in videos of modders attempting co-op, A.I. quickly fall out of sync because each client is running its own simulation and the game-state on two clients is almost never identical, leading the A.I. to behave differently. It just doesn't work.
 
The obvious solution is to have the host simulate the A.I. and then pass their actions and other key bits of data to the clients. While it could work in theory, it'd still be a huge effort to extend the netcode and the A.I. system to support it and without crunching the numbers for the bandwidth, potentially pretty laggy (at best it'd warp like multiplayer and need leading).
 
Ultimately, I don't view it as the the holy grail of modding as others do but rather as something that'd be cool to try once and then forget about. The community has never been particularly strong when it comes to creating original single player content, with even the best effort failing to achieve parity with a stock map (in my opinion) and I doubt many players are going to stick around just to play the campaign over and over. I doubt it'd do anything for player numbers.
 
TL;DR: Technically possible but with compromises. Not worth the effort though.

I meant to make this post a while back, so excuse the delay.
 
A.I. syncing isn't feasible unless some individual is willing to spend an inordinate amount of time on it.
 
Without wishing to state the obvious, the netcode just wasn't designed for it. Co-op, RTS and other A.I. heavy games often use an entirely different networking model (lock-step) to make it work. Lock-step is essentially where all clients advance a frame at a time and the simulation can only advance once every client has received updates from the rest of the clients.
 
The main benefit of a lock-step model is that assuming the A.I. is determinstic (behaves the same way given the same inputs), there's no need to use bandwidth syncing A.I between the players. It's also easier to ensure that the game remains consistent between all of the clients.
 
As we've all seen in videos of modders attempting co-op, A.I. quickly fall out of sync because each client is running its own simulation and the game-state on two clients is almost never identical, leading the A.I. to behave differently. It just doesn't work.
 
The obvious solution is to have the host simulate the A.I. and then pass their actions and other key bits of data to the clients. While it could work in theory, it'd still be a huge effort to extend the netcode and the A.I. system to support it and without crunching the numbers for the bandwidth, potentially pretty laggy (at best it'd warp like multiplayer and need leading).
 
Ultimately, I don't view it as the the holy grail of modding as others do but rather as something that'd be cool to try once and then forget about. The community has never been particularly strong when it comes to creating original single player content, with even the best effort failing to achieve parity with a stock map (in my opinion) and I doubt many players are going to stick around just to play the campaign over and over. I doubt it'd do anything for player numbers.
 
TL;DR: Technically possible but with compromises. Not worth the effort though.

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

The first. The second looks like a guy with a mohawk, grinning while he jabs objects into his eyes.
 

My point I was making was that it's overkill for being mainly used for in-home streaming. You could easily stream games to your phone with some app and an attachable controller. If you don't want to go the phone route, you could go with specs lower than a quad-core and 4gb of RAM because even that is really more than you need to simply stream the games.  
 

That's a lazy argument. Actually tell me which of those games would be more enjoyable. Find which ones could be played well on those specs, which ones could play well with the more than likely shittier version of the Steam controller, and which one of THOSE is probably not another indie game that's available on Android. Edit: So far, I've only found Super Meat Boy and Fez.

Go on...

I'll be honest and say that I remain entirely unconvinced but I wish you luck.

While I applaud your initiative, I do have to question whether you've bitten off more than you can chew. To produce an AAA quality Halo fan game with both single and multiplayer by late 2015 and get it on the Xbox One, with a tiny, inexperienced team that have never produced a game before is just, well, it's not going to happen. Then there's the prequel book you intended to write.
 
Unless Microsoft have changed direction in recent times, you'd be lucky to get the game finished before they slap you with a cease and desist let alone be invited to publish it on the Xbox One.
 
Other warning signs are that you mention it's been going since 2009, yet there's nothing to show. Another being that you suspended the development less than a week ago in favour of some other mysterious project and then randomly resumed it this week. 
 
 
I don't want to put your project down and while I'd love to be proven wrong, this is most likely going to fail as it stands.
 
Now for the constructive bit:
Narrow the scope of the project. Even an experienced team with millions at their disposal would balk at the idea of pulling something like this off. Don't try to develop the entire project at once. Pick a task (say, a single level) and get it working, it'll keep the team motivated. Set a proper website up and make the development open. People are more likely to contribute to something that they can see is making progress. Learn a useful skill that will allow you to make a worthwhile contribution to the project. I would never join a mod where the leader was just an ideas guy; they have no investment when things get tough (and they'll probably always be tough) and no way to keep the project going forwards even if they're on their own for periods of time. Post when you have something to show. Talk is cheap and making posts is fun but don't let it stand in the way of actually getting stuff done. Join other game development communities and ask for their advice too. Surround yourself with people that actually have experience and can help you keep things in check. The fanboys sure won't.

Nah.

You can find a bunch of sounds in this pack, if it's of any use:


 
https://mega.co.nz/#!vUVkxIhA!wYk0TyYWeJfk8Io15S_Tagq6Jb4-wz0uNiBcksMuh_4

You can find a bunch of sounds in this pack, if it's of any use:


 
https://mega.co.nz/#!vUVkxIhA!wYk0TyYWeJfk8Io15S_Tagq6Jb4-wz0uNiBcksMuh_4

Memory mapped files, my friend.

https://www.youtube.com/watch?v=gWJvptXsv4w