Id *like* a linux version of NS2, but id rather UW maximized their profit and was stable as a games company to keep producing games. So I'll let them make that call.
But it seems to me that Linux is an untapped market for ns2, UW and/or spark engine... ;-) and should get some consideration beyond just how hard it is to get ns2 going. (see last section ;)
If steam is the blocker, then steam is the blocker. (Although maybe if you said you wanted to make a linux version you could kick them up the pants about getting a steam client out ;) Or maybe UW could build their own linux supporting 'auth/voice' library to plug into the game for linux (atleast for future spark mods) ;-)
Most of the problem getting software going under linux is uninstalled libraries, library versions and compatibility and where stuff is located.
The solution is to install into your own tree and have almost everything you need under there (libraries, etc) and dont rely on external stuff. (Many windows games do this, or just reinstall whatever exact version of directx they need, so this isnt a 'linux problem').
Its not as friendly as the 'compile it from source' way, but this approach works fine for many big projects (eg the excellent 'splunk' log monitor), but people usually dont care as long as it works fine. You'll be installing 3G of stuff, so who cares if its 3.1G if you include your own libraries.
As as to linux/mac/windows users. Look again at the <a href="http://www.humblebundle.com/" target="_blank">http://www.humblebundle.com/</a> figures where linux users are willing to spend almost 3 times more money on games/charity as windows users. ;)
If I were a game company then knowing that platform X users will spend more money than platform Y users might just make me consider them as a market, especially if I could corner the market with my sexy new graphics engine. That would suggest to me that they might also be more willing to buy more of my games (especially if they were usable on any platform, including my beloved linux).
Also consider that NS2 on linux would get covered by a number of linux mags which gets the news out to the people who we previously determined were more willing to spend *cash*. So, small indy game on windows or one of the only AAAish games on linux (with free targeted advertising).
I agree that most linux people usually have a win/mac desktop, but that does not mean that it isnt an untapped market, especially if the spark engine worked on it, allowing mods to be portable to all three platforms.
And unlinke windows, you probably dont need a modern warfare type budget to corner that market. And the platform isnt as restricted as the mac platform (ie no apple tax) ;-)
If I was UW's Linux advisor, id suggest:
* get dedicated server out asap (Hell, I'll offer to help get it working if you need linux/compile help!! ;)
* concentrate on getting game shipped, then look at linux version (you probably want to tap that free linux magazine press with a working polished game people can leap right into).
Just my 6 cents (thats 3 lots of windows 2 cents ;)
Just wanted to note that UWE didn't say there would definitely be no linux client. Twitter post says "not sure" so well.. they're not sure what to do, which is better than certainly not :)
I could imagine something like "we do it if we get 1k preorders for linux". Or someone's willing to do a port (icculus & co. ftw).
And on the steam issue.. Desura could serve as a nice distribution platfrom (see <a href="http://www.desura.com/news/desura-linux-developement-downloads-and-bug-fixes" target="_blank">Desura linux development blog post</a>).
<!--quoteo(post=1866986:date=Aug 6 2011, 10:34 AM:name=QuelTos)--><div class='quotetop'>QUOTE (QuelTos @ Aug 6 2011, 10:34 AM) <a href="index.php?act=findpost&pid=1866986"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Just wanted to note that UWE didn't say there would definitely be no linux client. Twitter post says "not sure" so well.. they're not sure what to do, which is better than certainly not :)
I could imagine something like "we do it if we get 1k preorders for linux". Or someone's willing to do a port (icculus & co. ftw).
And on the steam issue.. Desura could serve as a nice distribution platfrom (see <a href="http://www.desura.com/news/desura-linux-developement-downloads-and-bug-fixes" target="_blank">Desura linux development blog post</a>).<!--QuoteEnd--></div><!--QuoteEEnd--> i mentioned the "not sure" remark but was ignored oh well
So I thought I would chime in on this as someone who has worked on games that were ported to the Mac and did linux development for many years. I will first admit that I am a Mac user I own several of them, they are not my primary gaming machines as I do own a rather nice PC for doing development on as well as for playing games on but none the less I am a Mac user. I also own and run a linux machine mainly for development and thats mainly because the Linux community really just lacks in its gaming and many other things which I will touch on in this post as possible reasons why one would not want to release a AAA game on it.
First off I would like to comment on the OpenGL stuff and these comments, “Well if you guys port it to Mac that uses OpenGL and so does linux so it will be easy!†While they both do use the same graphics sub-system I just want to say personally developing for OpenGL sucks! The APIs are very dated and are full of bloat from the years of extensions and 3rd parties adding mountains of crap on top of the base APIs and for years the Khronos Group has been touting how they plan to bring the API into the modern computing age and every time we see a new release of OpenGL they fall short. While many of those using a Mac or Linux are not usually supporters of Microsoft you cannot really argue that DirectX is a horrible API because its far from that and where OpenGL fails DirectX really shines and thats in the ease of use in the API and the constant updating with modern technology. Even the heavy users of the OpenGL framework like Id will admit that OpenGL has many short comings and in general is not the most efficient graphics API on the market.
With that said one is correct that they do use the same graphics sub-system and that part of the game would be cross-platform for both OSX and Linux but that in itself is not the main reason I would be turned away from porting to Linux; there are a lot of differences with relation to OS related tasks; OSX is not Linux at its core is BSD which means it has minor differences at the core level that require Linux and OSX implementations of certain things. On top of that we there are some differences in just interacting with the OS... One being the programming language that OSX uses which is Objective-C and Linux being C++ and I mean there is just a lot of subtle differences that make porting annoying but those are not even the reason why I would not port to Linux.
The biggest two reasons are Marketshare and Support... Most game companies who put out Linux versions of their games do not support them and sometimes they do not even updated on the same track that the normal game does and that has a lot to do with the Marketshare of the linux gamer base being much smaller than that of the Mac users; the reality is there is money to be made from selling your games on Macs mainly because their user base is in line with target audience for a game like NS2. Macs have a wide appeal to college age students which are more likely to play this game and opens the game to a much wider audience; the type of computer user Linux attracts would be people who are likely to play NS2 but then again the Marketshare of those users does not seem to be worth the resources required to get the client working. Beyond this I go back to the support issues I mentioned above since the Linux community cannot seem to agree on any real standard for anything. It makes supporting even one Linux distro rather complicated and so if there was a Linux port it would be as I mentioned before unsupported and I do not know if that aligns with our goals as a developer of games.
I am not saying there would not be a Linux port but I would like to say that all those people who are just saying well you did most the leg work why not go the extra mile... are just not seeing the whole picture and the complications of supporting more than one platform.
Aye, that's why I said I lack knowledge and would like an informed opinion. Thanks wacko.
It is quite true that most of the support is only distro-wide and done by that particular community. If something is not open source than there is a lot less support from the communities to expect since it won't be available in the maintained repos.
But don't those issues also apply to the server? Or is the server so much easier to maintain that the client that it isn't a relevant factor?
After looking at the steam mac data, only around ~15% of all macusers have good enough gfx cards to run* ns2 atm. Sadly most ppl can only afford the cheapest models to be one of the cool kids...
* doesnt mean that it would run smooth
UWE Mac challenge: get smooth fps around 30 on a 9400m with at least 1280 x 800res :P (intel hd3000+ could also be looked at, apple loves the new integrated gpus)
AsranielJoin Date: 2002-06-03Member: 724Members, Playtest Lead, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow, Subnautica Playtester, Retired Community Developer
edited August 2011
In a Q&A newspost, that the plan would be to open source the parts of the engine that are linux specific, so that the community could in part support it. Is this still the plan?
edit: I would also like to add that if there is a linux server, and an OSX client, then indeed porting the client to linux should not be that hard. And opengl became much more unified and better with the release of opengl 3+. The binary/proprietary linux GPU drivers support the latest opengl versions, so that should not be a problem. LInux is already at opengl 4+ i think, and osx "only" at 3.2, but that shouldn't be such a big problem. By the way, i doubt there is any objective C code to be written for a osx port. I wrote pure c++ apps that where ported to windows and osx (from linux), and i didn't have to write objective-c.
<!--quoteo(post=1867129:date=Aug 7 2011, 03:53 AM:name=Asraniel)--><div class='quotetop'>QUOTE (Asraniel @ Aug 7 2011, 03:53 AM) <a href="index.php?act=findpost&pid=1867129"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->In a Q&A newspost, that the plan would be to open source the parts of the engine that are linux specific, so that the community could in part support it. Is this still the plan?
edit: I would also like to add that if there is a linux server, and an OSX client, then indeed porting the client to linux should not be that hard. And opengl became much more unified and better with the release of opengl 3+. The binary/proprietary linux GPU drivers support the latest opengl versions, so that should not be a problem. LInux is already at opengl 4+ i think, and osx "only" at 3.2, but that shouldn't be such a big problem. By the way, i doubt there is any objective C code to be written for a osx port. I wrote pure c++ apps that where ported to windows and osx (from linux), and i didn't have to write objective-c.<!--QuoteEnd--></div><!--QuoteEEnd-->
I am just going to say Asraniel that I disagree with you and my years of experience with regards to developing games on Windows, Linux and OSX has shown me there is a lot involved when it comes to cross-platform development and I would be more than happy to share those details with you offline if you would like.
AsranielJoin Date: 2002-06-03Member: 724Members, Playtest Lead, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow, Subnautica Playtester, Retired Community Developer
Sure, would love to!
My crossplatform development is mainly non game specific, involving the Qt library which really makes it as easy as just recompiling the code on the desired platform.
I mean its not new to those who have been developing on both platforms that OpenGL is a pain to work with at times and needs some love and Microsoft has really made it easy for developers to use their API as I mentioned in my previous post :P
Asraniel I am pretty sure if you looked at the code for QT mainly dealing with the UI, audio, maybe some OS level parts of it you will find there is objective-c code :P I know for a fact there is. QTs job is to abstract away all the OS specific stuff we do not have that luxury all the time with games.
Just out of curiosity, does anyone have statistics of how many users play next gen games on linux? I would love to see numbers showing it would be a wise investment to invest expensive man hours into supporting a platform that is not meant for gaming... After all my years of gaming, traveling to events all around the world, I'd be lieing if I said I've seen/heard of one gamer using linux as their gaming platform.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...
<!--quoteo(post=1867371:date=Aug 8 2011, 09:14 PM:name=wacko)--><div class='quotetop'>QUOTE (wacko @ Aug 8 2011, 09:14 PM) <a href="index.php?act=findpost&pid=1867371"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Asraniel I am pretty sure if you looked at the code for QT mainly dealing with the UI, audio, maybe some OS level parts of it you will find there is objective-c code :P I know for a fact there is. QTs job is to abstract away all the OS specific stuff we do not have that luxury all the time with games.<!--QuoteEnd--></div><!--QuoteEEnd-->
Err. Actually, the job of an operating system is to abstract away all the hardware specific stuff and provide a clean, known interface. The job of the language is to abstract the requirement of machine code.
I think you will find that objective-c and c++ and all languages are really compiled into machine code and not actually run on the machine.
Code -> Compiler -> machine code + known api = program. Objective-C is never run on the host, it just generates machine code which is run and calls a predefined interface provided by the OS. All OS interfaces are really in machine code and thus technically available from any language.
The issue isnt the language (thats just compiled into machine instructions) but rather the API offered by the operating system and the data structures needed (thats really the part that cause the most headaches).
Id be *massively* surprised if the API on any operating system didnt just use native/C level data structures. Any OS that isnt useable from c/c++ is really close to shooting itself in the head these days.
Usually you just need to put a wrapper round your external calls depending on how the operating system likes to be called (eg to open a file on windows call XXX(char*,int) but on mac, call YYY(int,char*) and on linux ZZZ(char *, int, int))
c++ is available for macs. This is where all the fink (http://www.finkproject.org/) stuff comes from. It takes OSS software and compiles it on the mac.
<!--quoteo(post=1867390:date=Aug 8 2011, 10:42 PM:name=vizionz)--><div class='quotetop'>QUOTE (vizionz @ Aug 8 2011, 10:42 PM) <a href="index.php?act=findpost&pid=1867390"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Just out of curiosity, does anyone have statistics of how many users play next gen games on linux? I would love to see numbers showing it would be a wise investment to invest expensive man hours into supporting a platform that is not meant for gaming... After all my years of gaming, traveling to events all around the world, I'd be lieing if I said I've seen/heard of one gamer using linux as their gaming platform.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...<!--QuoteEnd--></div><!--QuoteEEnd-->
I think this is a chicken and egg problem. Which of these is the root cause noone playing FPS on linux? ;)
"Noone plays FPS on linux, so I wont port my FPS to linux"
"I wont port my FPS to linux so noone *GETS* play FPS on linux"
I _only_ run windows as a terminal to my unix machine and a games platform. If i could play games on linux, Id never run windows at all. So I guess id count me as a windows gamer then? Not through choice. I count as a windows user the same way we all count as oxygen breathers (its the only real option at the moment).
I agree it would be an investment in time and effort. And I agree UWE should focus on getting NS2 playable on windows. +10000 for that!
*But* if they want to offer the spark engine as a dev platform, then linux compatibility would be a very nice option (not a requirement) for the linux community (even without steam/voice) and does represent an untapped market that noone else is courting.
<!--quoteo(post=1868268:date=Aug 11 2011, 03:12 AM:name=slayer.faith)--><div class='quotetop'>QUOTE (slayer.faith @ Aug 11 2011, 03:12 AM) <a href="index.php?act=findpost&pid=1868268"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->If i could play games on linux, Id never run windows at all. So I guess id count me as a windows gamer then?<!--QuoteEnd--></div><!--QuoteEEnd-->
yeh same here, id be using linux all the time, but there are just a couple applications and games that lock me into having to dual boot windows. all my sound apps work through wine, but ns2 doesn't, and wine wont let me drag n drop so to keep my work-flow up do all my work in windows. i think linux has some clear advantages over windows, and if it just had more support from developers we would see alot more ppl using it
in the few years ive been using it tho ive seen wine come a long way, sometimes running the windows version better than native linux versions. i know that wine make a developers kit that helps devs make their apps work with wine maybe ns2 could do that. wine is for macs too.
AsranielJoin Date: 2002-06-03Member: 724Members, Playtest Lead, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow, Subnautica Playtester, Retired Community Developer
edited August 2011
<!--quoteo(post=1867371:date=Aug 8 2011, 01:14 PM:name=wacko)--><div class='quotetop'>QUOTE (wacko @ Aug 8 2011, 01:14 PM) <a href="index.php?act=findpost&pid=1867371"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Asraniel I am pretty sure if you looked at the code for QT mainly dealing with the UI, audio, maybe some OS level parts of it you will find there is objective-c code :P I know for a fact there is. QTs job is to abstract away all the OS specific stuff we do not have that luxury all the time with games.<!--QuoteEnd--></div><!--QuoteEEnd-->
I just checked the Qt sourcetree. Its all pure C++, no objective-c. If i had IRC here i would quickly check with the Qt developers, but it sounded illogical from the beginning.
Objective-c is just a wrapper around the low level API. Qt/C++ is just a different one.
Edit: i was wrong, there is actually some objective-c code in Qt, like that example: <a href="http://qt.gitorious.org/qt/qt/blobs/4.7/src/gui/dialogs/qfiledialog_mac.mm" target="_blank">http://qt.gitorious.org/qt/qt/blobs/4.7/sr...ledialog_mac.mm</a>
<!--quoteo(post=1866933:date=Aug 6 2011, 07:06 AM:name=Chris0132)--><div class='quotetop'>QUOTE (Chris0132 @ Aug 6 2011, 07:06 AM) <a href="index.php?act=findpost&pid=1866933"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Which means they can probably ill afford any more losses.
There is nobody in the world that uses linux because they don't have access to windows or OSX, therefore developing for linux is a waste of time, as anyone who could buy it for linux can also buy it for one of the other two platforms. Linux is a curiosity, it is not a market share.<!--QuoteEnd--></div><!--QuoteEEnd--> <!--quoteo(post=1868276:date=Aug 10 2011, 09:56 PM:name=Smug_Lobster)--><div class='quotetop'>QUOTE (Smug_Lobster @ Aug 10 2011, 09:56 PM) <a href="index.php?act=findpost&pid=1868276"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->*Image Cut* I DEMAND GNU SUPPORT<!--QuoteEnd--></div><!--QuoteEEnd--> <!--quoteo(post=1866835:date=Aug 5 2011, 08:03 PM:name=RichardRahl)--><div class='quotetop'>QUOTE (RichardRahl @ Aug 5 2011, 08:03 PM) <a href="index.php?act=findpost&pid=1866835"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->theres a reason that alot of companys don't support linux, it's a pain in the ass due to it being open source and there are so many things that can go wrong, something you have installed on your linux build might not be the same as someone elses.<!--QuoteEnd--></div><!--QuoteEEnd-->
I WILL CUT YOU ALL! CUT YOU 'TIL ALL THE RED STUFF COMES OUT!
<!--quoteo(post=1867079:date=Aug 6 2011, 07:29 PM:name=Jiriki)--><div class='quotetop'>QUOTE (Jiriki @ Aug 6 2011, 07:29 PM) <a href="index.php?act=findpost&pid=1867079"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Lack of standards. Library problems.
As a Linux sysadmin I completely agree.<!--QuoteEnd--></div><!--QuoteEEnd-->
You're getting cut too... Stop using Red Hat, CentOS, etc... you're a terrible person.
<!--quoteo(post=1867390:date=Aug 8 2011, 05:42 AM:name=vizionz)--><div class='quotetop'>QUOTE (vizionz @ Aug 8 2011, 05:42 AM) <a href="index.php?act=findpost&pid=1867390"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Just out of curiosity, does anyone have statistics of how many users play next gen games on linux? I would love to see numbers showing it would be a wise investment to invest expensive man hours into supporting a platform that is not meant for gaming... After all my years of gaming, traveling to events all around the world, I'd be lieing if I said I've seen/heard of one gamer using linux as their gaming platform.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...<!--QuoteEnd--></div><!--QuoteEEnd-->
See like the previous poster stated no one plays next gen games on Linux because no one writes them for Linux... and no one writes, because no one plays... etc... etc... etc...
People keep saying it's fragmented because of all the different types... but they're all based on the same Kernel.
In 2010 Ubuntu claims 12 million users... what ever else for derivatives, and it's based on Debian... so the user base is actually larger than it ever shows... The real problem is as a few have stated is that we have WINE (piss poor substitute) and can dual boot... Which is why the users go unnoticed... is because if the want to play games (the only reason I still have Windows plaguing my hard drive) they dual booted. So far every Linux user there's an additional Windows user, just so they can play games that WINE can get to run (curse you LoL).
And as a developer, who's currently developing our first title, for <b>all </b>major PC platforms... I can honestly say... it's not that hard. Use Abstraction correctly and it add's a negligible amount of work... like seriously... < 1%. 5% if you have to refactor a bunch.
And I'll take package management on Linux any day over Windows.
I'm getting really tired of people justifying not supporting something because no one currently supports it.
It's like finding starving a child, looking at him and saying "Meh, why feed him, no one else has..." WTF?
Comments
But it seems to me that Linux is an untapped market for ns2, UW and/or spark engine... ;-) and should get some consideration beyond just how hard it is to get ns2 going. (see last section ;)
If steam is the blocker, then steam is the blocker. (Although maybe if you said you wanted to make a linux version you could kick them up the pants about getting a steam client out ;) Or maybe UW could build their own linux supporting 'auth/voice' library to plug into the game for linux (atleast for future spark mods) ;-)
Most of the problem getting software going under linux is uninstalled libraries, library versions and compatibility and where stuff is located.
The solution is to install into your own tree and have almost everything you need under there (libraries, etc) and dont rely on external stuff. (Many windows games do this, or just reinstall whatever exact version of directx they need, so this isnt a 'linux problem').
Its not as friendly as the 'compile it from source' way, but this approach works fine for many big projects (eg the excellent 'splunk' log monitor), but people usually dont care as long as it works fine. You'll be installing 3G of stuff, so who cares if its 3.1G if you include your own libraries.
As as to linux/mac/windows users. Look again at the <a href="http://www.humblebundle.com/" target="_blank">http://www.humblebundle.com/</a> figures where linux users are willing to spend almost 3 times more money on games/charity as windows users. ;)
If I were a game company then knowing that platform X users will spend more money than platform Y users might just make me consider them as a market, especially if I could corner the market with my sexy new graphics engine. That would suggest to me that they might also be more willing to buy more of my games (especially if they were usable on any platform, including my beloved linux).
Also consider that NS2 on linux would get covered by a number of linux mags which gets the news out to the people who we previously determined were more willing to spend *cash*. So, small indy game on windows or one of the only AAAish games on linux (with free targeted advertising).
I agree that most linux people usually have a win/mac desktop, but that does not mean that it isnt an untapped market, especially if the spark engine worked on it, allowing mods to be portable to all three platforms.
And unlinke windows, you probably dont need a modern warfare type budget to corner that market. And the platform isnt as restricted as the mac platform (ie no apple tax) ;-)
If I was UW's Linux advisor, id suggest:
* get dedicated server out asap (Hell, I'll offer to help get it working if you need linux/compile help!! ;)
* concentrate on getting game shipped, then look at linux version (you probably want to tap that free linux magazine press with a working polished game people can leap right into).
Just my 6 cents (thats 3 lots of windows 2 cents ;)
I could imagine something like "we do it if we get 1k preorders for linux". Or someone's willing to do a port (icculus & co. ftw).
And on the steam issue.. Desura could serve as a nice distribution platfrom (see <a href="http://www.desura.com/news/desura-linux-developement-downloads-and-bug-fixes" target="_blank">Desura linux development blog post</a>).
I could imagine something like "we do it if we get 1k preorders for linux". Or someone's willing to do a port (icculus & co. ftw).
And on the steam issue.. Desura could serve as a nice distribution platfrom (see <a href="http://www.desura.com/news/desura-linux-developement-downloads-and-bug-fixes" target="_blank">Desura linux development blog post</a>).<!--QuoteEnd--></div><!--QuoteEEnd-->
i mentioned the "not sure" remark but was ignored oh well
First off I would like to comment on the OpenGL stuff and these comments, “Well if you guys port it to Mac that uses OpenGL and so does linux so it will be easy!†While they both do use the same graphics sub-system I just want to say personally developing for OpenGL sucks! The APIs are very dated and are full of bloat from the years of extensions and 3rd parties adding mountains of crap on top of the base APIs and for years the Khronos Group has been touting how they plan to bring the API into the modern computing age and every time we see a new release of OpenGL they fall short. While many of those using a Mac or Linux are not usually supporters of Microsoft you cannot really argue that DirectX is a horrible API because its far from that and where OpenGL fails DirectX really shines and thats in the ease of use in the API and the constant updating with modern technology. Even the heavy users of the OpenGL framework like Id will admit that OpenGL has many short comings and in general is not the most efficient graphics API on the market.
With that said one is correct that they do use the same graphics sub-system and that part of the game would be cross-platform for both OSX and Linux but that in itself is not the main reason I would be turned away from porting to Linux; there are a lot of differences with relation to OS related tasks; OSX is not Linux at its core is BSD which means it has minor differences at the core level that require Linux and OSX implementations of certain things. On top of that we there are some differences in just interacting with the OS... One being the programming language that OSX uses which is Objective-C and Linux being C++ and I mean there is just a lot of subtle differences that make porting annoying but those are not even the reason why I would not port to Linux.
The biggest two reasons are Marketshare and Support... Most game companies who put out Linux versions of their games do not support them and sometimes they do not even updated on the same track that the normal game does and that has a lot to do with the Marketshare of the linux gamer base being much smaller than that of the Mac users; the reality is there is money to be made from selling your games on Macs mainly because their user base is in line with target audience for a game like NS2. Macs have a wide appeal to college age students which are more likely to play this game and opens the game to a much wider audience; the type of computer user Linux attracts would be people who are likely to play NS2 but then again the Marketshare of those users does not seem to be worth the resources required to get the client working. Beyond this I go back to the support issues I mentioned above since the Linux community cannot seem to agree on any real standard for anything. It makes supporting even one Linux distro rather complicated and so if there was a Linux port it would be as I mentioned before unsupported and I do not know if that aligns with our goals as a developer of games.
I am not saying there would not be a Linux port but I would like to say that all those people who are just saying well you did most the leg work why not go the extra mile... are just not seeing the whole picture and the complications of supporting more than one platform.
It is quite true that most of the support is only distro-wide and done by that particular community. If something is not open source than there is a lot less support from the communities to expect since it won't be available in the maintained repos.
But don't those issues also apply to the server? Or is the server so much easier to maintain that the client that it isn't a relevant factor?
Library problems.
As a Linux sysadmin I completely agree.
Sadly most ppl can only afford the cheapest models to be one of the cool kids...
* doesnt mean that it would run smooth
UWE Mac challenge: get smooth fps around 30 on a 9400m with at least 1280 x 800res :P (intel hd3000+ could also be looked at, apple loves the new integrated gpus)
edit: I would also like to add that if there is a linux server, and an OSX client, then indeed porting the client to linux should not be that hard. And opengl became much more unified and better with the release of opengl 3+. The binary/proprietary linux GPU drivers support the latest opengl versions, so that should not be a problem. LInux is already at opengl 4+ i think, and osx "only" at 3.2, but that shouldn't be such a big problem.
By the way, i doubt there is any objective C code to be written for a osx port. I wrote pure c++ apps that where ported to windows and osx (from linux), and i didn't have to write objective-c.
edit: I would also like to add that if there is a linux server, and an OSX client, then indeed porting the client to linux should not be that hard. And opengl became much more unified and better with the release of opengl 3+. The binary/proprietary linux GPU drivers support the latest opengl versions, so that should not be a problem. LInux is already at opengl 4+ i think, and osx "only" at 3.2, but that shouldn't be such a big problem.
By the way, i doubt there is any objective C code to be written for a osx port. I wrote pure c++ apps that where ported to windows and osx (from linux), and i didn't have to write objective-c.<!--QuoteEnd--></div><!--QuoteEEnd-->
I am just going to say Asraniel that I disagree with you and my years of experience with regards to developing games on Windows, Linux and OSX has shown me there is a lot involved when it comes to cross-platform development and I would be more than happy to share those details with you offline if you would like.
My crossplatform development is mainly non game specific, involving the Qt library which really makes it as easy as just recompiling the code on the desired platform.
<a href="http://translate.google.com/translate?hl=de&sl=de&tl=en&u=http%3A%2F%2Fwww.tomshardware.de%2FDirectX-OpenGL%2Ctestberichte-240159.html" target="_blank">http://translate.google.com/translate?hl=d...hte-240159.html</a>
German: <a href="http://www.tomshardware.de/DirectX-OpenGL,testberichte-240159-8.html" target="_blank">http://www.tomshardware.de/DirectX-OpenGL,...e-240159-8.html</a>
PS: dont have any real knowledge, wonder whats going on now on the open gl front
Asraniel I am pretty sure if you looked at the code for QT mainly dealing with the UI, audio, maybe some OS level parts of it you will find there is objective-c code :P I know for a fact there is. QTs job is to abstract away all the OS specific stuff we do not have that luxury all the time with games.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...
Err. Actually, the job of an operating system is to abstract away all the hardware specific stuff and provide a clean, known interface. The job of the language is to abstract the requirement of machine code.
I think you will find that objective-c and c++ and all languages are really compiled into machine code and not actually run on the machine.
Code -> Compiler -> machine code + known api = program. Objective-C is never run on the host, it just generates machine code which is run and calls a predefined interface provided by the OS. All OS interfaces are really in machine code and thus technically available from any language.
The issue isnt the language (thats just compiled into machine instructions) but rather the API offered by the operating system and the data structures needed (thats really the part that cause the most headaches).
Id be *massively* surprised if the API on any operating system didnt just use native/C level data structures. Any OS that isnt useable from c/c++ is really close to shooting itself in the head these days.
Usually you just need to put a wrapper round your external calls depending on how the operating system likes to be called (eg to open a file on windows call XXX(char*,int) but on mac, call YYY(int,char*) and on linux ZZZ(char *, int, int))
c++ is available for macs. This is where all the fink (http://www.finkproject.org/) stuff comes from. It takes OSS software and compiles it on the mac.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...<!--QuoteEnd--></div><!--QuoteEEnd-->
I think this is a chicken and egg problem. Which of these is the root cause noone playing FPS on linux? ;)
"Noone plays FPS on linux, so I wont port my FPS to linux"
"I wont port my FPS to linux so noone *GETS* play FPS on linux"
I _only_ run windows as a terminal to my unix machine and a games platform. If i could play games on linux, Id never run windows at all. So I guess id count me as a windows gamer then? Not through choice. I count as a windows user the same way we all count as oxygen breathers (its the only real option at the moment).
I agree it would be an investment in time and effort. And I agree UWE should focus on getting NS2 playable on windows. +10000 for that!
*But* if they want to offer the spark engine as a dev platform, then linux compatibility would be a very nice option (not a requirement) for the linux community (even without steam/voice) and does represent an untapped market that noone else is courting.
yeh same here, id be using linux all the time, but there are just a couple applications and games that lock me into having to dual boot windows. all my sound apps work through wine, but ns2 doesn't, and wine wont let me drag n drop so to keep my work-flow up do all my work in windows. i think linux has some clear advantages over windows, and if it just had more support from developers we would see alot more ppl using it
in the few years ive been using it tho ive seen wine come a long way, sometimes running the windows version better than native linux versions. i know that wine make a developers kit that helps devs make their apps work with wine maybe ns2 could do that. wine is for macs too.
I DEMAND GNU SUPPORT
I just checked the Qt sourcetree. Its all pure C++, no objective-c. If i had IRC here i would quickly check with the Qt developers, but it sounded illogical from the beginning.
Objective-c is just a wrapper around the low level API. Qt/C++ is just a different one.
Edit: i was wrong, there is actually some objective-c code in Qt, like that example:
<a href="http://qt.gitorious.org/qt/qt/blobs/4.7/src/gui/dialogs/qfiledialog_mac.mm" target="_blank">http://qt.gitorious.org/qt/qt/blobs/4.7/sr...ledialog_mac.mm</a>
There is nobody in the world that uses linux because they don't have access to windows or OSX, therefore developing for linux is a waste of time, as anyone who could buy it for linux can also buy it for one of the other two platforms. Linux is a curiosity, it is not a market share.<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1868276:date=Aug 10 2011, 09:56 PM:name=Smug_Lobster)--><div class='quotetop'>QUOTE (Smug_Lobster @ Aug 10 2011, 09:56 PM) <a href="index.php?act=findpost&pid=1868276"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->*Image Cut*
I DEMAND GNU SUPPORT<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1866835:date=Aug 5 2011, 08:03 PM:name=RichardRahl)--><div class='quotetop'>QUOTE (RichardRahl @ Aug 5 2011, 08:03 PM) <a href="index.php?act=findpost&pid=1866835"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->theres a reason that alot of companys don't support linux, it's a pain in the ass due to it being open source and there are so many things that can go wrong, something you have installed on your linux build might not be the same as someone elses.<!--QuoteEnd--></div><!--QuoteEEnd-->
I WILL CUT YOU ALL! CUT YOU 'TIL ALL THE RED STUFF COMES OUT!
<!--quoteo(post=1867079:date=Aug 6 2011, 07:29 PM:name=Jiriki)--><div class='quotetop'>QUOTE (Jiriki @ Aug 6 2011, 07:29 PM) <a href="index.php?act=findpost&pid=1867079"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Lack of standards.
Library problems.
As a Linux sysadmin I completely agree.<!--QuoteEnd--></div><!--QuoteEEnd-->
You're getting cut too... Stop using Red Hat, CentOS, etc... you're a terrible person.
<!--quoteo(post=1867390:date=Aug 8 2011, 05:42 AM:name=vizionz)--><div class='quotetop'>QUOTE (vizionz @ Aug 8 2011, 05:42 AM) <a href="index.php?act=findpost&pid=1867390"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Just out of curiosity, does anyone have statistics of how many users play next gen games on linux? I would love to see numbers showing it would be a wise investment to invest expensive man hours into supporting a platform that is not meant for gaming... After all my years of gaming, traveling to events all around the world, I'd be lieing if I said I've seen/heard of one gamer using linux as their gaming platform.
Let UWE focus on what is important, developing a title the general public will buy and be able to play on acceptable minimum PC requirements...<!--QuoteEnd--></div><!--QuoteEEnd-->
See like the previous poster stated no one plays next gen games on Linux because no one writes them for Linux... and no one writes, because no one plays... etc... etc... etc...
People keep saying it's fragmented because of all the different types... but they're all based on the same Kernel.
In 2010 Ubuntu claims 12 million users... what ever else for derivatives, and it's based on Debian... so the user base is actually larger than it ever shows...
The real problem is as a few have stated is that we have WINE (piss poor substitute) and can dual boot... Which is why the users go unnoticed... is because if the want to play games (the only reason I still have Windows plaguing my hard drive) they dual booted. So far every Linux user there's an additional Windows user, just so they can play games that WINE can get to run (curse you LoL).
And as a developer, who's currently developing our first title, for <b>all </b>major PC platforms... I can honestly say... it's not that hard. Use Abstraction correctly and it add's a negligible amount of work... like seriously... < 1%. 5% if you have to refactor a bunch.
And I'll take package management on Linux any day over Windows.
I'm getting really tired of people justifying not supporting something because no one currently supports it.
It's like finding starving a child, looking at him and saying "Meh, why feed him, no one else has..."
WTF?