Vis portals and detail brushes

[Megistos]

The other day, to have a look at how it's really done, I decompiled Karith's .bsp into a .map and took a look. The page I got the relevant command from said there might be a few errors in decompiling it, and I noticed things like unaligned textures.

The thing I noticed most, however, was that all the fantastic detail on the walls was made up of structural brushes, not detail brushes. Is this an error created by the decompiling process, or is Karith doing something weird? I always thought it was best to make brushes detail if they didn't touch the void - the way Karith's .map looked, there would have been thousands upon thousands of vis portals.

[Kaleo]

Decompile Transit.

[Taiyo.uk]

The thing I noticed most, however, was that all the fantastic detail on the walls was made up of structural brushes, not detail brushes. Is this an error created by the decompiling process, or is Karith doing something weird?

Yes, this is what happens when you decompile. All brushes in a decompiled .map will be structural.

[n.o.s.brain]

i dont see why decompiling ruins the maps like that, i mean, all the information about detail and structural and texture alignment is in the bsp, amirite? could some coder make a decompiler that would fix that?

[TRaK]

This thread here mentions a q3map2 patch by div0 that fixes the texture alignment issues with decompiling. I don't know any of the details though.

[epsy]

the structural/detail flag is stripped when compiling, simply because they are only useful at VIS portals creation and that's done at compilation stage

that makes it impossible to know if a brush has been marked as detail or not

now if there is a way to "guess" it, i don't know
it may work detecting brushes that "overlap"(well every brush overlaps these) VIS portals, trough false positives may be quite common

[Lava Croft]

Decompiling maps in an attempt to learn from them is a bad habit in general.

[n.o.s.brain]

lolyeah, it induces theft of brushwork.