Hand-rails: Geometry Or Texture.

ViPr · March 2003

<div class="IPBDescription">shoot throughable. there's a way.</div> i just have to say something AGAIN coz so many people have a misconception and i'm seeing too many maps with this error even some of the official maps and it is really annoying coz i have to keep yelling at my team-mates that they cannot shoot through that hand-rail or they can shoot through this hand-rail. somebody needs to tell the mappers. this is ridiculous how there is such miscommunication. There is a way to make hand-rails that use that hand-rail texture so you don't have to make it will many polys AND at the same time make it shoot throughable while not walk throughable. i just don't know the mapping trick but it is possible. someone please post how to do it please for god's sake! i'm sick of this nonsense!

Cronos · March 2003

Make it a func_seethrough.

I've found that one can shoot through these entities, not sure if it works with railing but I'll have to look into that...

Jower · March 2003

Shoulden't this be in mapping forum ?

Farsight · March 2003

When I make handrails I make them a func_illusionary. Then a get the clip texture in the normal half-life wad and put that over exact same position and space as the handrail texture. As you can walk and shoot through a fun_illusionary but the clip brush stops you from walking through it, but you can still shoot through. Does anyone else use a different approach, i havtn tried func_seethrough though that might work.

blue2k · March 2003

no farsight your method is the correct and preffered way.

Cadaver · March 2003

Func_seethrough is for optimising the commander view by making certain objects invisible to the commander.

Farsight is right - normally you can make a railing shoot-through-able by making it a func_illusionary and putting a clip brush in the same position so that you can shoot through it, but not walk through it. The problem in NS is that skulks cannot climb on clip brushes; hence, if you use this technique to make a railing shootable, skulks will be unable to climb over it. There's no easy solution to this issue, unfortunately: so get your facts straight before you blow off at the mappers again, ViPr.

coil · March 2003

Cadaver is correct; mappers must currently choose between a railing that can be shot through, and a railing that can be climbed over. Maybe sometime in the future Flayra will add a property to the entities involved to fix this, but I don't know.

Yamazaki · March 2003

I got tired of having to choose, so I set aside a few extra polies per room for solid brushwork rails rather than transparenices. Not to mention the shoot-throughable railings tend to act oddly anyways, since you can shoot through the visible portions as well as the empty spots.

realityisdead · March 2003

Also keep in mind that many NS maps are on the brink of collapsing onto themselves during compiling because they're teetering on the edge of several engine limits. Ns_Nothing has railing that you can not shoot through, but I assure you that it is <i>not</i> due to complete ignorance or a "misconception" on my part. I'm at 100% "leaves" in the map: So full that if I add <i>one more single brush</i> to the map, it does not compile. One brush... and ns_nothing has quite a bit of handrailing. Creating shoot-throughable-but-solid handrailing requires twice the number of brushes. That'd push it quite far over the edge there!

I'm starting to become immensely frustrated having to deal with Half-Life's restrictive engine limits, but it just gets all that much worse when you have people ranting at you and declaring your incompetence because of it! :|

Anyway.

Yamazaki · March 2003

That's quite true. It lingers in my mind whenever I start constructing brushes, so much so that I begin to find ways around my old habits just to save a few polies here and there. My main problem with previous maps was an excessive use of brush entities, so I'd come really close to the limit and risk having the invisible entity bug pop up. So I use far fewer entities now whenever I can. Especially solid-mode railings, gratings, fences and the like. If I can avoid them, I do so.

BlackPanther · March 2003

The solution to this problem is easy:

Give the skulks another way to get to the other side.
It's that easy.

Personnally, im a ceiling walker when im a skulk. Dropping onto unsuspecting marines is always fun <img src='http://www.unknownworlds.com/forums/html/emoticons/tounge.gif' border='0' style='vertical-align:middle' alt='tounge.gif'>

Heist · March 2003

Hehe ken20banks... I know what you mean. I am also at 100% and anything I add has to be taken away from somewhere else.

Eldritch · March 2003

Here's how I have this small section of rail in ns_foray...

I compaired this method and the clip brush method. While my method adds 1 extra solid ent, the clip method adds to unique textures and overall texture memory, -1 solid ent.

I, of course perfer the way I've done it. While the con is 1 extra solid ent, the pros are...skulks can climb it, reduced texture memory.

Which method do you prefer?

crode · March 2003

ive found on walls that if the clipping brush is the exact same dimentions as the func_illusion (or any other brush), skulks can climb up just fine

KungFuSquirrel · March 2003

All my railings are at 32 units high, func_illusionary, and clipped. This keeps things consistent within my maps, is visually approximate, and allows full skulk mobility.

If a skulk is coming from ground-level, it's a single jump to clear it. If a skulk is coming up a wall, the momentum the skulk carries over the top will propel it over the clip brush and to the ground on the other side. I've never had this backfire on me yet, and I am 100% certain that there's a hell of a lot of people better at skulk navigation than I am.

Yamazaki · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Eldritch @ Mar 13 2003, 07:46 PM)</td></tr><tr><td id='QUOTE'> Here's how I have this small section of rail in ns_foray...

I compaired this method and the clip brush method. While my method adds 1 extra solid ent, the clip method adds to unique textures and overall texture memory, -1 solid ent.

I, of course perfer the way I've done it. While the con is 1 extra solid ent, the pros are...skulks can climb it, reduced texture memory.

Which method do you prefer?  </td></tr></table><span class='postcolor'> 
Eldritch, why do you use a func_wall for the solid portion of your railings? From the screenshot given, leaving the solid portion as a normal world brush would yield the same results and lower your entity usage.

Eldritch · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Yamazaki @ Mar 14 2003, 12:27 PM)</td></tr><tr><td id='QUOTE'> </span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Eldritch @ Mar 13 2003, 07:46 PM)</td></tr><tr><td id='QUOTE'> Here's how I have this small section of rail in ns_foray...

I compaired this method and the clip brush method. While my method adds 1 extra solid ent, the clip method adds to unique textures and overall texture memory, -1 solid ent.

I, of course perfer the way I've done it. While the con is 1 extra solid ent, the pros are...skulks can climb it, reduced texture memory.

Which method do you prefer? </td></tr></table><span class='postcolor'>
Eldritch, why do you use a func_wall for the solid portion of your railings? From the screenshot given, leaving the solid portion as a normal world brush would yield the same results and lower your entity usage.  </td></tr></table><span class='postcolor'> 
Hummm, good point. I have no idea what I was thinking... lol

Ollj · March 2003

its just ballancong between entity count (= good ping) and max_map_clipnodes (lots of level detail).
most times world brush is the better choice, it will cast correct shaddows without the opaque fix, too.

KungFuSquirrel · March 2003

I'm also curious how the clipping method adds to texture memory. I find it hard to believe that you don't have a single clip brush anywhere else in the level - and even if you somehow don't, 1 16x16 texture is hardly a noteworthy difference in texture memory.

Yamazaki · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (KungFuSquirrel @ Mar 14 2003, 06:44 PM)</td></tr><tr><td id='QUOTE'> I'm also curious how the clipping method adds to texture memory. I find it hard to believe that you don't have a single clip brush anywhere else in the level - and even if you somehow don't, 1 16x16 texture is hardly a noteworthy difference in texture memory.  </td></tr></table><span class='postcolor'> 
16x16, completely uncompressed and without mipmaps, would be 256 bytes. Considering there is some compression involved, and the mipmaps add an additional 1/4th, 1/8th and 1/16th (I believe there's three of them, could be wrong), you're looking at something that is still under 384 bytes. These aren't exact calculations, best way is to add a 16x16 texture to a WAD and check its filesize.

But, as Kung Fu Squirrel said, not very noteworthy <img src='http://www.unknownworlds.com/forums/html/emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif'>

Eldritch · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (KungFuSquirrel @ Mar 14 2003, 01:44 PM)</td></tr><tr><td id='QUOTE'> I'm also curious how the clipping method adds to texture memory. I find it hard to believe that you don't have a single clip brush anywhere else in the level - and even if you somehow don't, 1 16x16 texture is hardly a noteworthy difference in texture memory.  </td></tr></table><span class='postcolor'> 
I never said I didn't have clip brushes in ns_foray, I do in fact use it. I was just stating what the difference was between each method. When I tested the different methods I tested them in new maps. I then added the clip texture, it added to texture memory.

I still doubt I will use the clip method on rails though. Instead I will use normal solid(removing func_wall as pointed out by Yamazaki) and func_illusionary.

tommy14 · March 2003

<i>"If a func_seethrough is NOT opaque (i.e. rendered to an amount of 255 to the ground players) then it can be shot though."</i>

that was posted a while back. i haven't tested it, but i got this idea:
so if you set a func seethrough to 254 you should:
- be able to shoot thru
- have players be blocked from movement thru
- not have a commander built on top of the rail - i hope he can't build thru/into it.....
- skulks could still climb it - i think.

and 254 is still pretty full strength looks.....

btw, to reduce your planes by 1, use a wedge instead of a block for a railing. you can have the wedge point down so players cans still stand on top.....

<img src='http://www.gamers-forums.com/smilies/contrib/sz/yummyimu.gif' border='0' alt='user posted image'>

Merkaba · March 2003

I strongly suggest against using the func_seethrough render value bug as an asset in your map, as you'll probably find it will be fixed some time soon.

And also, using wedges instead of blocks will probably not help your planes. If you think about it, a wedge will create (theoretically) 3 planes, but so will a block since one of the sides will be on the same plane as the lower-wall beneath the rail. Also, if the rail is then put at an inclline, a block will generate 2 more planes (bottom and top - sides are still parallell), whereas a wedge will generate 3 (top and bottom two sides).
On the other hand, you may save on r_speeds...but only by a miniscule amount.

Gordon · March 2003

I use the clip method, and It haven't gave me any trouble.

-GF

BigD · March 2003

Depends on the railing. If it is short, I'll probably make it 31 or less units above the ground and make the posts a func illusionary. A skulk then can't go through it. If it's 32 and up, clip method. Another situation, is if it is 32 units off the ground and the railing is say, I dunno, 4-8 or up units wide. I would make the clip itself recessed, so that alien simply "thinks" he's hit the underside of the railing! Say, the posts are the "railing width" minus 2-4. Not a pretty clipping fix, (opposite what most clipping tutorials will tell ya) but you would learn to jump for a reason, and not just because you can't climb that tiny wall. When it comes down to it though, with the size of many NS maps, I could see that being the last thing a mapper worries about. (Unless railings are used extensively)

It's where you want a big fence or something where this could cause trouble. It would be fantastic to see a fix though.

Pod · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Farsight @ Mar 13 2003, 11:40 AM)</td></tr><tr><td id='QUOTE'> When I make handrails I make them a func_illusionary. Then a get the clip texture in the normal half-life wad and put that over exact same position and space as the handrail texture. As you can walk and shoot through a fun_illusionary but the clip brush stops you from walking through it, but you can still shoot through. Does anyone else use a different approach, i havtn tried func_seethrough though that might work.  </td></tr></table><span class='postcolor'> 
use NULL instead of CLIP : that way you get less wploys <img src='http://www.unknownworlds.com/forums/html/emoticons/wink.gif' border='0' style='vertical-align:middle' alt='wink.gif'>

KungFuSquirrel · March 2003

</span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (Pod @ Mar 15 2003, 01:32 PM)</td></tr><tr><td id='QUOTE'> use NULL instead of CLIP : that way you get less wploys <img src='http://www.unknownworlds.com/forums/html/emoticons/wink.gif' border='0' style='vertical-align:middle' alt='wink.gif'>  </td></tr></table><span class='postcolor'> 
What?

Clip brushes are not rendered. At all. They have no effect on wpolys. At all.

Hand-rails: Geometry Or Texture.

Comments