What are you working on?

[Nahuel]

[Irritant]

[Nahuel]

[ajay]

Still tinkering away. I would be really grateful if people could read and respond to this blog post, it would help me lots:
http://earthquakemod.wordpress.com/2012 ... 12_2012-2/

[leileilol]

The IQM pipeline experiment of making vweps along with the player model has proven to be a failure


Even worse when I can't define skin paths explicitly at any part of it, so I have to drag my PCX's into the root.
Ideally i'd want every vwep to use the same texture page from a specially crafted 256x256 models/weapons/world.pcx for the world and vwep models to use. To save a lot of memory and switching. Q2's data is rather inefficient, a total media from-scratch-do-over just io see better crusher.dm2 benchmarks would be interesting. I'd do the same to projectiles (one texture would have laser, rocket, grenades, and "hellrot" bomb) puffs, and ammo pickups.

and before you ask, no, I am not making an elf game(although she's out of place, it didnt stop the Q2 DM community from being LITTLE FAERIES 86% of the time(and badly animated cows)). This is the only readily available animated 2.6x rigify model I used to test. Its pipeline is IQM > MD3 > MD2. I have to convert to MD3 first because apparently -mdlnobase doesn't work (Noesis).

I won't be able to use the IQM pipeline for some models like explosions in which a vertex morph sequence is ideal to drive them.

(I should really pack her and that shotgun weapon into a blend file so taniwha might have a look)

[taniwha]

The whole thing sounds like bad tools (the iqm->md3 converter and/or the md3->md2 converter).

I actually made a start on adding sp2 support to QF (starting with qfspritegen), but got fed up with compiler errors trying to compile Q2 so I could test things and thus went back to qfcc hacking (dags and sets and internal errors, oh my!). The compiler errors in QF's Q2 base are from improvements in gcc: better aliasing problem detection, dead code detection, etc. Also, I think it was the first time QF's Q2 was compiled for 64 bits.

Anyway, it looks like the tools aren't properly separating the vwep from the model. Also, not being able to specify paths just sounds bogus as that seems to have been one of the main points of Q2's format changes (at least for sprites, I haven't looked into md2 yet).

[edit] PS. I'd gladly take a look at the blend.

[gnounc]

(8:44:11 PM) Rich: -mdlnobase is working for me
(8:44:22 PM) Rich: Tell him it's intended to not insert the basepose into the mdl as frame 0

[ceriux]

[mankrip]

[ceriux]

[sock]

[mankrip]

In that specific shot, the bicycle really looks too dark, given that it's right under the light. The van on the right could also be a little brighter.

Metallic objects and objects with shiny painting like these reflects more light than things like concrete and asphalt, so they should be brighter. But I guess the correct way to do this would be to use bumpmapping.

[qbism]

color curve adjust

[ceriux]

ew i miss the old screenshots!

[mankrip]