What were you expecting to do with an invisible, non-solid brush, anyway?
You guys realize that you don’t actually need to run VIS to generate a pointfile, right? VBSP is what actually creates visleafs., and once it detects a leak, VVIS will never run anyway.
As for toolsblack, just make it a func_brush with solidity set to never. Or, you could make it a func_illusionary if you’re feeling classy.
Eh, doesn’t make a great difference anyway, fast VIS takes about 1 second to do its thing, leak or not.
I also don’t get the problem with the toolsblack. Why on earth would you even want a toolsblack that doesn’t act solid in the first place? What’re you using it for?
I really am not sure what I would want these black textures for, but now I’m pretty sure they shouldn’t be there as is.
What I got when I decompiled the map for the first time was that these textures separated the rooms/large areas from the corridors and there were usually fitted to the adjacent brushes. I actually kept wrestling with this fact even after I told everyone I had solved it, since there’s no way all of those placements could have been coincidence.
So whatever happened, I was given all these brushes and looking at them, they came to me as world geometry. I had assumed that since they were a tool texture that they would have some special properties (they do; they don’t interact with light).
I think they are supposed to be hint/skip brushes. My first thought when I saw all the placement for all of these textures was that they must have something to do with the map optimization. It just goes to show that the decompile process isn’t perfect, and that I have to lookout for any other textures that might have been changed too.
Haha. I was going to tell everyone that this whole thing was your idea. 
I can’t imagine how much work it must have taken to make that model, because I have zero experience in making models.
But in full honesty, this work owes a lot more to you than just using the led light models. Heck, I wouldn’t even be here if you weren’t mapping. I should be plastering your name all over a billboard somewhere.
Looking forward to the development of this project. I’m all for recovering as much of HL’s OaR as possible and ramping up as many additional playing minutes in OaR that can be achieved.
Rest assured, I’ll gladly provide you with detailed/useful feedback, once you’ve released your initial videos and alpha builds and such.
Wonderful! You gave wonderful feedback to TEXT is his own thread, I am sure you’re going to be really helpful to me as well~
I did sorta promise that I would grab a couple of in-game screenshots so you all could see the environment. It’s a bit late, but better late than never.
Here is our box track.
With the walls we can see the second and only variation on the exit junctions (the other has a column and then a door behind it). This left a lot of plain concrete, so to keep that from being too boring I added a crack decal to it.
I can also tell that I’ll probably be deleting the front pair of lights, as I think the tunnel is too bright. It also keeps emphasis away from the pair of boxes.
And the houndeye room.
I showed this room in the editor screenshot that I took, and here is how it looks ingame. Interestingly enough, the tram already had some track paths up here, so that almost seems to suggest that they had planned for the tram to come up this way at one point and then never did anything with it.
I might have to move/delete a few objects here. I am sure there will be players that shoot the explosive barrel when they arrive here. That actually presents a problem because it can shower the track with debris. The tram takes up most of that space, so add to the fact that you have a couple full-size barrels being pushed by the tram, it just doesn’t look good.
There are also a couple things that I have found that are present in the stock maps too (on a related note these next two pictures are from the stock c2a2a), and I’ll probably make a bug reporting thread about them. In the meantime, I will probably try to fix these errors in my version of the map.
I’m not sure why this happens (because I’ve tried looking at this area in Hammer and everything seems to check out), but everything under the water is nodrawing.
I borked eeltric pussle. Halp pls lol.
I don’t expect most players to do this, but this is completely possible through normal gameplay. Surveying the scene also revealed a couple potential inefficiencies in which the puzzle is handled too, which I won’t detail here.
I lawf u
never stop plz :3
Nice progress.
OaR is turning out to be one heck of a journey for Gordon. It looks good .RK!
If stuff under the water is nodrawing, and/or you’re getting funny render errors in a few places, you probably have left in some areaportals/occluders from the decompilation process, and thus the areaportals are leaking. Decompiling completely breaks them, you have to rebuild them all from scratch. Otherwise weird things happen. Areaportal leaks do not break VIS, but they break the areaportal in question as well as any water on the map. It’s a pain in the arse but there’s nothing you can do about it. Also make sure no areaportals intersect the surface of water. The Source engine HATES that.
Indeed.
How are you going to have the tracks switch thing work? Are you going to have the switch right next to the junction or do you have to traverse your loop to find the track controls? Also what are you using for controls…something like Textfamguys design or…?
Lastly when will this be done because I would LOVE to test and play it!!
I almost figured you might bring that up at one point since I remember reading that you had a problem with these when you were remaking Surface Tension.
This doesn’t seem to be the problem for three reasons though; one is that this is the only spot in the map where it happens, everywhere else is fine; two is that after checking through the decompiles (via texture finder and entity reports) I found no trace of area portals or occluders; three is that this problem exists in the stock maps, so apparently it’s been there before I ever got to it.
I just don’t know, maybe something else is happening with the geometry there that causes that.
You will have to traverse the loop to find the track controls. There will be a room with a switchboard that handles the entire area; there the player will be able to throw the switch. For this, I’ll probably use a lever.
And I don’t see this taking longer than 2-3 weeks for a couple of reasons. It’s not a large project, and I have the entire next week to work on it too. I reckon the alpha build isn’t too far off!
I knows I knows, the double post. But I have hit a fairly substantial setback.
For those of you who don’t know, BSP has a certain limit to both the amount of brushes that can be within a map, as well as how many brush faces can be within a map. The developers wiki says these limits are 8192 brushes and 65336 faces.
Now, for some reason or another, I have surpassed one of these limits. I am inclined to say that it is the faces limit. I’m not entirely sure, but as it stands now the map cannot compile.
This puts me in quite a pickle, because it means I have to cut down on the brushes. Problem, the security station is not even done yet. It has a pretty basic design, sure, but I haven’t finished decorating it yet. The pipe room is done though and so are the sections of track.
I have a few options, some of which I already plan on implementing.
- I can simplify some of the more complicated shapes. There are a few places within the map that use a lot of unnecessary brushwork and polygon work for very simple things (some of which are my own) I can simplify these objects to reduce the brush and face count. None of this would be visible, as it would simply be taking things and making them much more efficient.
- Turn some brushes into props. Selectively of course, as models handle lighting differently than brushes to. But this could actually be very effective. There just happens to be this one area (the electrical room that you can’t get into) that if I turned the contents into a prop, I would save 145 brushes and around 900 faces. That’s a fantastic trade-off. And there might be other things too.
- Outright delete nonessential things. These would be very minute details that you probably would not notice were changed. There is one area in particular where I could deleted about 80+ faces and no one would be wiser. Admittedly though, I wouldn’t be too happy about having to remove dev design, so I might save this for a more extreme measure.
- Get my nodraw on and tag whatever I can. These textures do not count towards the limits so I should use them whenever I reasonably can. Obviously I can’t texture anything that the player can see.
- Move the transition area. The elevator heading out of the level is pretty expensive. But… I don’t think I would do this unless I needed an absolute last resort, because not only would I have to edit the next map, but it would also destroy any compatibility with Uncut. Don’t count on me doing this.
I will be fiddling around with the design, so be sure to stay tuned. I’ll probably report back tomorrow with my results.
I’ve already converted that whole electrical box setup in the side room on A into a model for OaR Uncut. Just dig through OaR Uncut’s models and you’ll find it, or you could decompile A1 and take it from there, I think it’s the same as the one on A. God only knows why they decided to make that brush based, the only reason I can think of is to use the lightmapping in conjunction with the flickering light that’s there. Seeing as you’ve already said you’re going to use something or other from Uncut, take whatever you need.
Nodraw won’t help you much. The BM Dev team were pretty studious with their nodrawing, so it won’t do much unless you’ve been a bit lax with it.
Converting tunnel details (the side cladding on the walls mostly) into models won’t work either. That was an approach I already tried when I was hitting brush face limits on A2. It looked horrible. The vertex lighting was not kind to it.
Moving the transitional area won’t just break compatibility with everything else, it will require you to edit the B map too. Certainly not a good compromise I don’t think.
Converting brush based details to models is your best bet, and it’s not terribly difficult or time consuming with the use of Propper. Or outright change the designs of certain areas. It’s surprising how much you can simplify sometimes.
The BM Devs were already pushing the limits when they compiled the A map. Your hitting the limit so suddenly suggests to me that maybe there was already some kind of extra track or design element in place where the phantom track leads to, which they then scrapped in the name of entity limits. It would explain the phantom track present on the A map. You don’t really need to create a track when you’re planning on making an area, because making a track takes the better part of about 10 seconds - there’s no need to really do it in advance. It suggests to me they had something there and then removed it later on, possibly to account for limits.
Good luck overcoming that issue anyway. It’s annoying as crap to deal with, and its just one of the many annoyances that comes with dealing with an outdated engine.
It’s always a good idea to try and use props when you have lots of complicated geometry, especially when you find that you have a ton of complex func_detail brushes that have lots of verts/faces. Usually static prop lighting can be resolved with careful placement of info_lighting entities. Can get very annoying though.
Another option is to split off the area with the expensive geometry into another map, and link them with a level transition. This has the potential to get very cumbersome though, especially since you plan to loop the tramride.
EDIT: Probably a better idea for me to read up everything before posting. Ah well.
It’s kind of a huge deal when you’re an impatient soul like I am. 
I kinda wish I didn’t have to. But then consistency would be lost. …Okay, we all know it’s because I’m lazy.
But I took your advice and snagged the powerbox model and I replaced it in A. I also reconnected this pipe that I never really liked anyways (and this allowed me to delete about 80 small triangular brushes). I thought these two things would save me about 1000 brush faces.
I wasn’t even close. According to what vbsp tells me, those two details alone slashed 5000 brush faces. Holy shit.
I might just have enough to place both missing sections in their entirety now, although the fact still stands that the office is not complete yet. So I might have to keep going and look for other things to optimize.
EDIT: [redacted]
EDIT: Ran into a problem with Propper. Have it figured out now.
Excellent. IIRC (though this may be the triangular brushes you were talking about fixing), down by the power panel near the water trap on the A map are a few pipes which are broken/disconnected. I remember about 3 of them. Those all have this REALLY WEIRD conical shaped end brush attached to the ends of the pipes which is ridiculously complicated and very small. They’re designed to make the pipes look like they’re not hollow. They’re extremely detailed, and I’d imagine by removing them you could save many more faces. I have no idea why the BM Devs decided to make them so damn detailed, it seems literally impossible to tell the difference between that really detailed end and just a simple flat cylinder, in game. I’d imagine many faces would be saved if you removed all instances of that brushwork, with practically no quality loss whatsoever - if you shine a torch on them in game, zoom in and look really closely, you cannot see any of the detail they’ve put into them anyway.
Heh, I sure I am finding lots of places where they missed opportunities for nodraw and brush efficiency.
I might very well be able to put this whole issue to rest soon. Cleaning up maps certainly works wonders.
UPDATE: Good news! Through hard work and perseverance I have been able to greatly cut down on the brush work in the map. I am standing at around 65156 brush sides right now. I realize that leaves me with a budget of 380 sides for my security station, but I don’t think I should need that many, as most of the geometry work was already done. And there are probably more things I could do to increase the budget, but I think it’s fine for the moment.
Even better yet! Aside from connecting a downed pipe near the electric pool puzzle, everything else looks the EXACT SAME. You won’t be able to tell that I’ve done anything with the design. Well, maybe that’s not entirely true; I fixed a few design errors as well.
As an aside note, even nodraw seems to count towards the limit, so that doesn’t help me at all.
UPDATEUPDATE: I managed to chop an additional 455 brush sides by taking out needless Skip Textures. …Which I suppose is weird considering skip textures do nothing. They were mainly skip textures around props and lights that were obviously used for placement but weren’t cleaned up…
Huh, this could be a neat little thing to go along with On a Rail Uncut (for those that want it anyways).
A note about this room and tunnel:
The Black Mesa devs actually did have a track through there at one point before they started over simplifying everything in the chapter.
You might consider looking at the pre-release screenshots of the chapter in the wiki if you haven’t done so already.
There were also a lot of prop doors/closed exits to make the place seem bigger than it actually is.
But I don’t know if you’re trying to make it like the earlier version the BM devs made or if you’re just doing it your own way.