Error Compiling And A "ghost Wall"

Zoc · December 2003

<div class="IPBDescription">relation between them?</div> When I compile, I see this error (warning?)

hlfix v0.81b by Jedediah Smith - <a href='http://extension.ws/hlfix/' target='_blank'>http://extension.ws/hlfix/</a>
Using epsilon 0.004
Reading input file C:\Felipe\co_duel\co_duel.rmf... done
Reading wad list file C:\Arquivos de programas\Valve Hammer Editor\tools\wad.txt
... done
Snapping vertices
Tesselating non-planar faces
(Entity 0, Brush 144): Tesselating non-planar face
(Entity 0, Brush 202): Tesselating non-planar face
Decomposing non-convex solids
(Entity 0, Brush 144): Decomposing non-convex solid
(Entity 0, Brush 202): Decomposing non-convex solid
Uniting coplanar faces
(Entity 0, Brush 145): Uniting coplanar faces
(Entity 0, Brush 145): Uniting coplanar faces
<b>WARNING (Entity 0, Brush 145): Uniting faces with different texture info</b>
Writing output file C:\Felipe\co_duel\co_duel.map... done
Mapfile backup application 2.0 - Built Nov 25 2002 23:15:43
bug reports -> merlinis@bigpond.net.au

this is a pipe that I made by editing vertexes... - but in game everithing is alright with it...

and, when I was testing the map with friends, there is a wall (a simple wall) that you can walk by it and fall in "void"...

Does it have anything related between them?

Thanks

EDIT: btw, what that means? - and why this happens?

Warning: Leaf portals saw into leaf
Problem at portal between leaves 285 and 288:
(-1262.000 -3087.880 -886.000)
(-1260.154 -3088.000 -882.309)
(-1256.000 -3088.000 -874.000)
(-1256.000 -2984.000 -874.000)
(-1262.000 -2984.000 -886.000)

ProStyle · December 2003

In all my experiences the "Warning: Leaf portals saw into leaf" error is due to bogus geometry. This ranges to something that's either Fundamentally backwards of Pure Genius and unlucky - so, you be the judge and you'll probably know which area this lies in. Also the coordinate list is a good place to start but like the leak point suggestion it can be rather vague.

The void or invisible wall sounds like you're having issues with clip node generation? Try changing the settings in CSG or just re-vamping those areas in the editor. This could also be related to the leaf saw?

Confused · December 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> </td></tr><tr><td id='QUOTE'>Leaf portal saw into leaf
There are fine tutorials here : <a href='http://members.lycos.co.uk/EditHalfLife/tutorials/leafs.htm' target='_blank'>http://members.lycos.co.uk/EditHalfLife/tu...rials/leafs.htm</a> and at DeWan's Old Tutorial on BSP and VIS.

Basicly this comes from an area whose floor is not square enough, square floor areas never give this problem. Note: often if there is only one leaf saw leaf in a map level, it can be ignored. So before getting busy, see how your map plays!
Solutions to try when you have a lot of leaf saws, or problems from leaf saws:

1. VIS -full compiles were designed to help with this problem, so the FIRST thing to try is recompile with VIS -full and see if it goes away.
2. func wall any brushes within the coordinates given for the problem, this will simplify VIS in the area because func wall brushes are invisible for VIS.
3. Move walls in the area to make the floors more square.
4. Put a HINT brush across the trouble area. HINT brushes cut VIS leafnodes, so they will reshape the area and may fix the problem.
5. Finally just start ripping brushes out of the area.

Fixing Leaf saw error has also been known to fix other errors in later RAD, like
FindTransferOffsetPatchnum returned -1 looking for patch 1570 in patch 199's transfer lists and other problems.</td></tr></table><span class='postcolor'>
- <a href='http://www.slackiller.com/tommy14/errors.htm#leafsaw' target='_blank'>tommy 14</a>

Jedediah · December 2003

Find the non-planar faces on brushes 144 and 202 and fix them. Most likely the non-planarity is only by a small amount and hlfix is cutting them into tiny sliver brushes. hlvis definately does not like tiny sliver brushes.

Yes, I'm aware of the irony that this is precisely the sort of thing that hlfix is supposed to fix but, well, it's not perfect (yet).

Zoc · December 2003

I've fixed the brush 145, and now the leaves are gone (and the ghost wall too)

thanks <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

btw, I think that I'm not getting problem with these:

Tesselating non-planar faces
(Entity 0, Brush 144): Tesselating non-planar face
(Entity 0, Brush 202): Tesselating non-planar face
Decomposing non-convex solids
(Entity 0, Brush 144): Decomposing non-convex solid
(Entity 0, Brush 202): Decomposing non-convex solid

these are 2 octagonal pipes...
What problem could this make ? (I can't see any problem until now)

Kage · December 2003

You could try turning up the epsilon factor in hlfix if it's doing something that it shouldn't be doing. The parameter is -e [#] (1.0 is the default).

Zoc · December 2003

I'll try it... but (noob question) - can you explain what that means and why turning this up could fix ? <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

Kage · December 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (HlFix Website)</td></tr><tr><td id='QUOTE'>Epsilon Factor

Certain numeric comparisons in hlfix require a "fuzziness" value in order to deal with inaccurate data that can be generated by VHE. The value of the -e parameter is a multiplier for this fuzziness value. If you find that hlfix is modifying geometry it shouldn't or producing unexplainable "Incomplete edge cycle" errors, try raising this value. If you find that hlfix is not modifying geometry it should or that your compiled maps have visible errors, try lowering this value. The base value (1.0) is one that testing to date has shown produces correct results all the time. This value may change as beta testing proceeds and ultimately, the user will not have to worry about this parameter. For the time being, if you find that you need to alter the epsilon factor to get hlfix to work properly, please report this in the forums. </td></tr></table><span class='postcolor'>

Your compile log shows that your epsilon factor is set to 0.004.

Zoc · December 2003

hmm... thanks...
I'll increase tha value (didn't noticed that was soo low... - I'm using batch compiler's default for this.)

btw, does that mean that something wrong is happening? (or could happen?)

Tesselating non-planar faces
(Entity 0, Brush 144): Tesselating non-planar face
(Entity 0, Brush 202): Tesselating non-planar face
Decomposing non-convex solids
(Entity 0, Brush 144): Decomposing non-convex solid
(Entity 0, Brush 202): Decomposing non-convex solid

EDIT: I tried to use "-e 1.0" - but in the log still shows 0.004 - some kind of "bug" when showing the log?

watch_me_die · December 2003

They shouldn't cause any problems, HLFix is fixing them <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

Zoc · December 2003

Great! <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

Thank you <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

Jedediah · December 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Zoc @ Dec 7 2003, 12:09 PM)</td></tr><tr><td id='QUOTE'> hmm... thanks...
I'll increase tha value (didn't noticed that was soo low... - I'm using batch compiler's default for this.)

btw, does that mean that something wrong is happening? (or could happen?)

Tesselating non-planar faces
(Entity 0, Brush 144): Tesselating non-planar face
(Entity 0, Brush 202): Tesselating non-planar face
Decomposing non-convex solids
(Entity 0, Brush 144): Decomposing non-convex solid
(Entity 0, Brush 202): Decomposing non-convex solid

EDIT: I tried to use "-e 1.0" - but in the log still shows 0.004 - some kind of "bug" when showing the log? </td></tr></table><span class='postcolor'>
Those messages mean hlfix is repairing those errors. Ideally, you should figure out what is wrong with them that needs repairing, but if everything works fine in-game, you can probably just forget about it.

Also, the epsilon <i>factor</i> you supply on the command line is a multiplier for the base value which is 0.004. So -e 2 would give you an epsilon of 0.008, -e 0.5 would give you 0.002 etc.

Cagey · December 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Zoc @ Dec 5 2003, 11:15 PM)</td></tr><tr><td id='QUOTE'>and, when I was testing the map with friends, there is a wall (a simple wall) that you can walk by it and fall in "void"...

Does it have anything related between them?</td></tr></table><span class='postcolor'>
If you're using the p12 tools, I'm currently collecting maps with bugged clipping in an attempt to make this no longer happen. Would you be willing to email me your map source (and any required wad files) for testing the p12 tools?

Zoc · December 2003

hmmm... sure... I'll contact you soon <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

(If I don't send any email to you, its because I got some little problems in the university... - fell free to email me... probably I've forgot to do that... <img src='http://www.unknownworlds.com/forums/html/emoticons/wink.gif' border='0' style='vertical-align:middle' alt='wink.gif'>)

Error Compiling And A "ghost Wall"

Comments