Slashdot Mirror
Does Linux Have Game?
kwpulliam writes "Tom's Hardware has an interesting writeup, discussing the difficulties in bringing games to Linux, and the dilemmas faced by the graphics card developers."
← Back to Stories (view on slashdot.org)
← Back to Stories (view on slashdot.org)
Why can't someone port the Direct3D API to Linux? This would save a lot of hassle of porting the games to OpenGL.
How am I supposed to fit a pithy, relevant quote into 120 characters?
Well, I hate to say it but one of the biggest titles coming to Linux was pre-empted from Linux, OS X, and even Windows in favor of the X-box. Yes, eventually it shipped for Windows and OS X, but Linux was left out in the cold when Microsoft purchased Bungie. Bungie had plans for simultaneous release of Halo on Windows and OS X to be followed soon by a Linux release. That all changed when Bungie was bought out. Honestly given the consolidation within the game industry, I don't see much hope for games on Linux for a few years yet which is sort of odd given Linux's marketshare as being so much greater than OS X. Perhaps Toms Hardware is correct when it comes to Linux being a true desktop replacement?
Visit Jonesblog and say hello.
This is the final goal for open source. They need to become a gaming platform, both commercially and not, if they want to win the war with Microsoft.
Game developers should only use OpenGL for newly written rendering code. It is a high performance, advancing standard. Microsoft often "borrows" from it for Direct X (n+1).
OpenGL makes Linux, BSD, MacOS X and other ports practical as well as not having to deal with Microsoft's arbitrary API's.
Ok, maybe Direct3D is the main GUI on windows but it is of no relevance anywhere else. Not on Linux. Not on BSD. Not on MacOS. Not on the PS2. Not even Windows-CE and therefore not on PDAs and Mobile Phones. Nowhere except Wintel.
Therefore any sane producer should use OpenGL which makes it very easy to bring a product to any console, any computer and after some years to PDAs and Mobile Phones.
Today a productive sellcycle for a game should be:
1. sell on consoles - they have less problems with piracy and people are more willing to pay (overpriced) prices.
2. after the first hot sellcycle - three to twelve months - port the game to Desktop-Computers. There are million people which do not want a console but still want to play a game. Also those games are a bit cheaper which gets you more customers. Your game gets pirated though but as sales for (1.) already are over you simply live with it.
3. Meanwhile port it to Unix-Desktops or in other words, MacOS and Linux. It is a piece of cake, given you planes porting through all cycles means you gave some thoughts right at the start about portability and this pay off now. 10-20% additional sales for maybe 10 working days.
4. After several years portable gameconsoles, PDAs and Cell-Phones are becoming a target - your game outsold on the desktop and consoles, people knew your game and have good memory. People actually are willing to pay for Mario/PDA after having played Mario/Console. So now you recompile your game to those portable plattforms, sell it for a real low-budget-price and most likely you are outselling all earlier sales.
All portings only take some days of work, yes this can be done with some planing.
This cycle and no other will give maximum earnings. And it simply forbids use of Direct3D.
"Life is short and in most cases it ends with death." Sir Sinclair
NVIDIA does have alot of third party extensions and other third party IP in their code...so if they open sourced that it would piss off alot of other companies.
Needless to say, I hope one day they can provide a nice driver that doesn't taint the kernel, however I am grateful that they are actually giving us something that is fully functional in Linux (x86, AMD-64) and FreeBSD.
It's because XP problems are often totally random. See, one of the fundemental issues with XP is it thinks its smarter than you. That means, that under certain conditions, it does some totally weird things. For example, I recently tried to figure out why one of my XP machines suddenly became unable to connect to our wireless network. It'd go into a continuous cycle of getting and losing the connection once every few seconds.
Long story short, it turns out this "bug" was actually a feature. You see, if Windows XP is connected to a network that isn't broadcasting its SSID, and it sees one that is, it'll try to disconnect and join the other one. This happens even if the non-broadcasting connection is the only one in your "preferred networks" list. What had happened, of course, is that our neighbor had just got a wireless network, and forgot to turn of SSID broadcasting like you're supposed to. This particular machine was the only one within range, and as a result, freaked out.
Come to think of it, this is probably the same problem that plagued another one of my XP machines. I spent months trying to figure out why it'd randomly drop its connection, and finally gave up, assumed it was a hardware problem, and replaced the PCI wifi adapter with a small USB keychain one. Since it worked, I assumed that the USB dongle got better reception. Now I realize that the problem was just the opposite --- since the USB dongle got *worse* reception, it was out of range of a competing network.
I have lost tens of hours to this single damn bug, as well as $60 bucks buying new hardware I didn't need. All because some idiot programmer at Microsoft thought he could make a piece of software smarter than me. Whoever wrote the "Wireless Network ZeroConfiguration" for XP deserves a special place in hell!
A deep unwavering belief is a sure sign you're missing something...