We NEED more developers!

[divVerent]

Being familiar in Lua is a good start, if you also know JavaScript it'd be perfect, as both QuakeC and JavaScript are some sort of "simplified C".

I started making a wiki page about the QuakeC language (not yet covering its functions, as many other sources already do that), and I suggest you reading that for a start, and then trying to read through some Nexuiz source and try to understand it.

I'll later extend the wiki page by links to other resources, and some example code for "typical tasks".

http://ouns.bplaced.net/dev:quakec

[KillaGrunt]

I don't know if this will help but I will try to help as much as i can. Once i learn it. Perhaps you could teach me? (just suggesting) If so; only enough to help.

[Asraniel]

I don't have time at all to code for nexuiz right now, but i have a suggestion.

Create so "junior jobs". Small tasks that would be "nice to have" in nexuiz, that somebody new to the codebase could do. I think it would be a nice way to introduce them. And also for the others, a general TODO page is certainly nice, because especialy new people will hesitate to add new features if they don't know if they are really wanted or needed

[take_this_cup_of_poison]

Ok, here's an update (also I've just mailed all the people in the math department at MIT and a good number at NYU to see if we can find someone to make a better vis algo):

The ffxml files are all gpl.

[take_this_cup_of_poison]

Here's the email I sent in search of someone who could make a more efficent vis alorithim:

Topic:
Need help on a visibility algorithim.

Code: Select all

I was wondering if you could help on an algorithim problem? An opensource 3D game called nexuiz uses a(n) (also opensource) map compiler called q3map2 to compile it's maps into a BSP tree and compute visibilty and lighting. Currently the visibility stage uses an algorithim called "PassageFlow" to compute the visibility tables. The thing is, for any slightly complex map the time to compute visibility seems to rises exponentially. The original maps that this program was used for were fairly simple so this was not a problem and was not fixed. As this program is used for larger and more complex maps the problems increase greatly (2.5 day vis times are not terribly uncommon for large maps, even with large blocksizes set). Could you perhaps help to find or develop a more efficent algorithim?


The program's source is available through svn (along with the map editor) at http://emptyset.endoftheinternet.org/svn/radiant15/trunk/tools/quake3/q3map2/
and the visibility computing algorithims are in these two files:

http://emptyset.endoftheinternet.org/svn/radiant15/trunk/tools/quake3/q3map2/vis.c
http://emptyset.endoftheinternet.org/svn/radiant15/trunk/tools/quake3/q3map2/visflow.c

(On linux or other unix-like systems the command (svn checkout http://emptyset.endoftheinternet.org/svn/radiant15/trunk/ NetRadiant) can be used to retrieve the whole package, it uses the ./configure; make; system to build)

The devs of nexuiz usually are avaiable at http://alientrap.org/forum/ and there is a development forum in there (Divverent is currently the main developer). Everything is opensource and under the GPL.

I hope there is a better algorithim other than n^n for this problem :P



Please join the emailing campaign (see if you can get media, code, etc devs for nexuiz, you can use my mails ofcourse if you want or make your own).

Also, for what it's worth, I've made bumpmaps for some of my celing textures.

[take_this_cup_of_poison]

Div, or anyone else in the know, I got a response on the vis algo email. Could you email him back with the correct info. In the meantime I will try to respond
(here's the email I recieved):

From: "Peter Shor" <shor AT math.mit.edu>
Dear Miguel

First, it would help you to ask computational geometry people this
question. Second, nobody is going to want to go through the code to
understand what the algorithm has to do. If you could just tell us what
task the algorithm is supposed to accomplish, you would be much more
likely to get an answer.

Peter Shor

[merlijn]

I would love to help on the coding part, but am currently struggling with a huge shortage in time. Hopefully things will change soon and I can attempt to be of some help.

What might be a good idea is to apply as a mentoring organization for Google Summer of Code (GSoC), which will likely run again in 2009. More information can be found on http://code.google.com/soc/2008/

[divVerent]

I doubt we have the needed number of active people in our team to also mentor students in a GSoC project.

Obviously, we DO have a severe lack everywhere.

[take_this_cup_of_poison]

I got a 58 on my calculus test (calc 1, 2nd test of 4 or so). I went to every extra help (twice a week, 2hrs each time) and did my homework.

I got a 58 out of 100.

This, I believe, shows that I am far too stupid to contemplate programming for nexuiz. So here's my coin in the hat of "not going to help because of some excuse but posting here anyway to tell everyone".

58

Maybe that should be my signature.

HELL! I even looked into the quake C code a bunch of times (including today) and can't make heads or tails of it. The only code I understand is my own perl, C, and PHP code. All other code I have no idea what is going on.

I got a 58 on my calculus test.
58.

[ai]

What now? Did you say you got 60 on your test?