Dave Taylor is an owner, producer, and designer at Crack.Com, a game development company based in Austin, TX. Crack.Com is currently developing Golgotha, a first-person real-time strategy game that will be released for the Linux platform. Dave has taken some time out of his busy schedule to give us some insight into his opinions on Linux gaming:

1) On the official Crack.Com site (http://crack.com/about/ports.html) it is stated that, "Crack, by making cool games available for Linux, has also helped push the OS authors towards making the OS the finest answer (bar none!) there is for game development." In your opinion, what does the Operating System/Dev environment currently lack that is vital to being a top-notch game development solution?

It's lacks only things which are hard to support in any OS right now. Things are getting a bit thready these days, which means you need to be able to debug them, and that's not easy, not on any OS. But I wouldn't be surprised if Linux is the first to pioneer useful tools for debugging threaded proggies. Linux needs an incremental linker. On large projects, it just takes too long. Lesse. Well, I've been thinking for quite some time that this whole idea of header files and C files is stupid. You should write all your code as functions and data, and store them all in a database. Prototypes should be generated automatically. And vi & emacs should be augmented to quickly reference the functions and prototypes. A lot of time in large projects is spent just dereferencing layers of crap. This could be sped up. Again, wouldn't be surprised if a tool in Linux was the first to adventure into this land.

2) What is your opinion of the General Graphics Interface project? Is the API noticeably improved from the Linux graphics libraries which preceeded it (SVGALIB, etc.)? Do you plan on using GGI in the Linux Golgotha port?

I think GGI is swell. Real easy to learn and wraps around anything. Yes, definitely better than SVGALIB. Yes, I want to use GGI on the Linux Golgotha port, but in the end, it's sort of Linus' decision. Whichever modern, general graphics API makes it into the kernel first will be my API of choice, assuming one goes in at all.

3) Do you feel that Brian Paul's Mesa accelerated 3D graphics library is an effective solution to developing OpenGL applications on the Linux platform?

Sure.

Are there any performance/compatibility issues with Mesa that you wish were resolved?

Don't plan to use Mesa for production games. Glide is supported natively. The other cards aren't accelerated yet (last I checked, which was a while ago). This will sound blasphemous, but I have been burned by 3D cards. Don't like to waste a lot of time worrying about them. Your game looks different on every one, and often times looks like crap on most. Seems better to go software then support cards you really like and support them natively.

4) In one of your .plan updates (June 15, 1998), you state that "We haven't released a Linux Golgotha demo because we've had trouble programming with threads in Linux." Can you elaborate on the difficulties you've had with porting your source code to Linux? Also, are there any compiler issues which must be resolved?

That thread problem went away, also documented in my plan file. Pretty sure it was our fault, although the difficulty of debugging it was definitely real. Not many other difficulties, really. Compiler issues? No. Egcs is great.

5)Can you describe your first experience with a Linux system?

I think I first experience was trying to install slackware from a buttload of floppies. I guess I'd say the experience was thrilling (what, Unix on my PC?!) and frustrating (how the hell do I pick my monitor timings!).