GameTree Linux Is Trying To Be Its Own Steam-Like Platform

[yogi_berra]

Is that the color management system's problem or a device one? And if it's a device one...why are you blaming the lack of a CMS for things there?

lcms is the pile of elephant turds here. Oyranos is a fairly good free CMS but no one packages it. Most studios that have the man power that use Linux, write their own CMS tools because of the asstastic nature of the "free" tools that you are arguing support for.

The proper answer would be: "Very few..."

Deanjo already corrected you on that. Believe it or not, quite a few people don't have a problem paying $700 for a program and $1300-$2000 for the suite of programs that surround it.

The biggest impediment is more perception on things than anything else. It's not CMS. It's not really Sound (I've witnessed as much woes with sound on Windows as we're having...). It's not really 3D, per se.

As much as I hate wiki: [citation needed], Mr. Cox provided better sourcing for his argument against porting than you.

[Wyatt]

As much as I hate wiki: [citation needed], Mr. Cox provided better sourcing for his argument against porting than you.

You're...you're telling someone who ports things to Linux for money...you're telling him that he needs to cite his sources? I very nearly laughed aloud.

Oh right, TransGaming. If their wrapper here is as good as Cider...

...they're doomed. Completely. Irreparably.

PS: PulseAudio is a pile of dog...feces.

[L33F3R]

Says who? Since I've switched to Linux, I just bought less games. Bought some indie games that do run on Linux, but in general.. I wanted to play GTA IV, would have happily forked out the money on release day, but it doesn't run on Linux. Only recently it's sort-of possible to run it on Wine, but it's slow and a LOT of trouble. So much in fact I didn't succeed in getting it to run.

I'd say that's a lost sale.

im speaking on terms of people switching to linux because quality games finally exist for it. That is key to my point. COD Black Ops prolly doesnt work very well on wine so thats a decent example. If Joe is heavy into BlackOps but prefers linux, he needs to pick whats more important, not unlike yourself. So lets say he decides to play his game and get stuck with windows. A common choice. He buys 1 windows title and 0 linux titles (also assuming he is only buying the game for personal use). If a linux version existed, on the principal that he is not a fucking retard, he will buy 0 windows titles and 1 linux title. Assuming both versions cost the same, say, $49.99, Activision stands to make no more money then they would have if they only supported windows. In this example i am not factoring in people already using linux, I specifically respond in respect to those who are using windows but would use linux if games existed.

I understand coders can work on either OS, but im sure you dont code for nickels and dimes svartalf. More code generally equals more time and people dont work for free. If some magical programmer could port a game like blackops in a year, and receive a decent wage (higher than 75k), activision would need to pull in over 1500 copies to pay for his labor alone. Therefore, my point of them catering to Linux for the fuck of it, sends them at a loss. I think you are getting confused as to me saying that they would lose money as a whole on a linux port, this is false. There is infact money to be made in linux. Im just saying that major publishers dont give a fuck about having people switch to linux.

[Remco]

I understand coders can work on either OS, but im sure you dont code for nickels and dimes svartalf. More code generally equals more time and people dont work for free. If some magical programmer could port a game like blackops in a year, and receive a decent wage (higher than 75k), activision would need to pull in over 1500 copies to pay for his labor alone. Therefore, my point of them catering to Linux for the fuck of it, sends them at a loss. I think you are getting confused as to me saying that they would lose money as a whole on a linux port, this is false. There is infact money to be made in linux. Im just saying that major publishers dont give a fuck about having people switch to linux.

I'm pretty sure a triple A title would sell more than 1500 copies in the Ubuntu Software Center. Let's say there are 1 billion computer users, and 1% of those use Linux. A lowly 10% of them wants to play games, and 10% of them like this particular game. That's still 100000 sales.

[deanjo]

I'm pretty sure a triple A title would sell more than 1500 copies in the Ubuntu Software Center. Let's say there are 1 billion computer users, and 1% of those use Linux. A lowly 10% of them wants to play games, and 10% of them like this particular game. That's still 100000 sales.

Lets put some more realistic numbers in there shall we. Let us use the best selling PC game in history, the Sims with 16 million copies sold which occurred during a period where PC gaming ruled the roost, assuming the % of linux gamers are at the same percentage of windows gamers by OS usage.

16,000,000 * 1% = 160000 potential linux sales including dual boot users. Now most surveys conservatively figure there is about 60% of users that dual boot so we can't count those as additional sales. That brings us down to about 64,000 sales. Now mark up on games are quite high when going from the distributor to retailer a $60 typically costs around the $30 mark with retailers like Best Buy (I'm actually using COD Black ops for reference here and what the retaliler would pay from some of distribution houses I have access to) and the distributor has his profit to pad as well so we will knock of an additional say $5 per copy there.

We now have the publisher selling the game roughly for $25. That would be roughly 1.6 million in revenue before the cost of development is subtracted from it. Let's low ball the programmer doing the port and pay him 100k of that so we are at 1.5 million in additional profit and that is in an absolute best case scenario with the best selling PC game to date.

There isn't a whole lot of beans there.

[deanjo]

This is all assuming as well that all those copies were purchased at full initial retail price.

[Wyatt]

Lets put some more realistic numbers in there shall we.

Okay, good. Now here's my question: where's your code reuse?

If we look at this from the standpoint of more than an isolated incident; as part of a sane longer-term strategy on a publisher's part, maybe the story is different. Let's try a different projection.

So the initial overhead to add another platform to the development workflow in terms of technology is...let's say your technology portfolio is fairly portable (properly agnostic code, aside from the platform-specific wrappers) because you do simultaneous development for 360, PS3, Windows and maybe Mac already so it's a relatively modest $5,000,000. Let's say we're looking at a five year period. So then maintenance for the Linux branch of your code base might require only about five people if you do it right, and you can cross-train your PS3 and OSX people because both of those use OGL (well, almost). If you do Wii, get them involved, too. But your five dedicated folk will run you $85,000 plus inflation. Since the economy is a bit rocky right now, let's say that means the five-year average is $95,000 (you do give raises to your employees, right?); 25 man-years at $95k is 2.375 million dollars. So we're up to around $7.5 million for a five year investment in a budding platform.

Now, from what I can see from a quick jaunt around, around five million units sold seems completely reasonable for a year (NPD only tracks retail box (in 2011? For shame), so it's hard to peg). Lead time on AAA titles seems to average 24-36 months in many cases, so if we figure the rate of platform growth for linux slows slightly (somewhat unrealistic, according to anecdotal evidence), starting right now, it'll be at around 6.2% three years from now. Even if we assume W3schools is off by a factor of around two (not necessarily true, especially when it comes to the populous that would be buying games according to classic (but still relevant) player profiles), that's still 3.1%

The PS3 and 360 consoles had an attach rate around 8.2 last I saw (summer, some time). If we assume it's only half of that on PC....oh, lets' just round it off to 4. (Not quite reasonable given the simplicity of digital platforms; who has Steam games they've never played? Many. This may, however, offset the early-adopter skew in attach, and normalise for you only being one company so let's leave it for now.)

So how many PC users play games? All of them. (this is not up for debate; there are no living, thinking humans that don't play) How many buy retail video games? Well, if we use your estimation method, that's 5,000,000 * .031 * 4 (attach rate) = 620,000 copies in the third fiscal year. If you're shipping them to retailers at $25 a unit (precluding the existence of special edition SKU), that's 15,500,000. Even if platform growth halts immediately, and games continue their trend of not adjusting prices for inflation, that's still $46.5 million in five years. 620% ROI on your overhead and maintenance. More realistically, the platform will grow because that's what the trend tells us. Let's go with 0.2% yearly (half the growth of W3's recent trend). Okay, so that's only another 120,000 units over three years, but the $3 million covers the salaries, at least: $49.5 million in sales and 660% ROI

Now, all of this makes some possibly-bold assumptions, both about the sanity of your decision-making process, and about the willful behaviour of the consumers. But even if, once again, we halve the end product, it's still $24.7 million plus consumer good-will. And only five years. Considering the overhead and the economy, another $17 million in revenue can't hurt, and it's a "charitable" bullet point for the PR folk in the short term while the platform matures in the mid and long (it paid for itself and grew a bit; you're sticking with it, right?).

[L33F3R]

these figures seem hardly reasonable. deanjo's are somewhat realistic. And hes right, best case scenario.

Blackops sold 5.6 milli in 24 hours
Crysis sold just over 1 milli in a year. Hardly a boring title that people diddnt know about.

The numbers are realistically very low and start to seem pointless, unless your an indie.

Edit: Note crysis was Windows only.

[deanjo]

these figures seem hardly reasonable. deanjo's are somewhat realistic. And hes right, best case scenario.

Blackops sold 5.6 milli in 24 hours
Crysis sold just over 1 milli in a year. Hardly a boring title that people diddnt know about.

The numbers are realistically very low and start to seem pointless, unless your an indie.

Edit: Note crysis was Windows only.

Exactly, the numbers I give are pretty much a "when all planets are in alignment" scenario. A typical "block buster" AAA PC game will usually fall within the 1-3 million copies sold and I didn't even factor in the slope of discounting as a title gets older or promo prices or the distribution of incurred cost over all of development for the shared assets.

[Wyatt]

Just curious, what about these figures seems unreasonable? While I acknowledged the engineering costs made some rather bold assumptions in terms of overhead estimation ($5M is possibly a little low-ball, but really shouldn't be), the rest are all derived from real statistics that are publicly available.

Inflation statistics (US) formed the basis for programmer pay, though perhaps using CPI as the indicator is frowned upon? How are salary increases usually estimated?

For the past two years, retail box PC game sales have been in the 25 million unit range. I think it's completely reasonable that a significant publisher (EA, ATVI) might have as much as 20% (or more) of this segment.

W3school's browser statistics are often cited as being biased toward "tech people", but so is PC gaming; the assumption of slowing growth in my projection is a mitigation factor, as is the slashing of the result by 50%. In reality, it could be more than twice what I have estimated.

The attach rate is not likely to have changed significantly in this short term, and I'm using half of the console estimate to reflect the nature of PC consumers and their tendency to play individual games for longer periods. I will grant that the use of attach as a multiplier may be somewhat flawed because of this; hence the rest of the mitigating factors (not counting digital anything, micro-transaction, DLC sales, or Sub Fees, ad revenue, etc.).

The NPD's digital tracking is largely third-party and somewhat suspect. But, even then, estimates for 2010 have digital outstripping physical in sales of full retail game SKUs. The revenue factor is 0.43 versus 0.57 in the physical space, but it's at or surpassing parity from greater unit volume.

If we're talking realistically, then I really don't understand what part of this projection is contentious. If you have anything to add, or can eliminate extraneous mitigations (that would surely be nice), I'm all ears, but quoting sales figures for two individual titles and calling it a wrap seems to miss the point.