Entity func_brush leaked

Hi all,

I'm getting a weird error on compile (after a lot of work on my map, so I'm not surprised).

Entity func_brush leaked. Okay, gotta find+seal a hole. loads pointfile

Pointfile starts in midair, blank space and ends in midair blank space. From nowhere to nowhere. A straight red line maybe 500 units long, starting in blackness and ending in blackness. I have two func_brushes in my map and both of them are completely sealed up in seperate rooms away from the pointfile line.

I haven't an idea why this has happened. Anybody?

Thanks.

---

-- midkay

It just means that vbsp started drawing the pointfile from the func_detail. It always starts at an entity. If there is any leak in your map, every entity is "outside world" as there is no clearly defined inside.

The leak may be somewhere else. Sometimes even the pointfile is broken, but try taking a closer look. If you want a new pointfile, move a few entities or hide them, slightly change a simple brush and see vbsp making a new one. Try hiding the two func_brushes and see what vbsp does.

Use Hammer's check for errors tools to make sure there isn't a rogue func_brush entity that isn't attached to a brush. You should also make sure you don't have any multi-brush func_brush entities that have their origin outside the hull, as I imagine that could possible cause a problem like this.

If you have a traditional leak and your pointfile isn't helping you out, go back to more traditional methods such as the big block method (covering half your map in a solid block, so that if it compiles you know the leak is in that half, then repeat by covering half of that half, and so on...).

Still no joy? Put a world brush over the area of the claimed func_brush. Not a "nice" way to deal with the problem, but it should at least deal with it!

Worst case scenario, you've got a non-existant leak - a claimed leak with no apparent cause. I've had a few of these lately, and they are exceedingly annoying. Normally you'll get these indicated by your pointfile going straight through a solid world brush, and normally everything is completely sealed. Doing things such as rebuilding the brushwork around the pointfile area can normally fix it, though I've even managed to fix one in the past by changing a texture for god's sake (honest to god - two versions of the same map, one with one texture one with another, neither some special type, and one causes a leak). Sometimes this engine really infuriates me :sad:

---

[img]http://card.mygamercard.net/sig/Default/reno84.png[/img]
Designer @ Haiku Interactive | ReNo-vation.net

Thanks guys, I was in a rush to bed last night and wanted to see if i I could get a few answers.. I hadn't much time for experimentation.

reaper47: "If there is any leak in your map, every entity is "outside world" as there is no clearly defined inside." - right, but I usually get a line leading from something to the "outside" rather than just a random line in the middle of nowhere pointing at nothing from either end :smile:

ReNo: Switching a texture, huh? Haha, wow.. appreciate the comprehensive response.

Probably be back in 15 minutes with "I did it!"..

---

-- midkay

I did it! Close to my prediction... around 15 minutes after my last post :wink:

I loaded it up again and had the same problem.. so I decided to hide the two func_brushes I had in the map and that changed the pointfile's position to a usable one: a prop-static leaking out through one of the roofs. Alright, so I sealed that one.. then I got a func_physbox leaking from nowhere to nowhere again (versus a func_brush before).. I only have one physbox in the map and the line was drawn like 512 units to the right of it, from nowhere-to-nowhere.. around here I got the idea that vvis was kind of "hanging on" to its old position (before I'd moved entire buildings over). So I moved the physbox "to world" and then back "to entity" and reset it up.. fixed that error! Went back and did the same to-world-to-entity to the two func_brushes and everything's A-OK.

Weird.. like it was drawing pointfiles from where those func_brushes were instead of where they were. I did try moving them a few units back and forth and recompiling but that didn't help.. I had to completely "reset" them.

Thanks guys. :smile:

---

-- midkay