Repercussions of targeting *nix

**ZekeDragon** · 10-26-2009, 01:23 PM

Commonly, software developers get themselves into the mindset of choosing to support three different sets of platforms. What I mean by this is most developers think their program should either target "Specifically Windows", "Windows and Mac", or "Windows, Mac, and other *nix" (commonly carpet defined as 'Linux'). However, I've been delving into the POSIX standards quite a bit, and to no surprise I found them to be quite helpful for all sorts of programming tasks. pthreads are really just the beginning, you can easily take advantage of the signal system in these operating systems, the native timing utilities, and a myriad of other techniques. What this did is it got me wondering...

What are the repercussions, other than not being on Windows, that choosing not to support Windows entails? For example, say you were to develop a commercial product that targeted any POSIX compliant platform, such as Mac OS X, Linux, (Open)Solaris, etc., do you think this would prove to be a viable/non-viable target for a small to medium sized developer, and why? I know of companies that used to target only Mac, so that has to be a viable platform in and of itself, but would supporting other platforms dissuade Mac users from using your product? Would a program developed in this manner still not look right on the target platforms?

I consider this because when you cut out Windows support, suddenly you're able to use so many of the advantages that come from POSIX compliance. Further, it's obvious that you should use OpenGL for graphics, and you know that the platform it's on supports it fully instead of some OpenGL wrapper for DirectX. I'd also like to hear what other advantages can be attained by not keeping Window support as a requirement are. If I programmed this way, I'd might as well get as much out of it as I can.

Just as a side note, I'm not considering support for BeOS-like or AmigaOS-like operating systems as of yet. Right now I'm thinking Mac OS X, Linux variants, (Open)Solaris, and (Free/Open/Net)BSD. Further research might open up Be/Amiga style OS', but I'd like to get more compatible systems out of the way first.

I just wanted to know what everyone thought. My experience has been that for a small business doing game development, Mac/Linux/BSD group can be a fiercely loyal group of people to have liking you. Windows users get used to having everyone target their platform, and I think it'd be a nice change of pace for other OS' users to find someone targeting them instead. What do you think?

**WingedPanther** · 10-26-2009, 03:41 PM

The biggest issue you face is market-share. POSIX-compliant is nice, but if 90% of your market is using Windows, you've limited yourself to 10% of the market. Telling your customers to install Cygwin is generally not going to be considered a valid option, in those cases.

Personally, I wish Microsoft would add Posix compliance to their OS. Then developing cross-platform software would be MUCH easier.

**ZekeDragon** · 10-26-2009, 03:51 PM

I see what you're saying, but what I'm proposing is not to tell my customers to install Cygwin, but instead relegate my customer base to that 10%. Nearly every company in existence ultimately limits it's market, no matter what you do. Stores in a small town or places that specialize in some sort of sales, such as knitting, computer parts, or arts also limit their market. The point is that this limitation can be strategic, as it makes it so satisfying the needs of that customer base become much easier and more cost-effective. Simply put, that sort of thing is necessary for a small company.

Being a small fish in a big pond is extremely difficult for the small developer, but being that same sized fish in a small pond will exponentially increase your chances of survival in that pond. You significantly limit the number of competitors you have, while simultaneously being something of a "bastion" for your users. I understand that market-share is a limitation, but is it worth it? That's the question.

**WingedPanther** · 10-26-2009, 04:13 PM

It depends. Be sure to check out the competition. POSIX has a LOT of free software out there in a lot of disciplines.

**ZekeDragon** · 10-26-2009, 06:55 PM

Well I'm not afraid to admit that my thought process on this was leaning toward games. Certainly there are several free games, however games have the unique property of being an instant micro-monopoly, since you can never have a game that works precisely like another one (except with trivial or standardized games, like Chess, card games, etc.). Caster was made by a single person, and is quite unique. Even if someone else did make a nearly precise clone of it, you'd still have a customer base who felt the other free game was of lesser quality or is underhanded.

Doing other software would be difficult as a business unless you had something really nice and it wasn't already available.

**Guest** · 10-26-2009, 10:16 PM

If a proprietary software developer supports only Linux and Mac, they will automatically lose a lot of Linux user market share. Most people that use Linux use it because it is free/open source. (At least, I'm pretty sure that is the reason).
Another thing: People have already learned to avoid Mac computers if they want to play a lot of games.