Stop Microsoft

Operating Systems => Linux and UNIX => Topic started by: Lead Head on 12 May 2010, 22:12

Title: Valve porting Games to Linux
Post by: Lead Head on 12 May 2010, 22:12
Quote
Found already within the Steam store are Linux-native games like Unreal Tournament 2004, World of Goo, and titles from id Software such as Enemy Territory: Quake Wars and Doom 3. Now that the Source Engine is officially supported on Linux, some Source-based games will be coming over too. Will we finally see Unreal Tournament 3 surface on Linux too? Only time will tell, but it is something we speculated back in 2008. Postal III is also being released this year atop the Source Engine and it will be offering up a native client. We have confirmed that Valve's latest and popular titles like Half-Life 2, Counter-Strike: Source, and Team Fortress 2 are among the first of the Steam Linux titles, similar to the Mac OS X support. The released Linux client should be available by the end of summer.
http://www.phoronix.com/scan.php?page=article&item=valve_steam_announcement&num=1 (http://www.phoronix.com/scan.php?page=article&item=valve_steam_announcement&num=1)

About damn time a developer takes Linux seriously. This may just be the push that *Nix systems need to become more useable for the average person.
Title: Re: Valve porting Games to Linux
Post by: Refalm on 12 May 2010, 23:48
As a Steam user, I really like this very much :o
Title: Re: Valve porting Games to Linux
Post by: Aloone_Jonez on 12 May 2010, 23:53
A group of companies need to get together and build an opensource console and split the profits. If it's cheap, easy to develop games for and provides good all round value to the consumer, the developers should come rolling in.
Title: Re: Valve porting Games to Linux
Post by: Kintaro on 13 May 2010, 00:29
This certainly would make it a far less boring desktop OS. It will however probably suffer stability problems, as there are many distributions yet there will be no source code so there will be no downstream patches to fix up those little differences. Yet, I am sure developers as big as Valve could have a very positive influence on the current and very chaotic development process of distributions.
Title: Re: Valve porting Games to Linux
Post by: piratePenguin on 13 May 2010, 00:40
This certainly would make it a far less boring desktop OS. It will however probably suffer stability problems, as there are many distributions yet there will be no source code so there will be no downstream patches to fix up those little differences. Yet, I am sure developers as big as Valve could have a very positive influence on the current and very chaotic development process of distributions.
http://ubuntuforums.org/showthread.php?t=1459832&page=2 (http://ubuntuforums.org/showthread.php?t=1459832&page=2)
Quote
Inspect the files yourself, and you'll se it actually contains binaries and libraries for GNU/Linux (32 bit). It also links to OpenGL, X server and steam's gui, so this is not a game server for GNU/Linux. The steam executable also contains the names redhat, fedora, slackware, debian, mandrake, yellowdog, gentoo, lsb and suse.
Perhaps there won't be so many stability issues.
Title: Re: Valve porting Games to Linux
Post by: Kintaro on 13 May 2010, 03:17
Well, lets say hypothetically Ubuntu decide to change something downstream and that requires changes to a lot of other downstream applications. In the same update, Ubuntu can easily add patches and change all those as well... The update goes out and suddenly the source engine doesn't work because there is no source code to patch, until Valve release a new steam engine build. Bugs like this popped up a lot when pulseaudio was new for anyone not using package manager packages, or using obselete ones. In a complete absence of source code this can get a lot worse

I don't think this will last, Valve will probably give up on Linux support because the operating system is more trouble than it is worth. The market share is tiny, and with a moving target for an ABI it means this disproportionate amount of users also requires more effort and time invested. To get real application support OSS devs simply need to focus on having a more organized ABI that isn't changed in the downstream at all. The newest LSB will iron out a lot of these problems, but the distributors seem keen to actively ignore them.