Serpentine

That said, I do think we could do with a new post every 2 months or something. I think we should also take the opportunity to post 3 new WIP screenshots or something. Current work going on: Standalone - We're getting there slowly, Tels has moved a lot of materials/tables over from doom3 that we need. Wulfen's textures and a few others I have made are waiting to go in - I need to check what they look like vs. what we use them for however (boring, but no point rushing it, they're not placeholders). AI - Greyman and NH have been knocking out some issues with AI not seeing the player well enough in low ambient environments.The AI search hinting entities/areas - allowing mappers to better make things interesting. Models - Spring's retouching of the AI models, new teapot/clock/chest of drawers models from an external source, Noss' new models and work on as well replacement zombie model. Textures - Normalmap compression for massive savings in memory usage as well as increased performance on medium and low spec machines. Engine - Tels has been bringing our translation backend closer to the engine. I've cut down a lot of extra work being done by the lightgem - giving a very noticable gain in performance. General code cleanup and re-flow has lead to many minor gains, as well as larger gains by fixing caching systems. I'd check in with BD/Melan for some screenshots from the campaign - just need 1/2, maybe some better shots of the AI touch-up's by Spring and A nice shot from the WIP thread.

Wow, that's... :3

Looking good

Didn't they already cancel continuing with this? Tho that might just be some publicity stunt. After season two of the series — the whole thing went to shit, I regret not stopping watching sooner. The spin off show was unwatchable... not sure why people have crazy ideas about this.

3. The loooong candle one has always looked wrong to me — since it's a combo entity — you could either map it back to the shorter candle version and just remove the long one or something? The 'fat' candle issue is not something that gets to me — is it worth limiting them to only skinny candles? meh I don't think so. Could always have just used a knife to make a tapered bottom — as one does irl. 4. The flame glare has always seemed to low for a candle imho — however I think that's to stop the sprite from intersecting too far down the candle if it's on its side. 5. That might actually be a bug, I know that there's some rounding error with greys that are close to white, being managed into bright colours on nvidia cards (there's a comment somewhere in the engine code) so just check if the default pink is not specified somewhere.

Just a note : some of the lights and fireplaces flicker far too much, they strobe and cause visible tearing without vsync. That should really be toned down imho.

Yeah throw greebo a pm - he's a bit MIA atm however I think he can allow signups, failing that try modetwo

Yeah he's one of the contributers to Opacity.us - which I pasted earlier

How is the one in St. Lucia handled?

Oh, if you're doing texture work - there are some situations regarding cached images which can become stuck if using reloadimages/decls. I've often found reloadsurface to be better for refreshing both the material and texture. r_showSurfaceInfo is also handy for finding material names or to aim the reload (tho it does get annoying with the lightgem material/player in the way, I will fix that in the future). Just some info.

If you print more than like once a year, just get a cheapish office laser printer (most of them are very well supported by CUPS too). I guess some people may use colour... but it's something people usually say and then never do. That said, I'd like a cheap scanner with ADF, rather than this horrible Canon X700, which uses ~special~ protocols for everything and is just a pain in the ass.

Just use the retail release. We don't support the gpl release as a client yet - we will be releasing our own version based on the source in the next version.

Interesting, while it's not unheard of - the current release version has not been very problematic. Have you fiddled with your config or put your doom3 directory in any strange place?

Squeezing performance out of the lightgem - hopefully the changes are accurate and make it to release. Still have a lot of juice to squeeze out of that gem tho. Mmm. BD — Looks great as per usual, very... TES

I'm pretty sure there is rat clip - or maybe it's called whatever the level of AAS is for them? I'm certain I added it to some areas of my local copy of Thieves... let me investigate or confirm my memory is terrible. Edit - Confirming that I have vivid memories of something that doesn't exist... WTS frontal lobes.

@grayman - I'm too lazy to search - but iirc you did fix the 'rats can open doors' issue? In regards to rats, just make sure that if you use rat clipping, enclose the area to stop rat AAS from sprawling to other areas. Quite easy to forget since they're often used in strange places without combat AI.

Hey there, Sorry about the late reply. We're a really tiny team as is - while a few of us work on the code, it's by and large something that's driven personally - while we might be able to answer questions, there's very little in the way of what can actually be mentored at this time. I'd suggest rather than starting directly with the generic code, starting out with making a simple map, follow the tutorial, play around a little bit and try to get a mental model of how things stick together. Working directly with the code is important, but understanding the underlying structures is paramount to fixing things correctly . If you have already done all of that, or have prior experience - then I'm sure we can get you some access and see what happens, however to get svn access and the forum access, you'll need to speak to greebo. The bugs that you have suggested all sound reasonable, grayman works on the AI stuff, so we do have someone that can answer questions.

You need the games assets, the GPL release is only the game code and engine. Not the textures, sounds, guis, maps etc. Without those - it wont start.

Snape kills Dumbledore

Hmmm, are you sure you're using the 1.7.0 final, from the DR site and not one of the betas linked on the forums or something? I remember that bug, which was fixed before release.

@BC - If you're using the tri count from the assimp viewer, that is most likely giving you the values of how many unique triangles there are. This value will typically be 3x higher than the model will actually use (1 very per normal, xyz). Without the normals (as the exporter works now) the count ingame will be the merged value. You can also check how many verts are used ingame with listModels or printModel . I have a new version of the exporter, but I don't have time to do real tests at the moment, sadly I'm stuck doing real work for the next 2 days

K, there actually arn't that many, mostly random statues, arrows and weapons.

Rebuilding the whole mesh to drop 1k tris really isn't worth it, especially on a mesh where you cant really decimate cleanly or do too much automagically. I'm not saying that everyone should go crazy and push out high poly models, but rather to not do a pile of work for no real gain. Specially when the mesh only has a single material and includes shadows/cm.Pretty sure there are a good few props that use 3-5k already.

Tris are cheap, 3k is fine, it's not like you're making a forest of them. So long as there's a shadow mesh it's not a big deal.

The error you're reporting is just a timeout. Windows has some reg key for how low the threshold is to kill things that are marked as 'not responding'. If you search for related info on it, it's most often reported during level loading times and such in games. This is almost entirely caused by the OS, as the actual app is just doing a lot of non-interactive work.