I suspect that the higher Linux sales is, in fact, due to pent-up demand.  There's actually no scarcity of native Mac games these days, not like the old days.  But Linux is still something of a gaming desert (not counting Wine).  However, I expect that to change in the coming years, now that Valve is so gung-ho for the platform.  The great thing about it is, as long as a game like GalCiv 3 is ported to either Mac or Linux, a port for the other platform is a might-as-well.  So I figure if we get one, we'll get the other.  Let's hope.

GalCiv 3 would look great running on a Steam Box.

The trick is to convince them to make that initial investment.  I'm convinced that there are enough eager Mac/Linux gamers out there to more than subsidize the initial development costs.  And from there, future development costs will be much lower.

unity would be awsome....

If you look at market shares for operating systems, for ex:

http://www.netmarketshare.com/operating-system-market-share.aspx?qprid=10&qpcustomd=0

http://en.wikipedia.org/wiki/Usage_share_of_operating_systems

Obviously this type of chart isn't optimal.  A better breakdown would be how OSes are distributed amongst people who play premium games on their computers, or better yet for this game, a break down of machines capable of 64bit and Dx10+.  Anyways...to be very generalized based on Sept 2013 data...

Windows has over 85%.  Linux isn't even on one chart and falls under "other" and is under 2% on the other chart. Windows XP, which is way old now, crushes Mac and Linux combined - and I doubt many XP users have 64 bit (and/or Dx10 capacity) and all of those users won't be able to run GC3.

So the question then is, is it worth the development resources to develop and distribute versions to platforms few are using for gaming (since few are using them at all and they're not all gamers) - especially Linux?  It would seem to be a luxury or even a waste of time.  And this is probably why so very few companies do it.  For GC3, which is going to be 64bit/Dx10+ only, and as such be a little less forgiving to suboptimal configurations of any type, it would seem to make even less sense.

For me, personally, if you're going 64bit/Dx10 only and excluding the still very large chunk of people saddled with WinXp 32 bit, why would you even consider Mac or Linux and their even smaller market shares?  I would rather see dev resources dedicated to pretty much anything else - more content for the PC, for ex.

I also tend to think that most people serious about PC gaming have (or wish they had) Win7 64 PCs.  Those who willingly choose to go with some other type of machine, like a Mac, are to me, just saying, I have other priorities than PC gaming.  Why should PC game developers bend over backwards to cater to people who have priorities other than PC gaming and open up additional cans of worms for their support staff to deal with?

You could postulate that by going 64bit/Dx10 you are simplifying your development and support in some ways, because this is going to exclude a ton of out of date hardware, drivers, and other garbage that tends to massively complicate things for PC developers.  Throwing Mac and Linux into the mix could reintroduce more carnage - and with those market share numbers - would it be worth it?

Lastly, I'm no expert on the differences between the OSes as it pertains to game development but I'm sure it's not all that easy to do, or more companies would do it.  Despite 64 and Dx10+ being around for quite some time now, very few games have come out that push the tech to require them, so in some ways, this game is going to be bleeding edge.  Why add more complication when bleeding edge is usually complex itself.

While true, the # actually correlate fairly well with steam's own hardware survey. And Steam's survey is inherently biased to those who actually PLAY games so 'in theory' is a better indicator of the market share of people who actually 'play games'. And Setam's own numbers show WORSE % of users on Steam use Linux/Ubuntu than the other surveys which are more 'general'.

Still the % are that, it appears, Linux/Macs represent around single digits. Again +/- for a lot of factors but it's still 'low'. You're certainly are going to  be hard pressed to show anythign as high as 20% no matter what survey or number crunching you do.

The problem is that there aren't any good numbers, for what linux/Mac game sales are like.

Now dont get me wrong. I'm not saying its a 'waste of time' to be on mac/linux. I'm not saying that devs shouldn't do it. All I'm saying is that if you're trying to convince me, a business person, with limited resources, to invest time/money into mac/linux, then you gotta come to me with numbers to support that decision. And if I was a business person, I'm not seeing those numbers being justified, especially if the up front costs are massive.

+1 for Linux.  Or, at a minimum, making sure it can run on WINE.

I'm switching my desktop over to Ubuntu in the next few weeks and installing Steam for Linux.  I'll probably still have Win7 running in a VM, and the motherboard does support IOMMU, but Win7 will definitely be my last Microsoft O/S on the desktop.

This is impossible given that WINE hasn't fully implemented DX10 (a requirement for GalCiv3)

Even older DX10 games don't work in WINE so it's basically impossible for GalCiv3 to work on WINE given that its likely they'll be pushing the API significantly.

It's a WINE issue not a GalCiv3 problem.