Quoting: CorbenSo, does anyone know what happens, when I buy it now on itch? Will I get a Steam key later? Maybe it's already mentioned somewhere and I just didn't find it yet.

Hey, Corben, it took me a while to figure out if I was going to do this, but the answer is yes [External Link].

View this comment - View article - View full comments

Quoting: reaVerWorthlessBums: I presume some people are a little on edge here because this kind of bad news just doesn't emit the signals to other developers that we'd prefer. I see your numbers are actually lower than my estimates (I'd expect 20k-40k for a port, due to time used, things that have to be learned, etc) and the fact most of the post release support is just staying compatible with the distributions should be music to these people's ears.

At the same time, as I stated before, your game doesn't look like something that would be interesting after a year of playing (I'm saying this from a distance, I haven't really looked at it as I skip most 2D games). If people downloaded your game or bought it on windows, it's going to be unlikely to see a return on the Linux version if you release it a year after. This has been one of the more common gripes with the AAA industry as it takes them over a year to port their PC game to Mac, while it takes the development team that same amount of time to release a new version of that game that then also needs a year to be ported to Mac (and this isn't even talking about Linux).

So if you're doing Steam Marines 2, I'd prefer to see a day 1 launch so we can see a more accurate comparison for the above reason.

Unfortunately my results are more or less in line with what a lot of other (indie) game developers have intimated to me privately and publicly. I mean we would also be happier if our Linux ports paid off more.

You're referencing behavior I haven't really seen evidence of: that people would buy a game for Windows and a year or so later buy the same game again but for Linux. I've only seen this behavior when there is a full re-release like a special edition or GOTY or something of that nature, something with a substantial update and not merely a port.

I admit I am curious if Linux builds would do better if launched on day one, but my feeling is that it's unlikely given that Windows and Mac builds of SM1 launched simultaneously and Mac sales are still underrepresented. I'll see what I can do.

View this comment - View article - View full comments

Quoting: liamdaweI don't agree with the argument that Linux having less games, means each game will have more sales. We are still normal people with limited incomes, so we won't buy literally everything, we still have to pick and choose what games we are interested in to pick up. Just having less options, doesn't mean we are likely to buy more, it just means we have less to choose from (and in certain genres like this, we have tons), I don't see it any other way.

Honestly the empirical evidence here seems to indicate your conclusion, if not necessarily your exact reasoning ;)

There could definitely be reasons to expect a lower Linux percentage than user percentage. It could well be the lack of options or something else. If we assume that the smaller potential library is a net negative (not unreasonable - a larger library is a significant reason why Steam is the big dog, right?) this doesn't really bode well for developers making games for Linux, or Linux gamers for that matter.

It really does beg the question if there are that many Linux gamers ready to buy these non-ported games, are developers leaving a lot of money on the table by not porting? As someone else mentioned my SM1 porting costs were actually extremely low, largely because of how I run my company. That there are Mac companies like Aspyr suggests that at least large non-Mac games publishers believe porting is both costly and risky.

I don't have the answer.

View this comment - View article - View full comments

Quoting: MaCroX95@WorthlessBums

I see that you've taken the correct approach this time by developing a new game in Unity :) There you can literally just push compile and the game has a great chance of already working on all platforms, just avoid using platform specific middlewares and you should have no time or cost loss there :)

Secondly, the testing. Trust me when I say that www.gamingonlinux.com is one of the best and nicest communities in general, sure there are exceptions, but make this community your friend, all in all it is great and trust me that a lot of users would love to do majority of the testing for you as well and report the bugs for the linux platform so early access or beta could be a great Idea regarding Linux. Even some Linux youtubers would probably love to do a review of your games if you let them know about them and promoting it!

Not really sure about how Mac stands, a lot of devs say that it is much more profitable to support Mac eventhough I honestly think that this is just because of the commercialism, Majority of Apple's hardware cannot even run the newest AAA games and those that can cost so much money that only small portion of people can afford it... so percentage of marketshare for Mac can be really misleading as long as AAA gaming is concerned.

Best of luck on your new game and let us know when you finish it!

Yes, I must also admit that C# is much nicer than Torquescript and C++ :D

Like I mentioned in my last reply to someone else, there are platform specific issues to work through like shaders and display driver problems and so forth, so it's not completely frictionless. I'm new to shaders, so even tiny roadbumps can be confusing!

GamingOnLinux has been suggested to me before when I was testing SM1 for Linux. That said I didn't encounter too many technical issues so it just didn't happen. It is more likely I'll take advantage of this in SM2.

SM2 performance on Mac is... not great. I recently bought a new 27" iMac (late 2015 model?) and it's completely fine for coding or watching Netflix or something, but it does not run games well. It's been a struggle optimizing SM2 for Mac.

The general expection seems to be Apple will give its hardware a kick if only for VR because as it stands it's unreasonably bad.

View this comment - View article - View full comments

Quoting: m2mg2Thanks for the additional detail and for supporting Linux! I honestly probably haven't and won't buy any of your games but I still appreciate your work. It's unfortunate to hear so many bugs on steam marines 2 using Unity, does this appear to be more work than it was with 1/no Unity? What version of Unity? Is it actually Unity or plugins/middleware that are the problem? Do you plan on releasing Mac/Linux day 1 with Windows? Comparing 3 years of Windows sales to 1 year of Linux sales doesn't seem very helpful. Would be interesting to see if the numbers are the same with simultaneous releases.

Np. Don't worry about buying my games, I'm doing fine! Also there appears to be at least one person here who'll be mad at me regardless of what I say... :whistle:

I'm only halfway through development but for now it seems unlikely that a Linux build will be as difficult, as a percentage of effort and time, compared to the previous game. That said a benefit of Torque2D was that I had access to the engine source code; I have not paid for access to Unity's so in theory something could come along and hose me.

A large source of problems in SM2 as opposed to SM1 is shaders. Stuff that just breaks completely depending on hardware and drivers and what render path you're in. I have a game breaking bug with my fog of war regarding Linux machines on Intel HD graphics. I actually use the Windows Unity Editor for all Win/Mac/Lin builds which cuts down on the Linux editor-specific bugs. Unity has an active issue tracker. [External Link] I'm currently on 5.4.1 - I'd prefer not to have to freeze my engine version if possible.

You're right that it's not completely fair to compare differing time frames of Linux to non-Linux sales, but that's also implicit in the cost of having a Linux build in that particular case. If it was costless to have a Linux build upon initial release that would likely have improved Linux sales, but it wasn't because of the time factor.

It's unlikely I'll release Steam Marines 2 with WML builds on day one just because of the volume of issues I'd have to work through. The current plan is to launch the Windows build first, hammer out any outstanding issues, then bring Mac and Linux builds to parity.

View this comment - View article - View full comments

Developer here ;)

First, thanks to everyone who said some kind things! Second, I'd like to clarify a few things.

Mac Support

In my experience, and apparently in the experience of other devs who have tweeted me since, yes, Mac gamers seem to be less technically proficient overall. Some found no support issues but that Mac was still not worth the extra dev cost [External Link], and others found similar support issues. [External Link]

Greedy Dev!

Someone tried to run the revenue numbers from Linux sales of Steam Marines 1, but it was both inaccurate and incomplete. I am under NDA as a Steam developer, but I can cast a bit of light on the whole issue.

I recently blogged that the game's entire development cost was $97,913.88 USD over a 30 month period. [External Link] That comes out to an average of $3,263.90 a month of development cost. About 70% of that cost was spent paying myself to work on my game (i.e. not contracting for art, audio, et cetera). That brings the monthly cost down to about $2,284.73.

It took me about a 3 month span to port the game over to Linux, but part of that was not full time dev, so let's just say it took about 2 months. That's a total cost of $4,569.46 to get the first Linux build up and running. The person who tried to run the revenue numbers came up with a gross figure of €4,440 - so you can already see the problem.

Steam takes 30% of gross. I pay taxes that depend on situation and taxable income, but let's say for the sake of argument I net about 60% of my gross revenue. To break even on the initial $4,560.46 I need to gross $4,569.46/0.6 = $7,615.77. If I want to actually make a profit I need to earn more than that.

Additionally, adding support for a platform is just that - support. Linux tech support was very low, likely a combination of better-than-average tech savvy and relatively low sales numbers in general. But every build I push from that point on requires additional resources to 1) push to various distribution platforms, 2) test, and 3) provide tech support. None of those are costless. Points 1 and 2 also do not necessarily cost the same per platform; Testing and bug squashing requires time, energy, and the necessary hardware.

So no, Mac and Linux sales are nowhere near profit territory with regards to their associated costs: supporting those platforms for the game was a net loss. They are largely supported by Windows sales. Early on the numbers for Linux were actually much better. [External Link] It's just that as time went on more Windows users bought the game and less Linux users did.

Linux users on average are willing to pay more per game. Unfortunately we're talking about bundles here, and an extra 25% sounds really good until you realize that's like 25 to 75 cents of the total price; adjusted by gross dollars as opposed to net unit sales does not change the math much. Linux users on Steam et al wait for sales like everyone else.

Estimating Linux Customers

There seems to be some commotion about whether distributors are accurately tracking what platform is associated with which sale. There's likely some wiggle room here, particularly regarding people who purchase a game on Windows, but play the game on Linux - or even Windows gamers who prefer to game on Linux if the option is available. I readily admit these numbers are subject to further scrutiny, but without extra data I couldn't say if there are really that much more Linux users I'm discounting. We can speculate but that's really all it would be.

Estimating is hard. I'd like to address the entire "2% is really good because we're 1% of the user base" argument. Not every game on Steam has Linux support. You'd expect the average number of Linux sales to be higher than the average percentage of Linux users, per game that does support Linux. If Linux was 10% of Steam's userbase but there was only 1 Steam game with Linux support you would not expect that game's userbase to be only 10% Linux users; the Linux users are simply flowing into fewer games than Windows users are. That said the Mac percentage numbers are low as well - maybe I'm missing something here.

Is 2% good or bad? If it's bad then there's presumably things I could do to get more Linux users. If it's good, what most people here seem to be saying, then that's actually really bad because that's a strong argument against supporting Linux in my games; they'd simply never even break even.

What did I expect? I didn't really have a round number in mind, truth be told. Steam Marines 1 sold quite a bit better than I expected overall and this was my first time porting any piece of software to Linux. If I had to do something similar in the future I expect that it would take me less time.

The Why and the What

I'm currently working [Steam Marines 2](http://www.steammarines2.com/), and the subject of Mac and Linux has already come up. This time there's no significant cost for porting because I'm making the game in Unity instead of Torque2D. That said I've already been performance testing on Mac and there are issues, and I have Linux-specific bugs a mile long. To support both those platforms is likely to result in yet another loss on the long run although not a foregone conclusion.

The main reason I'd like to continue supporting Mac and Linux in the future is that I grew up as a Mac kid. I appreciated companies like Blizzard who made their games for Windows and Mac. I'm still mildly bitter about Bungie and Halo. But it's fine. I make games for a living now which is pretty neat.

View this comment - View article - View full comments