hammer is stupid

i decided it was about time to start the attempt at reducing r's on the map i've been working on. some sky and some hint brushes helped a bit, but when i tried turning some rocks into a func_wall, hammer decided that they were somewhere else. when i ran the compile, it stopped, telling me that there was a leak starting at a func_wall @ 0,100,4. it's the '4' bit that's odd. none of the map is above -300.

covering the area with a sky brush seems to have solved it for now (compiling atm), but i'm not too happy with that. seems a bit 'chewing gum and sticky tape' ish.

i've seen stuff like this happen before with func_water, but never with a func_wall. just wondered if anyone else has.

Any of these rocks that you 'func_walled' on the edge of your map? An entity-based brush will not seal the level like a normal brush does, making a leak. Try changing the rocks back to world brushes and see if that sorts your problem... :smile:

also make sure that if and when you DO remove some of the pesky func rocks, that you do a check for problems, and delete the leftover enties....they seem to stay some times after you chop em up etc....just a thought....i found out the hard way... :heee:

and at the risk of a flamewar, i dont think its hammer thats causing yer problem there man....

:Doc Brass... :dodgy:

no adam, i knew about that one. ,)



i suppose its easier to see what i meant from the pic.

like i said, extending the sky over where hammer thinks the rocks are has sorted it for now.

hehe thats funny..

hammer IS stupid..

reign

Doc Brasso has got a good point: sometimes this error can be caused by empty solid entities. Check for problems if you haven't already; any occurences of "solid entity is empty" will need to be deleted manually (Hammer can't fix this itself).

Brush based entities (excluding certain rotating ones) will not cause a leak if they are outside the world. They can only "cause" a leak if you intended them to seal your level, which they can't do.

From the looks of your drawing it says the rocks are at the origin, which certainly suggests to me that the problem is the one Brasso and Gollum pointed to.

Keen are the eyes of the elves! I hadn't noticed that, but it makes sense.

ahhh.....elven goodness...../me reaches for box of Keeblers crackers....mmmmm

fishy, please post the results of your quest so it can be added to the data base ergo, the snarkpit "fountain of knowledge"....

Doc Brass... :dodgy:

still no definative answer.

i check for errors regularly. there is no instance of any empty entities (something i've not seen for a very long time, btw), or any other errors. anyway, that would put the entity at 0,0,0 (the grid origin) and not 0,100,4

here's another pick. i used edit>select all, to highlight everything, instead of the select tool. this would highlight even empty solids, but as you can see, there is nothing highlighted at the co-ords stated. the entire level is well below that area.



the evil carve tool has never been out of it's box, so i can't point the finger at it. all of my rockwork is done by adding edges to simple blocks, and moving the vertex points, so it's not like i've clipped a brush and ended up with floating points either. this is one of the reasons i use alt+p regularly. to make sure i've not made an invalid shapes.

this is definately more akin to func_water drowning you in parts of a level where it doesn't exist. some of you (if not most) will have seen it at some time. seems to happen when func_water is either a complex shape, or is next to one. when it happens though, there wont be any leak created if it extends into the void, as func_water can exist there anyway.

this makes me wonder how many times this has been the cause for the threads that go something like, 'whats wit dis f00kin leak s**t - my maps is teh perphakt'.

the answer would probably be none, but i wonder about some strange stuff at times. :smile:

.........................................................

wow, another strange thing just happened.

in preparation to offer up my bsp to thomas, i retextured the whole map (other than sky and hint) with a generic hl texture, so that he wouldn't have to dl the wads for it too. i lowered the top of the sky back down to below the co-ords to expose my phantom func_wall to the void again, and ran the compile proccess to confirm the leak was still happening. it was. but the other strange thing was in the log. i had exceded max clipnodes. seems very odd that retexturing something would increase the clipnodes. other than maybe free-standing water, which i dont use much of at all.

This is a shot in the dark, but worth mentioning... Have you accidentally hidden any objects at any point that might be in an invisible VisGroup? Maybe you had some brushwork up in that location, hid it, and then moved the rest of the map?

I'm sure this probably isn't the case, but if you haven't checked yet you might just want to be sure. I know I've made plenty of silly mistakes like that. shrug

Don't hate me for mentioning this, but I can't count how often I've been stumped on something only to find it was just a weird little mistake. :razz: I'll go hide in the corner now.

So hold up a second, from the diagram it looks like you put the rocks outside the level but they are appearing inside the level in-game, but from your explanation of the problem (in words), it seems like just the opposite problem is happening.

1: where did you put the rocks?

2: where does hammer show them to be (if not the same as 1, please explain)?

3: where do the rocks show up in the game?

KungFuSquirrel said:
Don't hate me for mentioning this, but I can't count how often I've been stumped on something only to find it was just a weird little mistake. :razz: I'll go hide in the corner now.

oh my, do i come over that serious? i'd better use more smilies and stuff.<!--StartFragment --> hate is bad. very self destructive. so no, i wont hate you. i might throw things at you or kick your dog or somesuch, only just as stress relief you see, never as hate.<!--StartFragment -->

i'm gonna bump the thread in the maps forum with a link for the compiled map, but if anyone wants to have a look at the rmf to verify/disprove/whatever this strangeness, i'd be happy to zip it up for them. (reputable enquiries only please<!--StartFragment --> )<!--StartFragment -->

dont know if im considered "reputable", but i'd be glad to take a look at the rmf and see what i can decipher....send it to my aol addy, itll handle the size limitations just fine....pm me if you decide to so i'll watch for it..... :wink:

bugmanrox@aol.com

Doc Brass... :dodgy:

Hehe, well, it was worth a shot. :razz: you never know.

ok fishman....a few things ive found right off the bat....would you care to discuss them here, since it is an editing forum, or would you like to get on icq or aim or something first?....yer call.....

i havent done a full compile yet, because of the clipnodes factor....i have since removed them/turned it off and am going to compile it overnight....ill let you know what happens....other than that, its pretty decent architecture, although the one texture covering everthing makes it kinda hard to appreciate....

lemme know....

Doc Brass... :dodgy:

mazemaster said:
1: where did you put the rocks?

2: where does hammer show them to be (if not the same as 1, please explain)?

3: where do the rocks show up in the game?

1. inside the level

2. shows them where i put them

3 show up in the correct place in game

its only when i try to compile that hammer says there is a leak, originating outside the level.

Doc, fire away. you should only need to run a bsp compile for the error to show itself.

Then what does it have to do with the rocks? I think its time for a full compilelog.

alrighty then...installment no. 1....

this is one i havent run across.....hint brushes all over the place....not that thats a bad thing, i just dont think that they are used in a very efficient manner.....and, im not real sure yet (still researching), but a few were hanging outside the "box" into the space outside the map.....im not sure what effect this has yet, but im still working the problem....when you use hint brushes, placement is fairly crucial, but im sure you knew that....

if anyone has some definative knowledge on this, please speak up, because this is one helluva search.... Gollum?...where are ya???....

and, many many brushes intersecting/overlapping...in a basic mathematical "world", which is what this is, everytime you lap or intersect them, it is compounding your compile times, and increasing your faces by a factor of at least 4....which in my mind kind of counteracts the effects of func_walls....now, this is just theory, but im thinking at this point that this is why you are reaching your clipnodes max....im asking for your permission to do some actual editing on your rmf....just to test my theory....btw, i compiled with -noclipnodes, and it basicly froze at about 62%.....over a nine hour period....

i will tell you, i like the layout, but i think you need to be a bit more careful in your brush placement.....be a little "cleaner" in your placement....may sound harsh, but that is not my intent.....

......to be continued....

Doc Brass... :dodgy:

HINT brushes are a pretty arcane subject. I advise removing all of them unless you have a very clear idea of what you want each brush to achieve.

Also, HINT brushes should only ever be placed as blocks (please, please don't make 32-sided cylindrical HINT brushes :wink: ) Only one face should be textured "HINT"; the rest should be textured "SKIP". The face that is textured with "HINT" will define the forced BSP split. Make sure that it exactly covers the opening that it is placed over.

Note that if you texture the whole brush with "HINT", there is a good chance that you will make 6 BSP splits where you intended only to make 1.

Finally, don't try to use HINT brushes as a substitute for actual VISblocking. You cannot "engineer" lower visibility with HINTs - you can only influence the way the compile tools cut up the level.

damn i am glad you showed up mike....the hint brushes are on all sides of the brushes used as far as i can tell....

//reaches for grecian formula to remove additional gray hair that is sprouting as we speak....

Doc Brass.... :dodgy:

Erk! That will cause nasty stuff to happen - at least, I expect it to confuse the hell out of hlbsp.exe

i was afraid of that....well hell, here we go.....btw, whats yer take on my lapping brushes,/face splitting theory in this case?

as soon as i get permission to redo alot of this, ill get on it.....tnx for yer help mikee, yer a good man... :wink:

...praise be to the god of boulders... :heee: ..../offers up first born daughter as sacrifice.... :heee:

Doc Brass... :dodgy:

You're too kind :redface: ....../accepts sacrifice :heee:

As for overlapping brushes - well, they should be avoided unless they are entities. Overlapping world brushes will cause compile times to increase, perhaps vastly, and may lead to r_speeds increases.

Even one set of complex overlapping brushes can send your compile times through the roof. God only knows what a level full of them would do.

I remember my first attempt at mapping - I overlapped all the walls slightly because I thought it was necessary to seal the level Humbling to think back to that.

again, as i suspected.....tnx for the reinforcement.....

.....as to the sacrifice, you havent seen her yet.....roflmao

Doc Brass... :dodgy:

Well if nothing else, she'll make good catfood.

Gollum said:
Well if nothing else, she'll make good catfood.

might be more humane as "boulder fodder"

man, she would literally kick my ass if she knew her dad was talking this way....

....and shes actually damn good looking...ive been putting the hex on her boyfriends since she was 15....

"a little fear is a good thing....." :heee:

you guys kill me.....i sure do like this site...

tnx again man.... :wink:

Doc Brass... :dodgy:

arghh

i've lost a post that i'd been at for nearly an hour

i'll re-do it once i calm down a bit.

hmphh

? posted by Dr Brasso
1. hint brushes all over the place....not that thats a bad thing,

2. many many brushes intersecting/overlapping...

3.im asking for your permission to do some actual editing on your rmf....just to test my theory....

4. btw, i compiled with -noclipnodes, and it basicly froze at about 62%.....over a nine hour period....

5....may sound harsh,

6......to be continued....

1. you make them sound random :smile: some of them actually helped out a lot.

2. honestly? the only world brushes that i can find to be overlapping, in the rmf i sent you, are brushes that had the clip texture on them before i re-textured the whole map. there's a compile paramater that gives you an indication of brush overlap in a map. i'll try it out and see what it says.

3. fire away. if you do anything that helps with the r's etc, i'd be glad to here of it. :smile:

4. retexturing the clip brushes i think is to blame here too. instead of reducing clipnodes on complex brushes, they are being horribly split by them. the compile i ran before the texturing only took around an hour and twenty.

5. not at all. i'm just a little surprised that you say there are so many overlapping brushes.

6. good, good :biggrin:

? posted by Gollum
1. HINT brushes.I advise removing all of them unless you have a very clear idea of what you want each brush to achieve.

2. Only one face should be textured "HINT"; the rest should be textured "SKIP".

3.Note that if you texture the whole brush with "HINT", there is a good chance that you will make 6 BSP splits where you intended only to make 1.

4. You cannot "engineer" lower visibility with HINTs - you can only influence the way the compile tools cut up the level.

5. Finally, don't try to use HINT brushes as a substitute for actual VISblocking.

1. i may remove some of them, but as above, some of them help out a lot.

2. nice one. thank you. i only wanted one face to have the hint texture, as i could see the double splits that were being caused when i was running with wireframe on. i looked for a 'null' texture that i've seen discussed, but i dont seem to have one. i didn't know i had a skip texture.
is this the same as the 'null' that i've heard of, and can it be used to kill faces that player wont see?

3. looked like more than 6 with wireframe on :/

4. hint brushes, used correctly(which would be the hard bit :razz: ), will reduce r's. whether this is done by how they influence the cutting up of the map, i dont know. but surely its the same as engineering lower vis with hints?

me and my damn semantics :razz:

5. i know they dont work as normal visblockers. i like to think of them as being a type of visblocker though, but only when they are not in my direct line of sight. it's not exactly accurate, but close.

i hope i dont sound pissed at your comments or the way i've answered them. i'm not. very much appreciated that you took the time. cheers.
i'd already written out a big post, and it got lost in a moment of madness. (i hit the back button..........duh...)

Dr B, you didn't say if you still got the leak error. i'm assuming you didn't if your compile is taking more than 9 hours. hehe, it would be one helluva map if it took 9 hrs before it found a leak.

thanks again guys.

fishy said:
2. honestly? the only world brushes that i can find to be overlapping, in the rmf i sent you, are brushes that had the clip texture on them before i re-textured the whole map. there's a compile paramater that gives you an indication of brush overlap in a map. i'll try it out and see what it says.

thanks again guys.

dunno about the rest of ya, but i always thought of one or more overlaps as alot... doesn't everyone?

1. hint brushes all over the place....not that thats a bad thing,

.....sorry...didnt mean to make them sound random....was not my intent....some were placed well...others i found to be a hinderance more than anything...

2. many many brushes intersecting/overlapping...

i took about ten screenies if youd like to see them....most that i shot were along the outside areas of the map, the perimeter i mean....and i quit after ten...lol :heee: i can zip em and send them if you like....

anyway, ill see if i can get er fixed up a bit, and send her back to ya....sound like a plan?.... :wink:

Doc Brass... :dodgy:

PM me a link to the map and ill see what i can do.

i think i have it under control maze, but if you wish, i will, with fishy's permission.... :wink:

Doc Brass... :dodgy:

Dr Brasso said:
i took about ten screenies ....i quit after ten...lol :heee: i can zip em and send them if you like....

anyway, ill see if i can get er fixed up a bit, and send her back to ya....sound like a plan?.... :wink:

yer, send me the pics please. fisheye@invalid-brush.co.uk

i've already done more to the map, rearranged some architecture and built some new bits, so it doesn't sound like the best of plans :sad:

knowing where you find any problems would be good though, especially if you find them on the properly textured rmf.

cheers

done...i sent them from my wifes email addy, kernsbe@cox.net mine seems to be b0rked at the moment.....lemme know yer thoughts fishman.... :biggrin:

Doc Brass... :dodgy:

cheers Al. some of those i knew about. especially the ones where the edges of the roofs cut into the sky. i assumed this wouldnt be a problem, as the bit inside the sky doesn't get rendered, so there's no extra faces. i'll tidy it up though.
where the hint brushes overlap world brushes is all sorted. (doesn't mean it'll stay sorted )most of them got deleted, and the ones that helped have only one hint face now. :smile:
himt brushes are still something i'm experimenting with .

some of the pics i cant tell where on the map they were taken, but 1 or 2 are(or should be) func_illusionary brushes(sewer water). you did point out a couple that i had missed. nice 1, thanks :biggrin: . i started to look for them after what you said, and found some more. mainly in bits that i had intended to work on anyway.
hehe, you did get one that i had forgotten about. there's a bit where the angle of a tunnel got changed, and it had 1 brush longer than the rest. after i'd made the doorway between the tunnel and a cellar that it led to, i noticed the overlap. i couldn't be bothered fixing it, as it wouldn't be seen in game. /me bad :rolleyes:

i think if i stick to a more uniform grid layout in future, i wont have as much snagging to do at the end up. part of the learning curve i suppose.

btw, Dr B, did you get a leak starting at 0,100,4 at all?

glad to have attempted to help....any time man... :biggrin:

and actually fishman, i didnt get a leak now that you mention it.....hmmm.....

Doc Brass..... :dodgy: