Getting indie developers on board with Linux shouldn't be hard in the coming future. Linux is rapidly growing and is started to be seen by the public eye as an OS some people would actually personally use for their own personal computer. Also, Unity and Unreal Engine 4, the two biggest engines indie studios use to develop 3D games, both easily support developing for Linux. Lastly, and most importantly, these guys don't have a choice. If they want to compete with the triple-A market they have to expose themselves to the biggest market they can.
To get the triple-A companies on board is a tougher problem. The most-thorough and simplest solution is to get more Linux users. Triple-A companies use engines built on top of DirectX because DirectX works much better than OpenGL on Windows and everyone uses Windows (like, everyone. Windows still has 90%+ market share). If you pulled the rug under that logic by having everyone use Linux, you would force triple-A companies to support multiplatform engines and graphics APIs. Just because this solution is simple doesn't mean it's easy. Linux market share is currently pathetic. There are more Vista users than there are Linux users. While on forums and discussions and other virtual agoras, evangelists annually preach "the year of the Linux desktop" and with each patch to every WiFi driver pundits will make blog posts about how better Linux is getting and how no one likes Windows 8 and just-you-wait, the fact of the matter is, before Linux topples Windows in the market share graphs, it has to start toppling "Other" in the market share graphs. And it's pretty obvious why Linux isn't getting the market share it deserves. 1. No one wants to install a new operating system on their computer, and the fact that people have to know what an operating system is at all is a tragic case of a leaky abstraction, because people don't want operating systems, they want pictures and videos and music and Facebook and e-mail, 2. No one wants to deal with WiFi and graphics driver issues, and 3. No one wants to think for a single second past "Best Buy" when they want to buy a new craptop for their lovely little girl going to college (which I think is why Win8.x is beating both OS X and Linux combined despite being an objectively worse OS), and you sure as hell won't be finding any Linux at Best Buy. Marketing exists for a reason. Older computer geeks with long beards remember buying the Macintosh 128K for being the first computer with an effective implementation of a GUI and a mouse; many other people will remember buying the Macintosh because the 1984 ad caused Apple to explode in the public eye and turned Steve Jobs from one of those computer people into a heavily publicized rock star. Apple would continue to master and practically define hype, and while OS X isn't faring as well in the market share war, what Apple really is focusing on -- iOS -- is killing it. Until the Linux folks discover how important marketing stuff is, they won't capture the people who don't care enough about their machines to think past "what's the most popular thing being used right now?"
These are huge issues, and until they get solved, you won't be seeing Linux with a high marketshare, which means no triple-A company will bat it a single eyelash.
The easiest solution, although less effective than the first one, is to increase marketshare in platforms that are not Windows. This is because most platforms that are not Windows will require use of APIs that are multiplatform, making it frictionless to port to Linux. Somehow get people to game through the browser by making it possible to run triple-A games in Chrome? Good. Linux has Chrome. Get people to game mainly on OS X? Good. OS X games usually use OpenGL, and that's easy to port to Linux. Because this isn't as direct, you will always get some people who give so little shits about a set of OSes that have less users than "Other" does that they won't go through the effort of clicking the checkbox next to Linux in the engine project setup page, but it's somewhat of a solution.
To get the triple-A companies on board is a tougher problem. The most-thorough and simplest solution is to get more Linux users. Triple-A companies use engines built on top of DirectX because DirectX works much better than OpenGL on Windows and everyone uses Windows (like, everyone. Windows still has 90%+ market share). If you pulled the rug under that logic by having everyone use Linux, you would force triple-A companies to support multiplatform engines and graphics APIs. Just because this solution is simple doesn't mean it's easy. Linux market share is currently pathetic. There are more Vista users than there are Linux users. While on forums and discussions and other virtual agoras, evangelists annually preach "the year of the Linux desktop" and with each patch to every WiFi driver pundits will make blog posts about how better Linux is getting and how no one likes Windows 8 and just-you-wait, the fact of the matter is, before Linux topples Windows in the market share graphs, it has to start toppling "Other" in the market share graphs. And it's pretty obvious why Linux isn't getting the market share it deserves. 1. No one wants to install a new operating system on their computer, and the fact that people have to know what an operating system is at all is a tragic case of a leaky abstraction, because people don't want operating systems, they want pictures and videos and music and Facebook and e-mail, 2. No one wants to deal with WiFi and graphics driver issues, and 3. No one wants to think for a single second past "Best Buy" when they want to buy a new craptop for their lovely little girl going to college (which I think is why Win8.x is beating both OS X and Linux combined despite being an objectively worse OS), and you sure as hell won't be finding any Linux at Best Buy. Marketing exists for a reason. Older computer geeks with long beards remember buying the Macintosh 128K for being the first computer with an effective implementation of a GUI and a mouse; many other people will remember buying the Macintosh because the 1984 ad caused Apple to explode in the public eye and turned Steve Jobs from one of those computer people into a heavily publicized rock star. Apple would continue to master and practically define hype, and while OS X isn't faring as well in the market share war, what Apple really is focusing on -- iOS -- is killing it. Until the Linux folks discover how important marketing stuff is, they won't capture the people who don't care enough about their machines to think past "what's the most popular thing being used right now?"
These are huge issues, and until they get solved, you won't be seeing Linux with a high marketshare, which means no triple-A company will bat it a single eyelash.
The easiest solution, although less effective than the first one, is to increase marketshare in platforms that are not Windows. This is because most platforms that are not Windows will require use of APIs that are multiplatform, making it frictionless to port to Linux. Somehow get people to game through the browser by making it possible to run triple-A games in Chrome? Good. Linux has Chrome. Get people to game mainly on OS X? Good. OS X games usually use OpenGL, and that's easy to port to Linux. Because this isn't as direct, you will always get some people who give so little shits about a set of OSes that have less users than "Other" does that they won't go through the effort of clicking the checkbox next to Linux in the engine project setup page, but it's somewhat of a solution.