Back again

well... i don''t think there are no bigger difference in producing games for any OS.

But if it should be a good book i suggest that u write
1) Something like a "first course" in game programming... something like the book WGPFD, but more compact.
2) Write about tools that can be used during the developement of a game... and some info about X11 games vs. console and how to get fullscreen and stuff
3) Write it damn interesting

cheers!



"Look mom, a demoscener! Can i keep it?"

You''ll probably have to emphasize differences in Linux and Windows programming. Such as signals vs. Windows messages, threading models, etc. I wouldn''t spend much time on networking, as Berkeley sockets are covered in mind-numbing detail in many other sources. You might want to cover briefly cross-platform libraries such as glut/mui and ClanLib.

Other topics to cover include sound programming, making X11 happy, and the ever-popular why-bother-making-games-in-Linux topic.

Well, I wouldn''t really want a book that spent all the time comparing the Linux API with Windows, since I can''t program DirectX

I''d buy a book that assumed that you know a bit about game programming theory, but didn''t know how to apply it to Linux. Although one that works from scratch would do if it''s not too patronising. Find your target demographic!

Things to include could be XFree86 4.0 (especially DRI/XXA) and a little on OpenGL/DRI 3D programming. Console mode (using framebuffers), X (including how to switch off the screensavers, getting constant focus, getting non-root fullscreen mode like Q3A, gaining priority over running processes), sound (using OSS, the standard Linux sound routines etc.), input (mice, joysticks in and out of X), CD-ROM access (mounting, streaming?) and so on.

Just to further oppose what SiCrane''s just said (sorry , I''d like a bit of network programming, simply because network programming for games is different from programming, say, an FTP server (but I wouldn''t go too in depth, and I''d point the reader somewhere else). If I was going to look at a cross-platform toolkit I''d use SDL rather than a high-level one like ClanLib. It''d be more fun not to use such a lib at all though. Saying how to apply threading and piping to a Linux game would be useful, but again don''t go too in-depth (point the reader to "Beginning Linux Programming 2e" Stones/Matthew instead).

A big topic to stick in the Appendix would be portability to other platforms (Windows, BeOS, not to mention other Unices).

On balance, then, I believe that a vast portion of the book would be on how to get fast graphics to run on XFree86 4.0 and get fullscreen to behave itself!

First of all, i just want to say that i think it''s really cool that you''re writing this book! I''m looking forward to seeing it on the shelves.

Next up, i think you really have to realize that a large part of the readership is going to be game programers coming from windows or other operating systems, and as such, you should expect a basic understanding of game theory and proficent programming skills. Otherwise, this book will bore people before they get to the good stuff.

Also along these lines, i suggest you look at cross-platform API''s, such as SDL or Clanlib, and as Arfa said, this would be a great subject for an Appendix.

Remember though, it is Linux game programming, not cross-platform game programming, so cover the Linux-specific stuff, even if there''s a cross-platform alternative. Some readers may be looking to port their existing games, and just need to replace key parts...

Also, please, please, please develop one or two projects throughout. Keep the emphasis on the API to a minimum, but feel free to use and cover one (SDL comes to mind). This kind of thing really helps the flow of the book, as well as being the perfect writing guide.

I''m sure there are a billion other things, too, that i''ve forgotten, but for now, that''s it.

-BenC