In this section you can check out what I'm currently working in, practically in real time.
If some of this seems cryptic it's because the logs seen here come directly from my personal productivity tool prio, and are really just "notes to self".
Nevertheless, if you are interested in what I'm currently working on, this space will certainly be updated FAR more often than the news section.
19:31:04 swörd CLOCK: 0.1 days (01:23 hours)
21:05:56 swörd CLOCK: 0.1 days (02:08 hours)
18:15:16 swörd TASK: builder feedback implementation
18:15:13 swörd CLOCK: 00:35 hours
17:37:29 sandblox CLOCK: 0.1 days (01:53 hours)
17:37:23 sandblox TASK: step up
17:36:40 sandblox TASK: builder2 pipeline
15:14:59 sandblox CLOCK: 0.1 days (01:51 hours)
15:05:32 sandblox TASK: builder2 smooth lighting
00:21:24 sandblox CLOCK: 00:23 hours
22:28:52 sandblox LOG
FloodLight stuff today, based on storing a huge grid of light values. Will be working more on it, hopefully a tighter / faster interface from the mesh creation stuff.
21:11:54 sandblox CLOCK: 0.2 days (03:37 hours)
17:15:00 sandblox CLOCK: 00:26 hours
16:40:25 sandblox CLOCK: 00:28 hours
16:09:33 sandblox CLOCK: 00:51 hours
15:17:59 sandblox LOG
Experimenting with flood fill algorithms as an approach to lighting, with good initial results in a 2d mockup.
20:00:11 sandblox LOG
Removed some bad linear searching from the chunk mesh generation, which was there in an attempt to optimize the meshes.
Raw build performance is now much better, improving lighting performance is next.
19:59:22 sandblox CLOCK: 00:52 hours
12:20:00 sandblox CLOCK: 0.1 days (01:14 hours)
15:18:46 sbtf CLOCK: 01:08 hours
10:44:50 sbtf TASK: check latest art concepts
16:36:24 sbtf LOG
Oh yeah, finally got my head around fine grained control of the UV-generation stuff in Wings3D, i.e. "so THAT's how you manually place the cuts!"
This kind stuff really sucks, as it is just so totally technical. Also it totally destroys your spontaneity; I have found myself wanting to add details to the mesh halfway through the painting procedure, and of course in that case I have to deal with repainting the whole texture after changing the mesh, but also the whole uv debacle. Totally makes me averse to iteration, and that's really bad.
I'm seriously considering taking the plunge into Z-Brush 4; first the trial version of course, but I'm about 90% sure that experience is going to make we want to buy the whole thing. UV-Master seems to solve all of the above in a very automated fashion.
...which brings me to the "animation problem". The things I've seen in rigging and animation tutorials (in for example Blender) just make we want to quit altogether; there just doesn't seem to be anything that measures up to Z-Sphere armatures and all of that stuff.
Also been experimenting with more straight up low poly modeling in Wings, looking into super-minimalist stuff around 100-200 tris....
16:28:21 sbtf LOG
Even more experiments with new enemy styles (visualizations), this time with animated sprites on the same texture. This brings to mind the need for more particle effects, a system for which could of course be re-used from ACE...
16:26:56 sbtf CLOCK: 0.1 days (01:58 hours)
19:19:52 sbtf CLOCK: 00:17 hours
19:19:48 sbtf LOG
Been spending a significant amount of (unlogged) time looking into Z-Brush 4 and comparing it to what I can find online in terms of character creation pipelines, from both an aesthetic and a technical point of view.
It really seems to me that Z-Brush represents a paradigm shift in terms of asset production, and what is really attractive to me is the tendency of the evangelists to emphasize speed and ease of iteration.
The Z-Sphere armatures in particular look very interesting, especially for a "non-artist" as myself. I have indeed acquired rudimentary "traditional" modeling and texuring skills over the last few years, mainly in Wings3D, but I don't like thinking about UVs or painting "flattened" models, and have zero experience with rigging and / or animation.
I have been trying out Sculptris, the free program that Pixologic have acquired from Dr. Petter, and while the whole sculpting thing is amusing I found that the mesh painting and Photoshop interoperability were very useful.
To be continued...
09:12:43 sbtf LOG
Procrastinating about portals / culling turned into some fun experiments with Quake 1 .mdl files. Love that stuff!
02:30:00 sbtf CLOCK: 0.2 days (04:57 hours)
20:34:56 sbtf CLOCK: 0.1 days (02:58 hours)
17:20:05 sbtf CLOCK: 0.1 days (02:11 hours)
00:15:17 sbtf CLOCK: 0.2 days (03:51 hours)
20:15:32 sbtf LOG
Got blendmesh enemies in there, using the Fast Creep for now. Number of drawing calls is now a real issue, need to look into some kind of robust visibility solution, thinking pre-calculated.
20:14:09 sbtf TASK: enemys as blendmesh, correct shader
20:13:33 sbtf CLOCK: 0.1 days (01:56 hours)
17:36:52 sbtf CLOCK: 0.1 days (01:47 hours)
23:48:07 sbtf LOG
Managed to move the Adrenaline concept to the Model, which is where it should be as it affects avatar speed.
A lot of talk about the conceptual direction of the enemies / beasts / aliens. Need to really hash that out some more. Right now there is a new proxy available, toggle style with B (2 total).
23:46:23 sbtf TASK: adrenaline to model
23:46:20 sbtf CLOCK: 0.2 days (05:22 hours)
19:17:53 sbtf TASK: rotating tracker gfx
23:03:38 sbtf CLOCK: 0.2 days (04:44 hours)
23:03:34 sbtf LOG
Seeker / Breach adjustments today, looking pretty promising.
Also added a typical directional motion tracker, absolutely easier to survive with that. Several upgrades come to mind, firstly the tracker itself, but also range upgrades as well as the "persistence" of blips on the tracker itself.
One could also play around with the frequency of the "sonar pulses"; in the movie Aliens the constant tick / blip of an "empty" motion tracker was kind of creepy and stressful in and of itself.
23:00:24 sbtf TASK: gui: directional motion tracker
20:21:04 sbtf TASK: entrypoint -> breach
19:48:57 sbtf TASK: mission log
17:45:09 sbtf CLOCK: 0.1 days (02:19 hours)
17:09:05 sbtf TASK: bug: seekers push each other through walls
17:08:14 sbtf TASK: exit / complete level
16:07:40 sbtf TASK: sprint speed based on adrenaline
14:59:38 ace TASK: review: thegeeksftw.com
14:32:38 ace TASK: review: jay is games
23:39:10 music TASK: payout to per
23:32:09 sbtf CLOCK: 0.1 days (02:19 hours)
23:32:02 sbtf LOG
Have been experimenting with actively seeking enemies (Seekers). Right now they are basically the same as the previous incarnation (Lurkers), but this in itself poses the problem of just way too much aggression. Need to tune the speeds.
The addition of randomly occurring "entry points" being punched through the ceiling really changes things around. Firstly the "need" for sentry guns comes to mind, but another way to deal with them would be to simply close doors around them. This doesn't help if you need to be in the same "room" as they happen to spawn in, rooms being much more common with the current level generation algorithm.
In the Space Hulk games I've seen previously, enemy entry points are always in spots that you needn't go, i.e. the end of dead end corridors. This may be a necessary mechanic to keep things reasonable.
I'm thinking that doors should stop enemies, but that they should also be able to break the doors.
Can you say emergence?
23:26:53 sbtf TASK: seeking enemies
23:26:50 sbtf TASK: alien entry points
23:26:47 sbtf TASK: mechanic: increasing (over time) population of actively seeking enemies
20:57:20 sbtf CLOCK: 00:51 hours