P: Provide support for Linux (2011)

It seems that there is a vicious cycle occurring here. Linux (or "Linux/GNU", to be a stickler about this) doesn't have the market to support large companies supporting the operating system, and the market isn't there because the user doesn't have all of the software he/she may need in their workflow (photoshop, 3dsmax to name a few).

@1049674: There is a market on Linux for server software. Just not so much for desktop software (I should have been more specific).

And Photoshop is few orders of magnitude more complex than the apps you first listed. Yes, the lack of standards and problems with APIs on Linux can be overcome -- if you have a market that makes it worth the hassle.

I think you've hit on gold here. If your development team were to make it a goal from now on to use "coding practices that avoid, as much as possible, OS-specific code, hooks, and APIs" it would put you in a better position to take advantage of future OS platforms as they come along and it would make it more likely that we'll see a Linux port someday. Who knows, the next major OS might even be Linux-based. If you count phones, I would say that it already is (Android).

@Chris: I don't know if you're asking me or Glen, but while I agree that PS is pretty complex, I don't think that Gimp, Firefox, OpenOffice, etc are *not* complex.
I don't really think it's a matter of complexity, so much as it is a matter of coding practices that avoid, as much as possible, OS-specific code, hooks, and APIs.

As far as your earlier comment that there is "no significant market for commercial software on Linux..." That really needs to be amended. There may or may not be a market for commercial *image editing* software (and there are some commercial NLE video editors on Linux, so I would go with "may"), but there is plenty of commecial software on Linux. Well, shucky darn (can't say "he77" here, jeez), my company pays big bucks (in the $100k range) every year to Oracle, which we run exclusively on Linux.

@Chris Cox
Chris, I agree photoshop is a complex program, and I think paul archer makes a good point that alot of software for linux is originally made for linux or developed concurrently. And while the examples of paid software for linux that Glen gave may or may not be as complex as photoshop, surely there are quite a few pieces of software that are. Take for Maya for instance by autodesk, and Avid makes quite a few programs available on linux as well.

Oh well though, I'm not even sure it's worth pushing for anymore to be honest, I recently sucked it up and bought cs5.5 for my mac before my student pricing ran out. Maybe someday my dreams of linux and adobe will come true. If not, there's always Blender (in place of after effects), inkscape, gimp, and scribus, while the last 3 are not as good as their adobe equivilants, I could make do with them.

Are you being sarcastic, or really trying to compare those applications to something as complex as Photoshop?

Adobe could do it. But the software you mention was either developed on Linux (or other Unix) first or developed concurrently on multiple platforms. Adobe has a much higher hurdle to face by porting PS & LR.
But the real answer is simpler: there's not enough money in it, especially when many of the people who would like to see these products running on Linux are currently running them on Windows or Mac. So if they're already paying customers, what's the point? Plus when you're making literally billions of dollars a year, what's the point of even worrying about a small market segment?

Gimp, Chrome, Firefox, OpenOffice/LibreOffice, and most of the software I use every day works equally well on Windows, Mac, and every major Linux desktop distribution. I currently pay significant money for IntelliJ IDEA and Crashplan on Linux, and they also support all of those operating systems. If they can do it, why can't Adobe?

I think that because each element of the system is designed and developed by different entities, the software doesn't have a chance to fully mature before the next project comes into popular use by most distros.

There is no one, comprehensive checklist that I know of. But yes, lack of standards and lack of system provisions for things that should be system wide (like fonts and color) are stumbling blocks. They can be overcome, but the OS really should already have standards and handling for things like that. (and "pick the window manager of the month" doesn't help, either)