The development of Quake II and hard coded data

[ses]

Firstly this port of Quake II to GWT is fantastic, well done to those involved!

I've been digging through the game's source code and I was wondering if anyone knew about the way in which id software developed the game, specifically why so much is hard coded, for example information to do with the 'monsters' and game menu?

AFAIK id had some great internal tools for building maps and scripting AI including there own file formats. Therefore I don't understand why a lot of other things appear right in the code which would make it harder for someone to pick out and edit if they needed to. Perhaps it is to do with performance?

[Stefan Haustein]

I was a bit surprised to see this, too. Perhaps some of the code was auto-generated by a tool? I must admit I don't have a deep insight into the original development process -- what we have looked at for the GWT port was already the result of the Java conversion done by the Jake2 authors...

Stefan

[Joel Webber]

It's been a long time since I looked at this stuff, but IIRC the menu code is just a big ugly hack, written by hand. But the other hard-coded data (especially character animation frame sets and the like) is generated by an export tool. That stuff was checked into the C code base, then the Jake2 guys just converted it into Java. It's really quite ugly, but it makes more sense if you think of it as generated from a tool.