"However while it may be free to develop and release a non-commercial game, should you wish to profit from your hard work the current UDK license states that 25% of all revenue earnt beyond $5,000 will be paid to Epic as your licensing cost. I consider this a small price to pay with no risk for developers who now have access to proven technology."
This is bad for a couple of reasons. a) I don't think the devs want to be giving away money that they spent on developing their own game, and equally this will make it more expensive for the players who would have to offset the extra cost (I believe the devs have said when(
if) this game comes to fruition the cost to the player will be mainly for running/maintenance costs) b) This will also tie them down to licensing/legal mushy jargon.
It is free "to develop and release a non-commercial game" and once you go over $5000 in profit (how they will define this I don't know) 25% goes to them. What you get with it is a toolkit, but no access to the source and a mix of documentation. You can weigh up whether that's a better deal, or whether you want the engine with an open source for license holders and an active developer who supports and listens to his customers)
"I forgot the name.....ARMA's Game engine would be good....(i know its probably impossible)...."
Arma2's game engine is very expensive, and ARMA's game engine is fairly restrictive, still has bugs, and is getting dated.
"half life 2 is on source. half life 2 has the crossbow. the crossbow bolts take ages to hit target > i call BS.
also with the source engine we could see quite a lot of realistic gore.. anyone tried headshooting the zombs in left4dead2? rather interesting. explosions cause loss of limbs and shotties tear pieces off, i guess we couldnt be able to have 4km maps but tbh i wont miss them. it is by no means perfect, but even with deviation problems it would be better than bf2"
Yes, the crossbow does have bullet drop and does take time to hit the target, however it moves very slowly and would probably be quite a hack to get it implemented with 'bolts' moving at 800m/s with little to no network troubles. As for realistic gore, that can be done with particles very easily and the blood spatter can be achieved through decals or neat texturing. Tearing limbs off is fairly pointless, wastes network bandwidth and isn't exactly realistic (you're gonna need a big bullet to do that kind of damage, equally the griefers would love it as an excuse to blow up friendlies). The maps would be very limited in size, and as we all know size
does matter

(also, what then would set the game apart from every other fps?)
Although it's no longer the "I want feature x in the new game", it's turned into "This engine is the best because it has x in it". Commercial engines will tie them into licensing, and many of the engines are getting dated or no longer being developed. Stepping up from a mod to a full game is a brave and ambitious venture, and the shire amount of work that goes into making games is vast. So having a lot of documentation, and a active developer who you can ask questions in an invaluable resource.
People please think before you suggest yet another game engine that you think is the next best thing because of a cool feature, or because x game is good on it and therefore it's engine must be amazing. With an engine you do want features, but features that are relevant to the developers and not just relevant to the customers. "Gore" doesn't count as a feature, because almost any engine can do it in one form or another and "X graphics" does not count as a feature especially since when people say good graphics they don't really understand what's good about it... Good graphics is not paralax mapping or real time global illumination, these things can be used and still make a really ugly game, the graphical features give the potential to the developer and it's up to the developers use of that technology to make truly breath taking scenes. PR has shown that the 4 year old bf2 engine can have amazing graphics, and WoW has achieved amazing effects using just simple old texture mapping and textured drop-shadows.
Ed (and yes I have gone off on a bit of a tangent)