Witchers, what about Linux?

[volsung]

@Gilrond,

What Guy N'wah and Gregski said about making money and ethics (respectively) does not necessarily represent their personal views, they are explaining how the game (and in general) industry works. A CEO and any company in general are investors, they want their money to generate more money and couldn't care less about ethics or moral values. If a project can't guarantee a large enough profit, they may not be interested in it. This is the sad truth, regardless of whether GNU/Linux is better suited for some tasks or not.

Now I do agree with you that it is hard to count Linux users, and especially those interested in gaming. Perhaps if somebody ran a worldwide poll?

In my line of work (scientific computing, research, computational science) almost everybody uses GNU/Linux, in their workstations or in a computing cluster. I *know* there are many high-end computers in people's offices and homes running a distribution of GNU/Linux, and they would probably buy AAA games if they were offered to them. I'm talking about computers that could run The Witcher 2 on Ultra minus Ubersampling in many cases, and in some even SLI configurations.

 

[Gilrond-i-Virdan]

2% is the high end of estimates; 1.4 to 1.6% is closer. It's not based on licenses sold, so that is not a foundation for believing it to be an undercount. It's based on Web activity, particularly identification of browsers used to hit counted Web sites.

The actual Linux user base is much larger, but that counts all those servers that aren't used for running desktop applications or games.

There is nothing in business ethics that requires you to lose money on a project or to serve a market that cannot offer a return on investment. Linux has to be a profitable market for game development, or it is not a market for game development.

My point was that 2% of desktop estimation is off as well (too low). I wasn't even talking about server installations, where Linux is a clear majority.

I didn't say that ethics require one not to make profit. See above, I explicitly mention that company should be sustainable. I was commenting on your maxima statement that "everything you do has to be for the purpose of making money." That's simply incorrect.

 

[GuyNwah]

My point was that 2% of desktop estimation is off as well (too low). I wasn't even talking about server installations, where Linux is a clear majority.

I didn't say that ethics require one not to make profit. See above, I explicitly mention that company should be sustainable. I was commenting on your maxima statement that "everything you do has to be for the purpose of making money." That's simply incorrect.

"Sustainable" and "profitable" are not, as it appears you would have them be, different.

A company organized for profit is "sustainable" only if it makes a profit. Any activity of that company that uses capital or labor must be for the ultimate purpose of making a profit. Proposed projects are evaluated on the forecast profit from the activity and the soundness of that forecast. That's not my opinion. I don't like it either. There are projects I would love to pursue that wouldn't make money for my company. But I can't, because it's a cruel fact.

What not-for-profit companies like the Mozilla Foundation and KDE Foundation believe is edifying but irrelevant. It does not change my conclusion: games on Linux must be a predictably profitable market, or it is not a market that for-profit companies will sell to.

 

[Licaon_Kter]

Do we have any data on this one, besides, you know, your words? How big is the demand? What opportunities are there for a company? How big is the market?

3 hints:
*Humble Bundle (average Linux price paid was always way above Mac/Windows)
*Steam (ITS COMING!!!1111oneone)
*Kickstarter (latest hits got Linux versions because the buyers demanded it)

There were efforts made to implement DX10, but they were abandoned years ago.

umm not quite true, DX9 is there already yes, but DX10/11 is handled as a whole and still implemented slowly. Yes it's behind, but the DX10/11-only game list is pretty much short ( keep an eye on the key word here: ONLY )

@Gilrond: not sure where they announced it, but, like trust me, it's native OpenGL, not WINE for TW2

 

[gregski]

In my line of work (scientific computing, research, computational science) almost everybody uses GNU/Linux, in their workstations or in a computing cluster.

And isn't Linux used mostly by people from "your line of work"? I mean, tech-savvy, aware, power-users? Which might be as well 0.58% of the whole gaming population?

3 hints:
*Humble Bundle (average Linux price paid was always way above Mac/Windows)

Average price might have been higher, but if you broke down the numbers by quantity, I bet the Windows one would be the highest.

*Steam (ITS COMING!!!1111oneone)

We'll see how this turns out and if it brings out any relevancy Linux might have in gaming.

*Kickstarter (latest hits got Linux versions because the buyers demanded it)

I would call it a vocal minority, nothing else.

 

[Gilrond-i-Virdan]

Average price might have been higher, but if you broke down the numbers by quantity, I bet the Windows one would be the highest.

The higher price paid indicates a demand that's not met with enough of supply. Opportunity for anyone interested. Simple economics.

I would call it a vocal minority, nothing else.

CDPR themselves are quite a minority in gaming industry (see their talk about black sheep approach). I wouldn't call it "nothing else". Minority can be not only vocal, but influential as well. There are millions of desktop Linux users, and it's a significant market, even if it's smaller than other OSes markets.

@Guy N'wah: As I said, if anyone thinks money is a self goal - that's already bad. Even for profit companies vary greatly in their values and approaches. From crooked and unethical, to decent and upright. Being decent doesn't require one to operate as non profit.

Not every activity is supposed to be "for money". Google for example gives their workers some time to pursue projects which interest them. Just for the sake of technology advancing and encouraging general technical curiosity. It might give Google some profit (if some of those projects will be something very significant and successful), but it as well might not. Still they don't consider it unworthy. It's just an example that not everything company does is supposed to be measured in money.

 

[GuyNwah]

I am puzzled by your view that a for-profit corporation exists for any purpose other than conducting a lawful business that turns a profit for the shareholders.

Companies that allow skunk-works projects are doing it for the good reason that they expect some of these projects to result in profitable new products or services. Not because it is rewarding to satisfy the curiosity of their employees.

Anyway, this is off the topic of whether a Linux port of TW3 is a suitable product. I still believe it would be a costly venture into an unproven market. Companies with a strong surplus of capital and few opportunities in their core business may be able to afford, or be compelled to attempt, that venture. I am not convinced that CDPR fits that model.

 

[Gilrond-i-Virdan]

It's not true that the market is unproven. Look at Humble Indie Bundle sales. Quite a visible proof. And the game of a caliber of Witcher series will not suffer from the lack of demand.

 

[volsung]

And isn't Linux used mostly by people from "your line of work"? I mean, tech-savvy, aware, power-users? Which might be as well 0.58% of the whole gaming population?



Average price might have been higher, but if you broke down the numbers by quantity, I bet the Windows one would be the highest.



We'll see how this turns out and if it brings out any relevancy Linux might have in gaming.



I would call it a vocal minority, nothing else.

I think your numerical approximations are still off, many average people use Linux. And I bet many more would want to get rid of Windows if they can run all their favorite applications. Now, regarding everything else I have to agree that, with a cool head, Linux is right now a potential but unproven market on which companies might not want to invest... yet.
You're such a ... business person!

 

[CostinRaz]

Humble Bundle is a far cry from a success. Linux has registered the smallest sales there in terms of overall cash. Sure it may have had higher numbers on average but that really doesn't mean much and it has made less then 20 million dollars overall, that's counting Windows, Mac and Linux together.

That's not a proven market.

 

[GuyNwah]

Humble Bundle is a far cry from a success. Linux has registered the smallest sales there in terms of overall cash. Sure it may have had higher numbers on average but that really doesn't mean much and it has made less then 20 million dollars overall, that's counting Windows, Mac and Linux together.

That's not a proven market.

Linux actually did quite well in the Humble Bundle campaigns, with something like 25% of the revenue, but it's important to understand that these were games with existing Linux ports, and the revenues from all the campaigns put together are only a fraction of what is needed to sustain the development of one title with AAA production values. It proves that there's a market for games on Linux. It doesn't prove anything about how much that market will sustain, and that's where I agree with you.

Even if Linux turned out to be a profitable market, I cannot find an answer to this question: If I could have a Linux port of TW3, or I could have something else CDPR made, would I rather have the Linux port or the something else? If the something else is a big new adventure for TW2, or the Cyberpunk game, a year sooner, I think I have to vote for the something else.

Companies that are resource-constrained have to consider opportunity cost. What do you give up by doing project L instead of project R? Even if project R is only delayed, you gave up some of its potential revenue.

 

[Gilrond-i-Virdan]

Don't forget, no one is suggesting for CDPR to target Linux alone. They have ways to fund the development for new OSes. Otherwise they won't be bothering to port anywhere (including Mac OSX and consoles).

The effort which goes into the system specific parts of development is only partial. The bulk of expenses falls onto the shared part (art, voice, actual engine development and etc.). OS adaptation of the flexible engine is still work, but it's not the most difficult part of development. Making the flexible engine itself should be the most involved task. And it benefits the whole development, since it allows porting it to many OSes easier (including may be console ones). So it's not even the question of Linux alone, it's about whether CDPR are interested in serious cross platform development already, or not. From what I observe - they are interested, but how good they'll pull it through - time will tell.

 

[GuyNwah]

The console port was supported by Microsoft's Xbox 360 development environment. It's not even a port in the traditional "make the software run on a dissimilar operating system" sense. Microsoft deliberately makes this manner of port as easy as possible, so long as you have not deviated from Microsoft's way of doing things. The Mac OS X port was supported by an operating system emulator. It does not share anything in common with the effort of porting to a dissimilar operating system. In short, the console port and Mac OS X port are not representative of the effort that would be required for a native Linux port.

While system-specific development effort is indeed only partial, that part can be very large, especially when the original development environment was a Microsoft one. Microsoft puts enormous effort into locking you into their development tools and their APIs. If you have significant experience with any Microsoft Visual Studio platform, this will not come as a surprise to you.

You made the assumption that CDPR has a flexible engine that is adaptable to a different OS with some reasonable ease. I don't know any more about the coding of their engine than anybody else who isn't one of their devs, but I would not be surprised to learn that that assumption did not have a solid foundation -- especially when it's evident from their employment announcements that they use Visual Studio heavily.

You also left out the costs of retooling and retraining. Visual Studio doesn't do flexible or cross-platform. All your developers will need new tools and will need to come up to full productivity on a cross-platform development environment that they may never have used.

In short, I don't share your optimism that the cost of developing a largely new flexible engine on a new development platform will be minor, and I still believe that the only value in it is the ability to capture the additional market. If that market is no more than 10% of the Windows + Xbox 360 market, the cost of entering it will have to be considerably lower than I believe it is.

 

[Gilrond-i-Virdan]

I got an impression that they are working on such engine from their interest in Mac OSX port. Using Wine doesn't sound the same to me. Also they did mention something about Playstation experience being a plus for their potential hires. That's already a different OS from the Windows walled garden. They might have developers for adaptations to other OSes already, so you can't really stretch the deduction from their open positions too much.

As an example also, take a look at these positions:
http://cdpred.com/engine-programmer-2/
http://cdpred.com/senior-engine-programmer-2/
http://cdpred.com/multiplayer-programmer/
Requirements are very generic, and nothing is said about Visual Studio.

Note there:

At least one AAA title shipped on different platforms
At least one AAA title shipped on different platforms will be a strong plus
At least one shipped multi-platform, multiplayer title

Sounds like aiming at cross platform engine development to me.

Getting out of MS lock-in is a worthy goal in general, even if it requires spending time on getting experience in more tools. It'll make their development more mature and professional.

Anyhow, we don't know their plans for sure. May be at some point CDPR will explain their plans clearly, but right now they aren't talkative about their development.

 

[grregg]

If memory serves they mentioned that in Red Engine they use 3rd party technologies as much as possible. It was in one of their dev diaries I think. If that's true, then they are facing not only moving the engine itself but also all supporting techs as well. I wonder how many of those technologies are cross-platform or have a Linux equivalent.

 

[jerf.674]

All these talks about "1.5% of users" and "unprofitability of Linux port" are complete nonsense.

Humble Bundle showed (and you can see their official statistics on their website), that Linux sales constitute from 1/5 to 1/4 of total money they get. While this number is, of course, lower than the corresponding number for Windows sales, it's already quite similar to the situation with Mac.
Also, all major Kickstarter games announced Linux support due to popular demand. You can check the comments sections there and see rather high number of comments asking for Linux support (much higher than one would expect from mere 1.5% of users).
And finally, Valve announced that Steam and Left 4 Dead 2 are coming to Linux (this will happen very soon, by the end of this month), and then other Valve games as well. You wouldn't think that Valve would do something completely unprofitable, would you?

But the most important point is that, provided a Mac port is planned, there is very little effort required for making a Linux port. All the tools relevant to games are mostly the same on Mac and on Linux. Some people raised concerns about third-party software, but if CDPR are going to make it work on Macs, a similar solution will work on Linux as well.

So, all-in-all, it's quite obvious that making a Linux port if they'll have a Mac port will be quite profitable since it'll require little effort, and the sales will be comparable to Mac sales (judging from Humble Bundle results and Linux users' activity in Kickstarter projects).

Thus, I'm quite surprised to see people being opposed to making a Linux port. Guys, if CDPR will make Linux ports of their games, they will have _more_ money to make their games better for you, Windows users, not less (while spending virtually no developer time on this). So you should, in fact, support this, not be opposed to it.

As a final comment, I'd like to address the point that some people raise, that there are too many Linux distributions and it's very hard to make a product that will work on all of them. This is, in fact, a misconception. If CDPR will make a port targeting just Ubuntu (like Valve are doing with Steam and Left 4 Dead 2), everyone will be happy. The point is that, first of all, a good share of Linux users use Ubuntu, so they will be immediately happy, while all other distributions will themselves make sure that the product released for Ubuntu will work for them as well, which will not require any additional effort from developer of the product. Note that I'm not using Ubuntu (I use Arch Linux), and I'll be still very happy if CDPR release their games for Ubuntu, since I know that the community will almost immediately make them working on Arch as well.

 

[ledgem]

Getting back to the original topic, my guess is that you'll eventually see it on Linux after Steam is ported to Linux. As has already been stated, the Mac version (which I'm playing) is simply a WINE port. I can't imagine that there would be too much tweaking necessary to get a wrapped version available for Linux. The bad news is that the Mac version is pretty unstable...

 

[CostinRaz]

Thus, I'm quite surprised to see people being opposed to making a Linux port. Guys, if CDPR will make Linux ports of their games, they will have _more_ money to make their games better for you, Windows users, not less (while spending virtually no developer time on this). So you should, in fact, support this, not be opposed to it.

Time is money for game developers. OR where exactly do you think most of the budget of the game is spent? It's spent on salaries for the development team. The more money you have the longer you can keep people working on the game and even hire a greater number of people to do so.

You go based on speculation and how cheap it would be but you fail to provide any cold hard facts on that. Linux is the smallest market, that is a fact, and you need to prove that CDPR would make quite a bit more money then they would spend to actually make it worthwhile, if you can't you are just wasting time.

With the console version they likely spent just a bit over 2 million making it, money which they made back and then some with just week one sales. In that case then I find it acceptable for them to do it despite my deep seated hatred for consoles and well some of the things imposed by Microsoft were actually good: I am talking here of the cinematic loading times with Dandelion's dialogue which give further story exposition which exist because Microsoft does not allow loading times that are long without them being interactive.

With the mac version. I am unsure if it was actually worth it for them to do it, but let's not ignore the fact the Mac market is larger then the Linux one and also that the Mac version of W1 was riding high from the success of the console version of W2.

 

[jerf.674]

Time is money for game developers. OR where exactly do you think most of the budget of the game is spent? It's spent on salaries for the development team. The more money you have the longer you can keep people working on the game and even hire a greater number of people to do so.

You go based on speculation and how cheap it would be but you fail to provide any cold hard facts on that. Linux is the smallest market, that is a fact, and you need to prove that CDPR would make quite a bit more money then they would spend to actually make it worthwhile, if you can't you are just wasting time.

With the console version they likely spent just a bit over 2 million making it, money which they made back and then some with just week one sales. In that case then I find it acceptable for them to do it despite my deep seated hatred for consoles and well some of the things imposes by Microsoft were actually good: I am talking here of the cinematic loading times with Dandelion's dialogue which give further story exposition which exist because Microsoft does not allow loading times that are long without them being interactive.

With the mac version. I am unsure if it was actually worth it for them to do it, but let's not ignore the fact the Mac market is larger then the Linux one and also that the Mac version of W1 was riding high from the success of the console version of W2.

As I said, Humble Bundle sales show that Linux users pay in total a sum which is very similar to what Mac users pay. This is a cold hard fact, which you can check on the Humble Bundle site.
Total number of Linux installations is indeed smaller then the number of Macs, but the point I was making is that the _gaming_ market is quite comparable. Which is proven by Humble Bundle results.

Some cold hard facts about effort required for porting to Linux: if they are making a Mac port (which they are making), they need to either use OpenGL, or use DirectX under Wine. OpenGL libraries are the same on Linux and Mac, so that won't require any porting effort at all. Wine is, in fact, _primarily_ developed for Linux, with Mac OS support being only a side effect. With other libraries required for a game the situation is very similar, they are available on both systems in the same way.

These two points above make it, imho, quite clear that sales of the Linux version will give the developers much more money than will be required to make a port, provided that they already have a Mac port (which they plan to have).

 

[CostinRaz]

Humble Bundle is irrelevant as an example considering the low amount of money they've made with overall sales and considering the low price point and of the fact that in each bundle you get more then just 1 game.

A game that costs close to 50 dollars is an entirely different thing and there is no AAA title on Linux to prove that Linux sales would be anything other then insignificant.

If it were as profitable and easy as you claim it is then why have no other companies done it that have Mac versions of their games. I look here specifically at Blizzard with WoW and SC2 who both have Windows and Mac versions, but not Linux versions.