Originally posted over at Q3W
Do you have any plans to attempt a distributed client version of q3map2?
Depending on the amount of usage required between clients, I know I for one would be willing to set it up on the net… It would allow people to get map compiles done extremely quickly if nodes were availiable on the net
Ydnar - Distributed q3map2 client?
Sorry
I’ve been concentrating my efforts at optimizing for single station compiles because that would affect all people, not just the select few who have bunches of hardware around for compiling. 
y
I guess it all depends on the kind of network traffic usage you’d be looking at. If a proper system were implemented that was fully functional over IP, and the network load was sufficiently reduced, a peer client would function just fine across the internet via broadband connections. The process would only have to facilitate the LIGHT stage, since in a quality mapping effort it’s the only stage that takes insurmountable time to complete. Plus it has the advantage of being theoretically divisible, no?
I would imagine that the option of an online ‘server farm’ for map compiling would be extremely tempting for mappers… And makes great use of the ‘free opportunity’ offered by the modern broadband revolution, and idle computing time due to excessive CPU advancement of late 
BTW If you cast your mind back to q3farm, it used relatively little bandwidth. And it had the right idea. The only problem is, it’s not up to spec with your current compiler.
It has nothing to do with network bandwidth. It has everything to do with programmer bandwidth. *
Q3Map2 is sufficiently complex that a full, proper implementation of this would be a significant project, even just the light phase. There are more pressing things on my time right now, even if I wanted to give it a shot.
If someone else wants to implement it, they should go ahead and do it. If it’s written well, I’ll integrate their changes into Q3Map2 proper.
y
- Except the transmission of floating point lighting data from slave nodes to the master for storage and creation of radiosity lights, which on large maps could be the gating factor on widely distributed compiling networks.
