- including patching for multiple game versions
- operating specific problems (source is not as solid as somebody here thinks because it is a c++ mess/hell, I know from experience)
- operating specific 3rd party tools
- "special hacks" (source games are full of them) which often don't translate well across OSes
- support problems (you have to support more OS which can put quite a damper on your reputation if you can't handle it like Windows)
The UEngine as somebody mentioned it here is a lot more portable than this mess called Source Engine. As mentioned, the problem for the games itself is supporting "another" OS and that's the real problem that steam does not solve. It makes it "at best" "slightly" less troublesome but if you are Sisyphus then it doesn't matter if one less little pebble is on your way as there are still hundred others.
Sooooo... Any news from Valve on the subject?
Who said anything about Linux being a development target? Bug fixing is what is important, compiling your code on multiple platforms with multiple compilers will expose more bugs than one platform and one compiler. Some studios compile on linux simply to expose and fix bugs.Linux is somewhat widely used for things like video and audio editing and post processing, offline rendering, etc. As an actual target for game development (released or not), not so much, and I know this first-hand.
Its really just a rumor until Valve announces it.This is why the possibility of Steam coming to the platform is such a big deal, and I happen to think it may well be very good for Valve to be the first.
And you've worked at how many tier one and two studios?But I have yet to see a first or second tier studio that does PC development do hands-on Linux development of their titles as well.