Page 1 of 2
Clipping and wallclimbing issues
Posted: Sun Oct 06, 2013 10:11 am UTC
by majkifajki
I noticed that levels in Unvanquished could make use of some extra smooth clipping. Whenever I try to wallwalk but sticking to wall/ceiling, I get a lot of trouble with "bumps". "Bump" is sticking out of surface, in perpendicular way, element.
Example:

Let's look at this from Alien's persepctie, which has glued itself to wall and it's wallwalking on that wall.

4 (FOUR) 90 degrees rotations of camera for alien's player to walk pass "bump". That's very, very annoying and makes playing with alien much harder then it should be imo. But there is solutuion:

Other approach, more precise:

And that issue is EVERYWHERE

I strongly recommend "clipping" treatment for all levels :p I know, PITA, but it would gameplay much more smooth.
Re: Clipping and wallclimbing issues
Posted: Sun Oct 06, 2013 1:29 pm UTC
by ViruS
Well, comparing the size of those ATCS pillars to a dretch is quite massive. If we just fill in the gaps, where will the trappers and dretches hide before ambushing humans?
Re: Clipping and wallclimbing issues
Posted: Sun Oct 06, 2013 2:04 pm UTC
by Viech
As already mentioned on IRC, our maps are Parpax, Platform 23, Thunder and Yocto. If you notice bad clipping on one of those maps feel free to document it here. The maps you show are placeholders we will only keep as long as we don't have enough new maps. Be aware that our mappers are currently inactive or working in other areas, so it is unlikely that the issue will get fixed soon.
Also ViruS is right, I personally would only clip gaps up to 8 qu, those pillars look like they have a thickness of 24 qu. Not only because dretches should be able to hide but also because it looks very odd when wallwalkers walk on invisible walls. Trappers wouldn't necessarily be affected (there are different types of clip brushes, some just block players). On IRC you said that the geometry should be changed so that clipping is possible and I agree that this is a good approach but doing so on every wall would look and play rather boring. Have a look at how I made the ceiling and one side of the long hall on parpax completely plain while the other side has a lot depth. I think this is a good compromise.
Re: Clipping and wallclimbing issues
Posted: Sun Oct 06, 2013 2:20 pm UTC
by majkifajki
This is good for deeper bumps

Re: Clipping and wallclimbing issues
Posted: Sun Oct 06, 2013 8:28 pm UTC
by Ishq
If you disable auto-wallwalk pitching, you might find it less disorienting. I play without wallwalk pitching because I find it annoying. Also, change wallwalk turn speed to "instant"
Re: Clipping and wallclimbing issues
Posted: Thu Oct 10, 2013 7:01 am UTC
by ViruS
Ishq wrote:Also, change wallwalk turn speed to "instant"
Is it because sometimes the server makes the smoothing jittery?
If I use pitching + instant, I find it hard to navigate when there's multiple small bumps with little degree of bend usually and I set smoothtime to 125-200.
Re: Clipping and wallclimbing issues
Posted: Thu Oct 10, 2013 4:52 pm UTC
by Ishq
Re: Clipping and wallclimbing issues
Posted: Fri Oct 11, 2013 10:06 am UTC
by ViruS
Yet the jiterry effect seems simultaneous with the world's rotating objects/movers and animations such as telenode rings, and seems to occur on servers that needs a rest...
Re: Clipping and wallclimbing issues
Posted: Fri Oct 11, 2013 2:48 pm UTC
by Viech
ViruS is right, long runing (64 bit?) servers indeed have an impact on client side calculations such as viewlocking and shader effects. We might want to talk to the ioquake3 people about the problem. If they don't have it then we can probably exclude the network stack as a source of error. We should also test with a NaCl server on a 64 bit machine to see if it's related to the handling of game code (maybe lcc does some bad stuff).
Re: Clipping and wallclimbing issues
Posted: Sat Oct 12, 2013 9:55 am UTC
by ViruS
Why not just host on a 32 bit system?
I don't see any penalties for it apart from possibly moving hosts...