Add new comment

Here's the relevant comment, it was something I wrote in 2005 when I was wrestling with the Torque engine:

There's a heirarchy in game programming: Rendering Engine > AI > Game Systems > Architecture > Networking > UI > Tools. Everyone wants to be higher on the ladder, the tools are the absolute least interesting to work on and are generally given to the least experienced programmer, who does everything he can to get moved up to something more interesting. Since everyone else is dodging UI work, usually he winds up there (but he won't want to stay there, either).

So it shouldn't be surprising that no one has bothered to fix blatant bugs in the Torque world-building tools in the last 3 years. Or that the GUI building tool is almost as bad. But it's sure a pain, I'm left with a choice of fixing them myself, extracting the data and building/using other tools, or just do the best I can with them (they are usable, just flaky and unreliable).

That, at least, was easier in ordinary software development. Since none of the work was particularly interesting, you could assign someone a piece of it and expect them to finish it, since the only satisfaction and hope for recognition they had was to do it well. But no-one wants to do a good job on tools or GUI work, because then they might get stuck on it.

WoW's tools do seem to be pretty good. Certainly their world-builder's tweak the output from the algorithms, good procedural tools are not a replacement for world-builders, but a way to create an overall consistancy and improve their efficiency.

--Dave