Linux Version

[Shai]

Hello,

Me, and i'm sure many other simulator pilots would love to see a linux version of a good simulator....
i boot to windows just to fly (unfair)

Thanks :)

[Charlie_VFP]

Porting everything to a second platform will take away time and resources that could be spent making a single platform operate better.

I voted no.

[Zimluura]

i voted yes, for two reasons.

1) microsoft has mentioned plans to charge a yearly fee for using windows. oddly enough ms press releases are the best linux/mac advertisments i've seen.

2) coding multiple platforms forces more abstraction and helps stability for all versions. example: if the developers code 100% for winxp then there is more chance that when longhorn (or even service pack 3) comes out that there will be a compatibility problem and the devs will have to spend a long time hunting down the problem, and then rewrite several different pieces of code to fix things. if the same codebase can compile on winxp, macosx, and linux with smaller, separate, sections of os specific code the developers should be able to get that kind of thing fixed much faster.

[Shai]

(sorry for my bad english)

This is true.

and from my expirience if the game is coded for linux it will actualy run better than Windows... linux is much better with heavy load on the system because of the simple reason that it have less coding mistakes (because of being Open source OS and used by many people and coders).
Example to this is America's Army game which from my testing running better on my Linux OS.

If it compiles on Linux it shouldnt be problem and should take too long to make Windows port.

in fact making it multi platform game will get many people to buy it.
bigger community.
thats what we all want.... i think :)

[Antler]

and from my expirience if the game is coded for linux it will actualy run better than Windows... linux is much better with heavy load on the system because of the simple reason that it have less coding mistakes (because of being Open source OS and used by many people and coders).
Example to this is America's Army game which from my testing running better on my Linux OS.



Are you sure? I don't mean this in a horrible way at all, but was your linux system as heavily loaded as your windows system (e.g. a virus scanner, firewall, etc). However, I do know that linux is less likely to crash, and is arguably better at multitasking. That's why they usually use Linux on Web servers and such.

Having said that, I'm not that much of a linux fan. I have used it, and I didn't really like it. But, this is not a discussion about Linux as an OS so I'll stop there.

As for having multiple versions and higher abstraction, Java comes to mind. I'm not for one second suggesting anybody use Java - it would be far too slow. But, my point is this: has anyone tried using Swing to code a Windows interface? It looks like Windows, but as soon as you start using it, it becomes very apparent it isn't: the feel is wrong, and it's a bit slower. An example of a Java interface program is JBuilder.

[Charlie_VFP]

More time, more programmers, more money. All of the things that G2I doesn't seem to have.

Also, coding it across different platforms means that yes, very few parts of the sim will actually be reliant on specific OS related issues. But this means that it cannot take advantage of OS specific issues.

We have a saying in the Martial Arts world... good at all, master of none.

Sure, it will run "good" on all platforms... but it will be nowhere as good in the overall picture.

Stick to Windows, code it to the teeth, and make it beautiful.

[Zimluura]

Terry, i must respectfully disagree on the good at all, master of none part. the best example i can use to draw this conclusion is id software (always a technology leader, and always on every os)

it's difficult for me to guage your programming experience, so please forgive me if this sounds condescending. but linux, macos, and windows all have c compilers, and they all should adhere (more or less) to the same language standards.

i also can't think of any windows specific features that would benefit coding a game.

the time and resources are definatly a question though. but from my somewhat limited experience with game programming i feel that the time investment would be either negligible in the big picture or could even amount to a savings during debugging sessions.

at this time it is certainly not financially sound to provide individual technical support to linux users (too many distros) but linux users are, in general, more technically competent than windows users and shouldn't have issues with community based tech support, which is free.

anyway, this is all IMHO, no way of knowing what kind of situation the developers face. i'd just like to see more cross platform games.

[Antler]

Linux, MacOS and Windows do all have a C compiler, but the difficulty comes in where you have different libraries that are platform specific. If you have a function call to a certain function within a platform specific library, it may not exists in another library's API, or the function might be called something different.

As far as I know, ID software doesn't make linux ports of games (or at least they aren't shown on their website). I do know that Unreal Tournament does have a linux port, but they always come out after the windows versions.

[Zimluura]

of course Antler, platform specific libraries will increase difficulty in this sort of thing. but there are cross platform multimedia libraries now.

iirc, the quake3arena demo was available for mac and linux before it was available for windows. i don't think they sell a linux version anymore, but they have binary downloads available for free on their ftp site (you need to buy a windows cd, but you just copy the media files and dl the executable).

and i *thought* linux patches have come out for ut2k4 only a few days after the windows patches (which could be due to qa latency).

i'm not really saying that going out of your way for linux support is a good commercial move. but i think mac support is. and with the cross platform libraries that are out now, you may be able to do win32 & mac and get linux support for free.

i'm not certain about cross platform networking libraries, but i'd be surprised if none existed.

[Charlie_VFP]

Agree or disagree, they are all opinions. With time and money, everything is possible.

My only programming experiance lies in Windows based environments with C++. However, and the last poster stole my thunder :), that there are different libraries for each compiler. Even on the Windows only platform, something done with a Microsoft compiler or a Borland compiler will have discrepancies in the library class and function calls.

I am not very well versed in linux myself. I have played with it quite a bit and found that I like Windows better. Just personal preference. That being said, have they ever figured out how to get DirectX games to run nativly in Linux? I just checked up on the few projects trying to, and on sourceforge, and found nothing. Seeing as how the core of the graphics runs on DirectX, this could be a serious problem in porting the game to linux or Mac graphics wise.

Look at XPlane.... its graphics cannot compare to MSFS2004. However, it is still a good sim. Why? Realism. But its sales are no where near MSFS2004 either.

Good at all, Master of None is a very appropriete term for this. The resources, time, and money are no where near where they will need to be to make Fighter Ops a top of the line sim, as we all expect it to be from the information they have released so far. Seeing as how G2I is already relying on outside coders, I cannot imagine them trying to bring in so many that they can port it across 3 different OS's and keep it at the same level on all.

Even with all their money and coders, I do not think Microsoft would try a project like that (all of the politics of Linux vs Windows aside).

The cards all point to one thing in my mind: Use all available resources to make Fighter Ops the best it can be on the Windows platform. Until Linux becomes more graphics / game freindly, stay away.

I would think they would port to Mac before they would port to Linux. More users in the Mac arena.

Again, just opinions.

[Zimluura]

Terry, there may be some general mis-understandings here.

i think maintaining 3 separate code bases is definately a mistake. the idea in not so much to port code but instead to build a single codebase that will compile on a variety of platforms with a variety of compilers.

if using directx on windows (and opengl for everything else) then i think the "good at all, master of none" analogy is quite accurate. but if using opengl on windows too then there wouldn't be 2 separate renderers to work with.

i was under the impression, perhaps wrongly, that with fighter ops the developers were starting without any legacy code from falcon, and also that they were still very early into development, where they could still decide to use opengl. please correct me if i'm wrong. if they're 1 week away from previewing their d3d9 graphics engine then it's probably too late to look at cross platform development.

and for the record, i'm not very well versed with linux either :)

[Vlerkies]

It won't be too bad if it moves across platforms later down the line, but not for the initial release version

:drink:


:military:


~S~

[Charlie_VFP]

The devs stated they already have an 'extensive' code base.

If you use OpenGL for everything, then you're going to hurt the Windows users and DirectX fans. I have an ATI 9800 Pro AIW card and it runs DirectX much better than OpenGL. When I ran a GForce 4 and a GeForce3 card, they both ran DirectX better.

I am not debating which is better, simply what has always worked better for me.

I do not think it is even possible to write code that will compile over 3 OS's, let alone two. Well, let me re-phrase.... I do not think it is possible to write a sim that will.

Unless you're dealing with a very basic set of code, you are going to have to port it. The more code, the more time it takes to port.

Download some simple programs writen in C on Linux and try to compile them in Windows. Won't happen.

Too much time, too much money, and too much hassel.

[a.m.i]

I want to see a linux version that can be runned as dedicated server...

[Zimluura]

Terry, again my experiences differ from yours.

for me opengl has always been faster and always (allot) more stable on win32 than d3d. i have limited experience with ati cards, but they seem to handle ogl ok. now, i think we all want to skirt the 3d graphics api wars here. i am, like you, stating which one has worked better for me.

also, a year ago i wrote a 3d engine with a friend. and we got the same sourcecode to compile on both linux w/ gcc, and windows w/ vc6. never had any library problems, and only occasionally ran into differences in relaxed syntax for each compiler. it's true that our program wasn't as detailed as a sim, but we had all the graphics, input, event-handling, and file-io ground work laid out, so the 2 areas left were sound (covered in the sdl lib we used, but not personally tested by me), and networking. all the additional details of a sim would be with higher level stuff that shouldn't interface directly with the os apis anyway.

what i'm really trying to get across here is that if you focus on designing your software for more than one platform early in the project's life, you can stave off non-standard syntax and design mistakes that will make you fork your source tree to support another os (a *bad* idea, that's where prices will go crazy)

if the devs already have allot of code done it would probably be difficult to rope it into a cross platform design. but it really depends on how much code interfaces with the os without a solid abstraction layer.

<phew> long post.

[Charlie_VFP]

OK, running with that train of thought and just assuming this to be correct...

Is it really worth their time and effort for the small base of linux users?

I am inclined to say no or else others would have done it already... those with more time, money, and resources.

Just no need. The core players out there use Windows. The second highest group uses Macs.

Simply not worth it regardless of how hard / easy it is. If it is easy to code for both, then it would take twice the testing.

Just not worth it.

[Vlerkies]

And just to add to Terry's post. If they do make a linux version, who is going to Beta Test it? Like Terry said most simmers use either windows or macintosh. It is going to be very hard to find enough linux users for Beta Testing :thumbsup:


:drink: :drink:


:military:

~S~

[Zimluura]

totally agree, the demand for a linux version isn't enough to warrant much beta testing (only to make sure the server component works well), or much in the way of technical support.

the safest way to do it would be, at release, to put the linux (and maybe even macos as well) executables on the site for download as an unsupported beta version, and setup a community support forum for them. so users can help each other troubleshoot.

if it consistently passes all tests there might be a way to roll it into the next cd press and update the box art with macos and linux logos. to more actively advertise alternative os support.

i firmly agree that the win32 platform it the most important, then macos, and far behind that is linux. but i think macos support is probably worth the effort to heavily abstract the codebase from the os. at which point running on linux *may* (depending on compiler and lib choices) become a freebie.

[Charlie_VFP]

I think that the entire beta team for the linux version would be every person who would use a linux version. Which means there would be no point in them buying the final version. :)

Im sorry, but I do not even see Mac support as being a worthwile endevor. A good friend of mine is a graphic artist and her whole company swears by macs when it comes to graphics, video, and sound editing. For everything else..... well... I am sure you get my drift. :)

Why divide resources amongst different OS's? Why not put all resources into the Windows version and make it the very best it can be? This is especially important for a small company that has very little resources, and no funding, to begin with.

[Vlerkies]

That can also be very true Terry

:drink: :drink: :drink:


:military:


~S~

[Zimluura]

Terry, you misunderstand.

for the method i'm describing you'd need to buy the win32 version to beta test the linux version and the latest executables could be freely available on the site (as they essentially always are anyway with patches, etc.). the media (models, textures, table-data) would be something you have to pay for, but not something you have to buy more than once (because media is already crossplatform). the only point in putting them onto the next cd press would be to advetise full support should the beta test succeed. because some people don't have windows at all, and won't buy it if they're not sure it will work on their os.

as for resource division i stand by what i've said before, that i think the windows version would benefit from the support of other oses. i'm sure resources are limited, and it would take some resources get it set up. but i really think that with the inherent complexity of a such a detailed simulation it's the best idea financially to keep the codebase well abstracted from the os. you can even make a case for that based on previous windows upgrades that have caused some games to break. and if you abstract it well enough you can get it to run on other platforms at little cost.

if they decide to support other oses during development or after, the key is to get it all to work from the same sourcecode. allot of time ports are outsourced to another company, which forks the sourcetree, and doubles the cost of maintaining things. that is the kind of thing that shouldn't be done by a small company.

[Charlie_VFP]

Well, we are obviously never going to agree.

I think that it is a bad idea to make the sourcecode to run on all the OSes and waste time doing such.

While it would be a great thing in the "ideal model" world, there is no such thing. We live in a world right now where Windows dominates the market in both OS and game design tools. I would also venture to say that 99% of the people reading this post are doing so on a Windows machine, and if they are not, then they at least own a windows machine as well as whatever they are on.

It is a bad idea, if not downright looney, to divide your resources in such a manner. There is a LOT more things going on than just the code. There is all the logistics of the matter; testing, tweaking, integrating, and having someone dedicated to keeping track of it all. Where as you keep it to one OS, then the complex picture gets much simpler.

We will apparently never agree so there is no need to keep debating back and forth saying the exact same things over and over again.

[Napalmski]

Bumped into this thread rather late but here's my $0.02. If I look at the industry I see that every FPS serious about online gaming has a dedicated linux server. That ought to say something.

If Fighter Ops wants to encourage online gaming on persistent worlds they might want to look into this. Preferably also with a setup which would allow running a dedicated server and a client behind one NAT firewall (Like it was done for Falcon 4 with the F4mp5-1 patch.)

I have a 2048/1024 ADSL connection and I wouldn't mind running a 24/7 linux dedicated server. Preferably one which can be administrated remotely through a web or telnet interface cause it will run on a headless server.

About the porting stuff: OpenGL & performance. As Doom 3 is coded in OpenGL, I think that's 'nuff said. Secondly for UT2004 there is ONE guy who provides the linux port so it can't be that much work if you do your code right.

From what I have understood porting between windows and linux shouldn't be too difficult if you keep some things in mind:

1) stay away from the MFC and DirectX stuff. There are lot's of cross platform toolkits which can replace MFC and if you're coding a dedicated server you don't need DirectX stuff. If you do need the 3 D world you can use OpenGL and OpenAL (as shown by UT2004).

2) use C/C++ code which can be compiled by GCC. This code can be compiled without problems by any other C/C++ compiler

Sure it wil cost resources, the benefits depend on how important you think it is that the community can run its own dedicated servers. I don't think many people are going to buy an extra windows licence to run a Falcon or Fighter Ops dedicated server.

Nap

[Charlie_VFP]

Dedicated servers run just fine on Windows too.

[Dave_VFP]

Time to bring out the Commodore 64 disclaimer again :)

[jhook]

Yes,
I would like to say that any other version of "Fighter Ops", should be developed by a seperate team. A team that is dedicated to there systems and would not draw on the primary team for support/data edits.

[Bladehawk]

Porting everything to a second platform will take away time and resources that could be spent making a single platform operate better.

I voted no.

Exactly my reason to vote no ON Linux and Mac version or any other (refer to mac pool The reason to vote no is the same...) :nono:

[krazyscotsman]

Terry, you misunderstand.

for the method i'm describing you'd need to buy the win32 version to beta test the linux version and the latest executables could be freely available on the site (as they essentially always are anyway with patches, etc.). the media (models, textures, table-data) would be something you have to pay for, but not something you have to buy more than once (because media is already crossplatform). the only point in putting them onto the next cd press would be to advetise full support should the beta test succeed. because some people don't have windows at all, and won't buy it if they're not sure it will work on their os.

as for resource division i stand by what i've said before, that i think the windows version would benefit from the support of other oses. i'm sure resources are limited, and it would take some resources get it set up. but i really think that with the inherent complexity of a such a detailed simulation it's the best idea financially to keep the codebase well abstracted from the os. you can even make a case for that based on previous windows upgrades that have caused some games to break. and if you abstract it well enough you can get it to run on other platforms at little cost.

if they decide to support other oses during development or after, the key is to get it all to work from the same sourcecode. allot of time ports are outsourced to another company, which forks the sourcetree, and doubles the cost of maintaining things. that is the kind of thing that shouldn't be done by a small company.

I think you need to make the move to Project Phoenix. We are planning to port to linux. Our graphics engine is written for both Windows and Linux (OGRE). We are open source, so no we are not trying to make a profit. We have no time tables so....

http://www.phoenixosfs.org - visit the forum

Krazyscotsman

[Playloud]

I think a Linux release would slow down development on the Windows version. There should only be one version. While I would be willing to install Linux, most people who would see this game in the store would not. Keep it simple, and create a single version that everybody can use. Windows XP Pro works fine for me. If Microsoft starts charging a subscription for using Windows, then maybe we can talk about porting FO to Linux.

[Charlie_VFP]

I highly doubt Windows would do that. It was probably a little suggestion on a "I wonder what would happen if we...." and after being leaked to the press, it is suddenly a huge deal.

Sort of like Bill Gates saying that one way to fight spam is to charge one cent for every email. Suddenly, every one was talking about how that was going to happen and what it meant.

I wouldnt worry too much until you see it happen.

[Stuntie]

Firstly, a dedicated server can be linux or windows.

You only need to handle the data, and any associated checks. You don't need to be able to run the game as a user. The data goes via standard network protocols, so creating a linux server should not be an major task.


Secondly, FO running on Linux - I would say No. At least not by the Devs.

I'm a programmer so I do have an idea of the kind of issues involved from the coding point of view. (Borland C++ Builder 6 is my main IDE at the moment, but I have used a number of IDE's for a number of languages).

It would be a major task in both time and effort for a minute number of potential customers.
If a team can be got together to work on a linux port AFTER the Windows release, then so be it, and good luck - you'll need it.


Increased abstraction = decreased speed as you need to go through an abstraction layer to get to the API calls that you could have made direct.

Not using the MFC - rather you than me. The MFC was added to get around the fact that the Windows API sucked. The MFC calls are invariably simpler and clearer to understand.

Using a Gnu compiler - I would hate to do a major job on one of them!
They are invariably lacking the advanced tools and features that start to come into play on big jobs like a sim. This is not 'hello world' or a basic text editor.

DirectX vs OpenGl. Don't want to get into that debate as it comes down to personal preferences, and individual system setups. But I will point out that Doom 3, UT2004 and other OpenGl games are often ones that are planned for a variety of platforms such as Windows PC, Playstation, Nintendo whatever etc. etc. I.e. Ease of porting to non-pc based systems is a significant factor in development.
How many of these have actually come out on linux though?

And if there is a linux version, then there will need to be a Mac version as well. That means three lots of code, three lots of testing, three lots of patches, three lots of ...... etc. etc.
Lets just stick with one for now.

Quick pause for a drink: :alcohol:

Having said all that it's no doubt irrelevant already. The dev's will have been working on the code for a while, and they will know if it's possible to tweak it for linux, or whether thats just too much hassle and too much code that would need rewritting for them to even contemplate it. The clocks ticking and there is a product to get out. Let's just leave them to it rather than have them jumping through hoops just to please a few, who would have run it on Windows anyway.

[Bladehawk]

VERY... VERY well said Stuntie :clap:

[Napalmski]

I'm not interested in turning this thread into a windows vs linux war so this will be my last post on this subject.

First, (to Terry) it's not the point that a dedicated server can run on windows, the point is that I and a lot of other people are not going to buy an extra licence to do so. Every online game which has both a windows and a linux dedicated server (Quake, Halflife/CounterStrike, UT2004, Tribes 2 , Savage to name a few), has the overwhelming majority of servers running on linux. You can draw your own conclusions from that.

Secondly (stuntie) GCC is a compiler not an IDE. There are loads of professional quality IDE's which support GCC: Eclipse, Kdevelop, Visual X-tools to name a few.

Thirdly there are good replacements for MFC, QT is a good example.

But maybe we should ask Austin Meyer (the guy behind X-plane, the simulation engine for fighter ops) he has ported part of this work (space combat) already to linux. He's the expert.

It all comes down to how you setup your code, if you lock yourself in right from the start with MFC and DirectX, porting the code is God's own work. If you use cross platform tools like QT, OpenGL, SDL, OpenAL, it's much simpler. I think I have given enough examples to remove the myth that cross platform toolkits give lower quality games.

Nap

P.s (to Terry again) Software Assurance for corporate customers is already a subscription licence model. You can bet your a$$ that MS right now is figuring how to apply it to the consumer market.

[Bladehawk]

Dear Napalmski, all your arguments cam be very true but that will not change the fact that to FO have the "constant" grow and Hyper Realism that is announced the team must be directed 101% to one platform, and don´t forget that PC means W98,Millenium,2000, XP etc... they seems to look a like but they are not !
Like Terry_VFP well said...
"Porting everything to a second platform will take away time and resources that could be spent making a single platform operate better."
The word better = :clap:
Porting = :nono:

[StretchG5]

Yeah, but which is going to be more profitable? Bringing a very good (but not perfect) flight sim to platforms and userbases that don't see too many flight sims of even medium quality, or making a terrific flight sim on one platform that already holds the lion's share of flight sims anyway? No one can say for sure, of course, but my 2 cents says porting will be profitable.

And hey, it means that you guys are nice enough to throw us a bone once and a while, eh...

[Zimluura]

________________
Stuntie wrote:
Increased abstraction = decreased speed as you need to go through an abstraction layer to get to the API calls that you could have made direct.
________________


i can't see how anyone could code a project as complex as a flightsim without having it riddled with abstraction layers. coding a big project any other way would result in a bloated, buggy, and terribly difficult to maintain source-code.

i suppose, if you're really worried that the one extra abstraction layer from the os function calls is the straw that breaks the camel's back you could always "inline" said functions.

i don't have a very heavy background in mfc, only have written one program in it, but uhh, are you sure mfc is a good idea for a game? i mean, the model importer/rigger, sure. the terrain editor, again, sounds fine. but performance critical graphics engine? that raises (and i'll be the first to admit, perhaps incorrectly) a red flag for me.

in any case I think we all can agree that:

a) the developers should spend a little time thinking if a linux, mac, ps2, amiga, or c64 version is a good move.

and b) that they should be the ones to make all such decisions.

i think the cross platform library support is there now (it wasn't there in 2000, but now i think it is), others disagree. i'm not optimistic about other os versions for this product, though if the xplane code was delivered with interfaces to more than one os it would be nice if the developers don't go out of their way to break it. i can only imagine why the medal of honor linux port took soooooo long.

[maverick6664]

I vote for Linux version. But it will take time, so I will buy both as a result.

1. Probably it will have no (or much less) stutterings. I guess we have stutterings because multi-processing code in Windows isn't mature for Falcon4 especially bms. As you see in other applications, if you run multiple applications on Windows xp at once, by default the foreground application runs fast, and "intermittently" background applications run and at that time the foreground application halts. On linux, background processes with low priorities (high nice values) also run smoothly as the process with high priority runs smoothly. In Falcon4, there seem to be 2 or more processes(though in task manager F4-BMS is shown as one process. I'm talking about the analogy); one for rendering graphics, others for calculation/sound/etc. And I guess stuttering happens on Windows when this background things are busy, because the multiprocessing code of Windows kernel isn't nice enough to handle this situation. In linux it might reduce FPS, but wouldn't cause stutterings, I think. I guess (yes, just a guess) the timeslice of each process is larger in Windows than in Linux and it causes stutterings.

As an experiment, try to run a background application such as seti@home(see setiathome.ssl.berkeley.edu) with Falcon4. It doesn't reduce FPS of Falcon4, but it causes micro/macro pauses in Falcon. It means process-switching of Windows is immature. I used to run flightgear(flight sim for Linux or any other OpenGL OS's) on Linux with seti@home, but I've never had such pauses or stutterings at all. Instead FPS has been lowered very slightly(yes, it depends on nice value.) usually 3-5 percent;negligible.

So I expect no (or much less) stutterings with Linux version.

2. Linux is free. It is not so popular as WIndows yet, but many people are using Linux. By Fighterops for Linux, the number of Linux users will increase :smile:

3. If Linux version is made, it should be easy to port fighterops to other OS's with OpenGL. Market will be large :wink2:

4. Above all, I will not have to run Windows just for FighterOps :evils:

[Charlie_VFP]

Same old arguements in more flowery words. Nothing changes. And I really have to dig up the motivation to continue this arguement as it is another one of those that each side will continue to say the samn damn thing over and over and never give an inch. So, why even continue to bother debating it?

The facts still stand.

Porting ANY program to multiple OS's takes resources. G2I is working with limited resources. Enough said.

Games that are forseen, prior to the first bit of code being written, to work on multiple platforms, mainly speaking of porting to PS2, XBox, ect), are easier to do so. FO already has an extensive code base. The developers are making their own code for flight models, weapons systems, ect. They are doing it in C++. Who is to say they are working with Windows based utillities? I would bet a cookie they are, but dont know for sure. IF FO was forseen to run on mutliple platforms, we would know it by now.

Look at the number of linux users. Look at the number of Windows users. Look at the number of Mac users. Look at the number of simulations being written. Look at how many are written for Windows only. Look at the prefered OS for simulation users. Needless to say, appealing to a linux crowd will do nothing for the sales of FO. There would be a MAC version of FO long before a linux version, but even those sales wouldnt even come close to comparing to those of Windows. Dont beleive me? Then tell me why all the games that require a thought process (not doom, not quake) are all made for Windows only?

This is not an opening to hear the horse apples comments about how First Person Shooters and 'wanna be role playing games' take thought processes.

A dedicated server, as stuntie said, requires no graphics. It will run just as well on Windows as it will on linux or mac if it is programmed correctly. So, there is NO, again NO, advantage to running a dedicated on a linux machine.

And as for the comment about FO on linux would not have as much stuttering as windows.... :poundit:

[DEX]

I don't know if a Linux version would be practical. From a profitability standpoint, I'd have to go with no. The market would be somewhat larger w/ 'nix users but not THAT much larger. A 'nix version DOES have atleast one interesting implication -


This sim uses lots of processing power and does lots of multitasking. Linux excels at this stuff. In addition, 'nix is very very friendly towards multiple processors and networking. Imagine having one box dedicated to the AI, another to sensors [radar, thermal, visual, AI/player], etc, all communicating over a LAN with UDP or TCP (LAN mind you, not internet) with your box, which only has to worry about sound/graphics. Performance would be other-worldly. Correct me if I'm wrong, but from what I understand, XP doesn't even support more than one processor.

Just a thought.

[Zimluura]

___________________
Terry_VFP: A dedicated server, as stuntie said, requires no graphics. It will run just as well on Windows as it will on linux or mac if it is programmed correctly. So, there is NO, again NO, advantage to running a dedicated on a linux machine.
__________________


that's a bit off man, the reason why you see linux dedicated servers show up for games without linux clients is due to linux being widely seen as more secure and stable than windows. many people who work at ms would debate this. but if the ms guys are wrong then linux would simply have a better presense of servers on the internet than windows...oh wait. well anyways the reason is that a linux dedicated server drastically increase the amount of dedicated servers that are available for said game on the internet. the licensing issue is probably a bit of a concearn too (each copy of windows does cost allot, and they have a ludicrous per cpu licensing scheme in smp arenas).

it sorta seems like, with this threads resurrection, that things are getting added to the topic again. or that it's at least somewhat informative. but i'm sure no one will think less of you if you feel you don't have anything interesting to add. :)


ok anyways, to my knowledge no one has tried to release a commercial flightsim for linux. so we really can't draw any really solid conclusions about the commercial viability of such a move. there are two things which i think are most worth noting on the uncertainties involved in a hypothetical project like this.

1) in general linux users can handle steeper learning curves than windows users. thus, in some respects they *could* really be the best target audience perhaps more profittable than windows. well i'd be ready to sign if it weren't for this next one.

2) i would guess that allot of linux users don't ever play games and have no interest in ever doing so. until i knew the compliment of this percent i'd be a cautious about expending too many resources for a linux executable.

[StretchG5]

Terry, saying that all games with a thought-process come out only for Windows is absurd. You're denying that -- on the Linux side, games like Homeworld and FlightGear, and on the Mac side, games like Falcon 4 and Myst (which even came out for Mac first) -- have no thought process?

The number of thinking games that get ported to Linux (or other platforms, since you said "Windows only") vs. the number that come out for Windows alone is pretty much the same as the ratio for point-and-shoot games.

[Charlie_VFP]

Stretch: The numbers dont lie. Look it up.

Zim:

Maybe the fact that no real games have come out for Linux is a sign...

Besides... many webhosts run linux for one single reason.... it is free. You can argue security and stability all you want. The fact of the matter is that it is free.

I would almost agree with you on the security issue. Almost. The media will lead you to beleive what they want you to beleive. Since Microsoft is king crap on the market, it is a better story to talk about Windows problems than Linux.

Furthermore, Linix's learning curve would scare away many potential customers. Who wants to learn how to go through all the garbage associated with Linux to get what they want done? Dont get me wrong, I have used Linux and my general nature has my interest tweaked on it. I just do not have the time to really mess with it yet.

However, every company I have worked with (including the military) has built their network backbones on Windows technology from the beginning. If Linux was really the 'All That' as you claim, then those companies would have switched over long ago.

But this isnt a Windows Vs. Linux thread. It is a "To port or not to port" thread. And frankly, in a discussion like this... stability, security, and ability does not come into play. Only one thing comes into play....

HOW MUCH MONEY WILL IT MAKE?

The answer is, sadly, very very little. So little that it will not be worth it.

[Zimluura]

Terry:
real games have come out for linux (id games, unreal, neverwinter nights, and you should never forget nethack :) ), but for the sake of hyperbole you're correct. it's not allot.

if a linux port means any net gain at all it would be worth it. the worry is that it would be a loss. but i guess that's what you meant.

the fact that there are allot of linux servers out there on the internet (for whatever reasons) is the motivator for making a linux dedicated server. it's seen as commercially viable because it enhances the windows version.

i, like you, am pretty damn sure a linux client wouldn't make as much as the windows version. but still, it *could* make enough to justify doing it. if the fighterops team were to make their finances and sourcecode public, we'd be able to say something about the worthiness of deciding to make a port. but they won't (and they shouldn't).

without looking at the sourcecode any arguments we could make about the money required are baseless. if you have access to the source you may be in a better position than i am to guess.

[Charlie_VFP]

No need to even look at the source. Arguements are all in above posts.

It all boils down to money. Plain and simple. Because of Bill Gates, Windows has the majority by a long shot. Not just by a little... but by a long shot. Because of this, the market for the PC is much higher and well beyond anything else; mac and linux included.

Most people that run a linux box do it because they are the type of person that likes to hack through things and learn. Windows is a simple point and click so any person of any intelligence level can use it. It is a fact of life.

Even if it were possible and not that hard to do, I would rather those resources go towards simulator related projects.... such as the add-ons and such, and not into getting it to run on a second or third OS.

I run Windows... so I am biased. A Linux version will not benifit me at all. With that aside, I have given as many facts and details as I can in my arguement to avoid being biased.

And that is all it boils down too.

[Bladehawk]

well said :clap:

[Zimluura]

Terry, man, you seem so negative sometimes. like, lets just say, hypothetically, that the choices for this poll were

1) no, even if support for other systems is easily achivable.
2) sure, provided it doesn't hurt the windows version.
3) yes, i won't be able to play without linux support.

now, from your stance thus-far i'd guess you would pick 1. puh-lease tell me i'm wrong on this. i mean i see that you have good, justifyable reasons for believing alternative os support is a bad move, but you seem to want to quash any discussion as to how, why, and under what circumstances it might be beneficial.

[Stuntie]

A better way of restructuring the poll that would help represent the view of people like Terry and me would be:

Create a linux version:

1) only if it took very little time or effort and did not cause features to be dropped.
2) If it took quite a bit of time and effort, but we could rertain all features.
3) Even If it took at lot of time and effort, with or without particular features being dropped.
4) Create one no matter what time and effort was required.

This adresses the two major worries that some of us have 'time and effort' and 'features'.

Most would vote for number 1. Some for 2, but very very few for 3 or 4.

Number 1 - is that possible?
It is not an easy job to port a complex system over to another, and especially one development is already well underway.
You can say 'use this API don't use that API' and talk about GCC++ but try doing it in reality - it's a minefield that will take time and effort away from other things. And trust me it will not migrate smoothley - nothing ever does.
And will all the features get in? What happens if a particular features makes extensive use of propriety APIs that are not available for Linux? Will the devs write new ones or convert to a different set of APIs - all very time consuming, or will they just drop the feature? What would you do?

So moving on to the other options we need to decide if they are worth pursuing.
I would say no. At least not at the moment.
There are too few Linux users who would be interested in FO to justify the work involved, and any potential problems with features.
I'm not being against Linux - I'm just being realistic. linux is a small market at the moment. And most flight sim fans have windows PCs for simming on - so they can be catered for already. If we port a sim to one neiche market - i.e. linux, then the inevitable cries will come from other neiche markets such as Mac's.

G2i needs to make money to survive, and there is just not enough money in a linux to cover the costs that would be incurred.
When FO isl aunched and selling, then there may be time to look at the feasability of Linux and Mac versions.

[Charlie_VFP]

Zim: Why is it that when someone disagree's, they are being negative? I have disagreed from the beginning but my posts have always continued details, sound logic, and attempts to explain my position in depth. That is not being negative. If you think everyone who disagree's with you is being negative, you have a very rough time coming.

Stuntie has nailed it once again. If it were possible to not take away ANY resources from the things already planned for FO and there will be NO negative impact to the planned goals of FO, then I would say it would be ok for other OS development. I still wouldnt use it.... but it would be OK.

However, *I* do not beleive it to be possible based on my experiance. I do know that there may be something that I do not know about. If that is the place and that knowledge becomes available to me, I will re-evalute my position on this topic (as I did with the civillian traffic topic). But until NEW information becomes available that shows there is a flaw in the information I currently have, then I will not.

Disagreement is not negativity. Trolling is negativity.

[Aragorn]

Bah - Humbug..... :poundit:



It'll take MORE coders to MAKE a Linux version than the GLOBAL TOTAL number of Linux users who'll FLY it...


Niche groups....


Can't live WITH 'em; Can't kill 'em..... :guns:



The Gorn

[Zimluura]

*seem* negative, *seem*. that post was all about trying to figure out your tone.

i would say that you still *seem* to have some hostility as the line:
"If you think everyone who disagree's with you is being negative, you have a very rough time coming."

is pretty condescending.

i mean, you're certainly not as openly hostile as Aragorn here, but he might have just been trying to lighten the mood.

and as for lightening the mood, what's up with the "pound it" icon? looks like...well...i won't say what it looks like an invitation(?) to do... but please someone explain.

[Charlie_VFP]

I'm not even close to hostile. :bigsmile: Nor would I become hostile here in a professional compacitance.

Take my posts at face value. There are no hidden meanings or tones.

(I was going to explain more in a PM to you, but you have your email and PM's turned off. Not very professional in public.)

[Zimluura]

turned off email, because of worries about spam. had no idea my pms were turned off.

[Charlie_VFP]

God, I know the feeling of spam. I tried a few new spam block programs and server side applications but they were all annoying. They worked good.... but I lost alot of good emails along with the spam. Truely annoying that SPAM thing.

[Aragorn]

SPAMMERS.

As per Linux users - can't live WITH 'em; Can't KILL 'em.





:gleam:

[StretchG5]

Why exactly can you not live with minority users? Not all of us think Windows is the greatest thing since sliced bread... Plus, you guys seem to be forgetting that most people who would port to another platform would be willing to work for free (i.e., for the good of their platform).

[Charlie_VFP]

Still going.... nothing out lives the Linux thread.....

<Fighter Ops Pacman glides by beating a drum>

[jhook]

As I said before, I woudn't see a problem with platform extentions, as long as you have a seperate group supporting that platform. The current team is buisy enough with all of the features this sim will have. If independant teams (independant from G2I and Fighter Ops) spends there time with there platform, working the code (under supervision and legal stipulations) could create an extension code for Mac and other platforms. As long as it would not take away time needed for the main frame release, I see no problem with it.

[longshot]

If you don't make a linux version, then test it with transgamings software under linux. At least that way those of us who use linux and not M$ can fly what looks to be one heck of a sim. I've been waiting for a replacement for Falon4.

Longshot

[Totose]

hi all

yes! :)

I use Linux for many other games but running Win2K just for Falcon4.0. :/
Win*** su** :p

[longshot]

One thing M$ is saying on their longhorn forum is that longhorn, when ever M$ gets it out the door, bugs and holes, is that may programs that currently work will be broken. At work they are dreading when machines come forced loaded with longhorn because the majority of our programs will be broken.

This will apply more to games.

2 cents worth.

BTW.. I do spend money for linux games and most run better under linux than windows.

[SUBS17]

I'm going to avoid getting that OS until I really have to get it.

[shamandgg]

I'd love to have a combat sim on Linux. If it happens so.. I'd be a happy man.
I'm sure a lot of Linux users will be. Windows users are being very selfish here saying they want their version to be released faster - so they vote for no Linux binaries at all.
I thought you wanted to have a bigger and better FighterOps community? You really should vote on Linux version.
I actually follow FighterOps project, because I still hope it will be working on Linux (as Xplane does). If it will not, I think I might skip it.

[jamse101]

Voted yes. I'm downloading X-Plane 8.16 Linux version right now, I compiled kernel 2.6.13.4 today and got latest ATI driver and I just want to try X-Plane now with all graphics settings set to "INSANE" :D

[svetron]

I'm a Windows user, but if the developers are making it compatible with Linux, and this means it'll take a little longer before release then I'll wait ! I want as many simmers to join in on the fun as possible. If this sim was Linux only then I would change to Linux, NO WAY I'm going to skip this sim !!!!!!!!!

[Buckshot]

While I definately agree a Linux version would be great, the Windows version is going to get total focus at the moment (although it is likely a linux dedicated server will be produced with or soon after the first release).

Windows is going to be our main income stream for obvious reasons, so we need to establish that first before we can branch out into full Linux and Mac versions.

[krazyscotsman]

Smart thinking about going after the windows revenue at least to get established before porting. After all, I believe most major game companies that do release linux ports initially release in windows.

Krazy