I'd think Demigod II is the way to go. Keep at least the current Demigods. Twelve would be a really good number, though, especially if they get more depth than now.

Still, I'd like to see the unique maps above. Get something the competition doesn't have in really awesome map design.

Honestly - connectivity is really the absolute number 1.  Even if you just ported dg1 over to a new engine or changed the network structure, as long as the connectivity is solid, we'll have a good community for quite some time.  I think we'd actually still be going very strong now if connectivity issues weren't there - bugs n all. 

No - I think you are reading correctly - it doesn't sound like there will be any connectivity improvements for the current iteration of Demigod.  Way too much money involved for little gain (or so that's how I read it).  That would come in demigod 2.  The question is - will there be a DG2 and will there be any new patches for Demigod as it stands. 

Sounds good to me, no lag please.

Sweet .... I'll expect something on my desk by say tomorrow afternoon then?

re: the_Monk

I'm not sure what the benefit of making the lobby client/server would really be, if all clients still have to connect to one another upon game start.  It'll just delay any connectivity problems until you launch.  That wouldn't affect any mid-game connection drops either.   The only thing this might be able to address is lobby drops (likely caused by all 7-9 other players in the game attempting to connect to the joining player at the same time), but GPG has that problem with SC2 as well, and they haven't been able to solve it despite being aware of and actively working on the problem for 6+ months.  However, for what it's worth, SC2 does not drop players mid-game.

I'm not trying to be overly negative here, but it's definitely impossible to change the game's core P2P model in a patch of any kind, so the changes that could be made are very limited.  Unless there's an overlooked flaw buried deep inside the netcode that's causing this, there's not much that can be done about it.

Ok, that makes sense.  Use a client-server layer to maintain connectivity and prevent or repair inappropriate drops-- presumably by pausing the sim until all P2P connections are re-established or the P2P-dropped client drops from the server as well?

I couldn't tell you whether or not that's doable in a patch, but it makes sense within the completely-P2P confines of the moho engine's sync/sim model.  At the very least, it could be a strong consideration in any sequel that also uses a P2P model.

Well... traditionally, games do get expansions.  I'd assume they'd end up going that route if its not f2p.  I can't imagine them refusing to create new dgs if dg2 is successful... I just don't think we need 75 of em.

Yup.  You've got the gist of my suggestions now.  Of course what I originally stated in reply #85 still presents the problem of additional network overhead cost whenever you encapsulate/tunnel traffic.  Sure you get the benefits I listed, but you also may make larger games unplayable (since at the end of the day it all has to come down to available bandwidth compliance as well).  I don't have any real numbers but by my mind you still wouldn't be able to play games much larger than 4v4 since the original P2P network model of DG doesn't seem to be very efficient to begin with.  Adding the fixes I'm talking about might fix the connectivity and sync issues but might increase overall latency even more if the available bandwidth isn't enough.

At the end of the day the real answer is to make sure any DG2 ships with more efficient netcode to being with.  I don't believe that the P2P model necessarily needs to be abandoned since it does provide for some benefits over the client-server model at the application layer for RTS games.  At the transport/session layer however the client-server model is much better at initiating and maintaining connectivity so again maybe a hybrid technology of P2P and client-server could be used in any future DG project.

the Monk

More demigods is fine if they're very, very thoroughly tested, well-balanced, and fill an actual niche in the game.  As others have mentioned, the 'more new stuff' model works well for F2P games, and it also can create temporarily-renewed interest in decent games that aren't balanced well (like, well, Demigod), but it can and usually does create as many problems as it solves.

A game that's well-balanced from the ground up with a good variety of different hero paradigms, especially one with 2-3 possible builds/roles per hero, does not constantly need new heroes to build hype and keep interest.  Nevermind that nobody cares about new stuff for a game that's known to be broken and un-fun to begin with, and when the playerbase has dwindled as much as Demigod's has, there's nobody left to care.  Stardock could have funded and released a DLC with 10 new Demigods 8-9 months after launch, and the game would have sold maybe a few thousand additional copies at best.  Probably none at all if the price of the game and the DLC combined broke $20-30.

Maintaining and tweaking emerging metagame balance, fixing bugs, not botching your game's launch, and fostering a vibrant competitive scene with strong matchmaking and ladder systems are much better and more proven ways of keeping a game's sales strong than random new crap.

I think the more well thought out Demigods the better honestly.

can't disagree.  Just say that a nice starting point is 10-12.