tmg
Quake 2 Mapping Club 
Posts: 266
|
Post by tmg on Jan 13, 2022 6:50:04 GMT -5
Using trenchbroom latest version, and I go to compile my map and it leaks. I know where it leaks in a general sense but not exactly where it leaks. Also no leak point file is made so I can't load it into trenchbroom to see where it is.
I know it's in a tunnel, and my map is still very small because I spend too damn long nitpicking eeeeevery detail instead of just getting in and blocking stuff out to see how it plays roughly...
Anyway, I am using KMQbsp3, standard vis and ArghRad to compile, and really don't know why kmqbsp3 isn't writing the pointfile...
|
|
|
|
Post by knightmare on Jan 13, 2022 10:45:55 GMT -5
What output do you get when it fails to write a .lin file?
The only failure state I can find where it fails to write the leakfile is when there are no entities exposed to the void.
Also try a more standard qbsp3 to see if it also fails to write the leakfile.
|
|
tmg
Quake 2 Mapping Club
Posts: 266
|
Post by tmg on Jan 13, 2022 16:49:29 GMT -5
I took a screenshot of the compile window along with my map in trenchbroom. where the tunnel is, the doors are not func_doors, so there's no leak there. Something about qbsp3 not splitting a plane? Could be a bad clip-cut in trenchbroom. I haven't this problem in QERadaint for a loooong time. Attachments:
|
|
|
|
Post by Guggenheim on Jan 13, 2022 17:35:28 GMT -5
I took a screenshot of the compile window along with my map in trenchbroom. where the tunnel is, the doors are not func_doors, so there's no leak there. Something about qbsp3 not splitting a plane? Could be a bad clip-cut in trenchbroom. I haven't this problem in QERadaint for a loooong time. Looks promising. Will it be SP map? |
|
tmg
Quake 2 Mapping Club
Posts: 266
|
Post by tmg on Jan 13, 2022 18:38:02 GMT -5
I found the issue. there was a leak in the tunnel but for some reason I didn't catch it until I used QERadiant's internal compiler. It's fixed now. Could be a bug in your compiler Knightmare...
|
|
tmg
Quake 2 Mapping Club
Posts: 266
|
Post by tmg on Jan 13, 2022 18:38:48 GMT -5
I took a screenshot of the compile window along with my map in trenchbroom. where the tunnel is, the doors are not func_doors, so there's no leak there. Something about qbsp3 not splitting a plane? Could be a bad clip-cut in trenchbroom. I haven't this problem in QERadaint for a loooong time. Looks promising. Will it be SP map? Indeed it will be an SP map/unit. It's all I do. |
|
|
|
Post by knightmare on Jan 13, 2022 21:43:47 GMT -5
Do you have a backup of the bugged version of the map? You could try using the -verbosebrushes parameter to display which brush has the error.
I also reverted some changes in kmqbsp3 having to do with brushes having double precision. That could be related.
|
|
tmg
Quake 2 Mapping Club
Posts: 266
|
Post by tmg on Jan 13, 2022 22:33:50 GMT -5
Double precision...dude I got no idea what that means lol - my programming chops suck. I imagine the change is good?
I'll keep using trenchbroom for now, til another version of radiant comes out that I like...maybe 😉
Anyway. The compile works now that I've found the leak, and Arghrad's lighting is great too. I did try some updated tools off github but they just did not work the way that yours and argh's do, or those builtin compilers in radiant.
|
|
