Testing large maps with FitzSDL-0.85beta on my desktop i'm getting quite a few crashes, sometimes every 5 minutes. And it just killed Xorg/Linux on my laptop while playing "timedemo demo1".

Hmmm... playing "-nosound" and have yet to get a crash. (Still.. negations can't prove a theory). 

I played multiplayer some days ago at office with FitzQuake0.85... it was running like a charm till it hang up, without any possibilities to unblock the game, except Ctrl+Alt+Del, and kill the process... and no error message at all... Maybe a network protocol issue, or something else, not sure...
Did you ever had heard about such issue before ? 

no, never heard of that. Can you tell me what map, what mod, and what you were doing at that moment? 

Hang up happened twice with 2 different mods: Reaper 0.80 and Killer 0.90.
The map was THF the first time, and DM3 the second time.
I was just running (bunny-hopping) and firing spikes with SNG on an another guy the first time, and the second I was ambushed with RL..
I am actually trying to educate my collegues at Quake DM at office :) 

kinda liked the laggy monster movement animation v0.8 provided more than smooh v0.85 one. any way to set the imprefect monster movement at v.085 pls? would really like to downgrade this feature. thanks 

r_lerpmodels 0
r_lerpmove 0 

the documentation states that they both default to 0, but it is wrong, they default to 1 :P 

i'm not certain, but i believe i've found some kind of odd bug to do with .blocked for doors.

this happens in fitzquake, but not in aguire's glquake, which leads me to believe it's a bug with the engine.

basically, you get crushed by a door when you're standing on top of it while it is traveling downwards. this shouldn't be able to happen since you can't block a door in that manner.

the instance where this was happening was originally in a map that was breaking many limits, but i've transplanted the bit of map into a seperate map and the problem still occurs...

any idea what this is? i can give you the bsp/map if you're interested. 

Have you increased your max FPS at all? At very high framerates glitches like that can start to appear, and it might be that aguirre's engine keeps a hard cap on server frames. If you can reproduce it at 72 fps then it isn't that. 

if you can upload the map somewhere and email me the URL, that would be useful. (or just mail it to me if that's easier for you.)

Also what Preach said. 

my host_maxfps was set at the default 72.
i set it to 10 and 60 and the bug didn't happen.
setting it to 71 (and 70, 69) makes it happen at a different spot, but setting it to 68 makes it go away, however, i get a single 'unstuck' message as the door hits the top of it's movement so i think it still has the potential to happen.

i'll send you the stuff in the mail. 

should have thought to test this sooner, but the bug also happens in normal glquake.

i'm pretty confused right now, because i've never had this problem before. i'm trying to figure out if i've built something different somehow, but as far as i can tell, it's the exact same combo of entities i've used many times before. o.o 

Yes, this happens at the start of Drew's speed map and at the start of White Room as well. Any kind of long distance door like that which you ride downwards has the potential to start hurting you for no apparent reason. It sucks. 

it's good to know it's not a bug i personally created. Still good to know about it, though. 

So I'm finally playing warpspasm on fitzquake .85 sdl mac beta, and it's lovely (the way it was meant to be played, no less), but I notice that the background music is directional; that is, I only fully hear it facing a certain direction.

I don't know if this is an issue with fitz or warp, but I confirmed that it doesn't happen playing warpspasm in Darkplaces. 

It's an issue with all engines that don't do something special to fix it. Basically all sounds in the level have an origin point, even if they are attenuation = 0 sounds. I don't know what darkplaces does specifically to fix it, but it must treat those sounds specially, forcing them to be the same volume in both channels. 

I hacked it by putting four sound entities in each level, at the primary cardinal points.

The best fix is a fair bit of messing about - null sounds and playing them from outside. But nothing breaks immersion like alt-tabbing.

Basically sounds in quake can only be heard when seen. ATTN_NONE makes them full volume despite distance. 

I suspected it had something to do with attenuation. It didn't really detract from the atmosphere anyways, I was just curious.

Keep up the good work and all; I really appreciate and enjoy it. 

metlslime, I made others tests in multiplayer.
I played with 5 collegues, and no errors / hang up / issues occured during more than an hour, hence the fact I think the issue comes from the mods itself... weird... 

That's why more engines need Ogg Vorbis cdtrack emulation (with cd play # support). 

what causes that? every so often, a wall torch (or quoth brazier) is full dark, even though the floor directly under it is very bright. 

... because it is in a wall ? I already experimented this dark-ish behavior with impaled corpses... maybe same issue.. though... :P 

here's a screenshot: http://necros.quaddicted.com/temp/dark1.jpg

as you can see, the entity is inside the level (the origin on the braziers is up where the flame is) and the floor underneath is nearly full brightness.

it doesn't happen in aguirre's glquake though. 

so this is a fitzquake-only bug? Perhaps it's due to entity lighting interpolation (the code that uses the floor lightmap brightness to determine lighting, in fitzquake interpolates between neighboring lightmap samples.) 

i am hesitant to declare it's a fitz only bug since i can only really test in fq and aguirreGlquake (henceforth aglq, curse you aguirre for never actually naming your engine!). do the tools used to compile (in this case, aguirre's light util) make a difference when the engine reads the data?

otoh, i cannot ever remember seeing this happen in glquake. the 'zone' where an entity is full dark also appears to be more than just a single pixel or whatever. i've had it continue to happen up to 16~ units away from where i first noticed it.

in any case, if it's interpolating, then it's definitely not getting some lightmap data at all since that entire area is full of light.