I'll try it.

In the meantime, I've confirmed that the worst part of the dirtmapping bug is now fixed; liquids won't cast dirtmaps on the world anymore.

They still receive dirtmaps, but at least this problem can be worked around by func_detailing them with _dirt -1: screenshot. 

Wow, 0.15.7 is incredibly fast! I don't have specific times, but it speeds up light on my map by a huge amount. No idea what kind of wizardry was performed for this, but it's really cut down on the iteration time for my map. Thanks so much! 

Pritchard - awesome :)

got some bugs reported and fixed already (to be in the next release):
- the "unmatched" target warning was broken
- skip-textured func_'s with "_shadow" "1" were broken (for making invisible shadow casters)

This one was tricky due to how I migrated light away from using the BSP for raytracing. The workaround I implemented has a limitation that the entire bmodel has to be textured with "skip", otherwise the skip textured bits don't cast shadows. Hopefully this covers the common use cases for that trick, though. 

That looks pretty damn cool. How did you get the water to draw shadows like that? 

thought I would report it just in case.

vis.c: In function �LoadPortals�:
vis.c:1120:9: warning: ignoring return value of �fscanf�, declared with attribute warn_unused_result [-Wunused-result]
fscanf(f, "\n"); 

Using independent surface caches for texture & lighting, and combining them in realtime.

The drawing routines of liquid textures (and now, of power-of-two textures too) were mixed with the drawing routines of regular surface-cached routines, combining the texture & lighting in realtime before blending the result into the framebuffer.

It's not difficult to do, just really tedious. Involves modifying and rewriting a lot of code. 

@ericw: I'm curious if it would be possible to add support for func_viscluster brushes in order to negate large open spaces. I'm of course assuming that vis leafs automatically chop up empty space into leafs every 1024 units. https://developer.valvesoftware.com/wiki/Func_viscluster

Spikespasm:
Host_Error: Mod_LoadLeafs: 121741 leafs exceeds limit of 70000. 

AFAIK, there is no chopping of space every 1024 units in this qbsp, that feature was added in quake 2's.

func_viscluster - this sounds like purely something to speed up the vis computation. It's probably possible to add to Q1 tools, it wouldn't help with reducing the number of leafs though. 

This just popped up now as I'm working on my map. Bounce lighting started generating random black spots in this one room of my level:
screenshot 1
Here's how it looks with bounce lighting turned off:
screenshot 2

Any idea what's causing this? I was changing my light entities around, but I can't figure out what I did that caused this. Has anyone even seen this before? Help would be greatly appreciated :s 

At least somebody's happy about this...

I tried running the map through 0.15.5 and the issue went away. I'm guessing then that this is an issue caused by the new approximation methods in 0.15.7... 

I hate making a third post like this, but I tend to write in a very "stream of consciousness" manner... Anyway, I tried the -novisapprox flag on the command line and the issue persisted, so that's not the problem. 

I've never seen that, definitely a bug. mind sending me the bsp? 

Sure, I just shot off an email.

I'm guessing it is some kind of regression from 0.15.5, although I couldn't say what other than that it seems to be bounce-related. 

I got an email back from eric and the bug was fixed. Thanks for the quick response! :) 

Could somebody help me with this little thing that annoys me so much :

https://drive.google.com/file/d/0BwxYkKdSD855Mzhuc3ViNVo5YzA/view?usp=sharing 

can you turn off texture filtering in game and see if this still happens? 

bilinear filtering -- the last row of pixels blends with the first row of pixels, since it's a tiling texture. 

In the black/red colored view it looks like the texture is mirror imaged. Is it one continuous surface going from from red to black or is it actually two different brushes? 

It is same exact texture but one of axis is mirrored "-1" it is 128x128pxl, and those brushes are exactly 128x128 and next to each other.. so filtering is making this to happen, that is looks like it is.. like 0.5 pixels or something? 

..so filtering is making it to look, like it is 0.5 pixels off or something* 

get up close and toggle between gl_nearest and gl_linear and you'll understand what's happening. You'll need to hide this pixel-wide blend by not using the full width of the texture on that polygon. Either scale it up slightly or make the polygon slightly smaller. 

Thanks metslime, I haven't tried scaling up yet - maybe that will work. 

Any time there is a difference between the textures on adjoining surfaces there will be a noticeable line if GL filtering is used.

I call them GL seams, and any kind of rotation, x/y shift, flip, mirror, etc. will cause it. Even just pixel color differences will cause it. 

It is interesting.. it really looks like there isn't different pixels in this case, if I understood what you meant by that. I have no clue how GL filtering works, so basically I can't avoid these things to happen.

For example I should create texture that points out on every main angle? Then I should have seamless street texture to fill everything in middle.. I was trying to create streets which has some messy garbage on both sides but middle is a lot cleaner. It really seems like that brown/black line in middle is like texture's starting pixel line, because its end and beginning has different pixel colors. The way GL filtering works.. does it look every texture individually and doesn't care about rotation, x/y shift, flip, mirror at all and that is why something like this happens sometimes? 

Does it try to blend blend pixels with next pixels.. if I offset it by 1 it is already in beginning, so it tries to blend with those pixels?