[Gtkradiant] roadmap, developement process (Was: realtime lighting)

Pablo Zurita gtkradiant@zerowing.idsoftware.com
Sun, 01 Jul 2001 11:33:01 -0400

I actually made a flow chart in a whiteboard to keep track of the goals and 
how I was going to implement the stuff but that wasn't the problem, I think 
the problem were 1. Quake engine construction method isn't the best to 
implement this kind of stuff. 2. The code just wasn't ready for such 
implementation, the "environment" has to be prepared before start coding. 
3. I think I need more knowledge on the subject to make something good. I 
did have a plan and everything, but every time I was going to do something 
I always ended up coding other stuff to make my plan work.


At 12:35 PM 7/1/2001 +0200, you wrote:
>Ok .. there are plenty of other areas to work on anyway. I've tried to keep
>this list updated about the current status and to produce roadmaps for next
>release, so I hope it's rather easy to select critical areas that need working
>A few notes though: I think the problems with real-time lighting is typical of
>something I see happening with SmallPileOfGibs too. We all have long-term
>objectives on GtkRadiant, because it is much more rewarding to think about
>radical changes when planning new stuff. But the most important thing is to be
>able to write down a clean way to get there, something that other developers
>will be able to integrate with their own work. Experience shows that this
>requires a minimal documentation effort, and step-by-step implementation.
>Radiant is not a good example of clean source code. The software has been
>holding together because so far there was mostly a single developer working on
>it. The state of the source code and the developement model wouldn't scale to
>several developers and a more distributed effort. I have emphasized several
>times on the importance of documentation in the last few months, and I might
>get to a point where I will write it down as a FAQ item prior to any
>developement effort. But the summary is very short:
>- list the general objective
>- set up a list of features required and criticial issues that will have to be
>- propose an implementation (classes, interfaces, new files and headers) which
>is based on our current coding patterns.
>Pablo Zurita wrote:
> > After "wasting" around 3 months I have decided to stop my work on real-time
> > lighting because I haven't been able to get good results. That doesn't mean
> > that I will stop working on GtkRadiant, I will just move the focus to other
> > stuff and leave the real-time lighting for the future when I get more
> > knowledge on the subject, I will keep doing experiments but it wont be my
> > main focus. I know people expected real-time lighting but the result of my
> > code wasn't acceptable. Ok, cya.
> > _______________________________________________
> > Gtkradiant mailing list
> > Gtkradiant@zerowing.idsoftware.com
> > http://zerowing.idsoftware.com/mailman/listinfo/gtkradiant
>Gtkradiant mailing list