ok I started reading this thread normally, but then as I got to some posts that suggested the coders are simply too lazy to make this work I got **** off...
perfectly normal post to begin with of course, but then the idiots crawled out from under their rocks...
if you've followed NS for awhile you will know that the team's tried to solve this many ways but it's simply too difficult... they're already bending every rule of the HL engine, and having what is basically locallized gravity is perhaps a bit too difficult or impossible to code...
asking a "rhetorical" question such as (paraphrased): "should the mappers make lamer maps with no details, should I bother learning how to navigate detailed maps, or should the coders get off their lazy **** and get this working" is down right immature and irresponsible... not to mention ungreatful
*EDIT* I decided to finish reading the thread and now have a few more things to say...
the AvP wall-walking code was a core component of the game's engine... the engine was designed so that A: predators could cloak B: you could implement vision modes C: aliens could wall walk
that's pretty much all that engine was based around, and that's all it did, and it did that well, to a point...
if you'll try to think way back, remember how an alien could never ever crawl down a flight of stairs, but had to instead stand up to do it, as crawling would make him jerk back and forth as he tried to navigate the tiny 2 inch gap of the first step? didn't think so... trust me, it's there, boot up AvP and try it if you want... everything seems better when you remember it fondly
I'm becoming annoyed with Frahg too, very quickly... he was the one whom I paraphrased up there... just because something has been done, doesn't mean it can be easily reproduced, especially when the setting (in this case the HL engine) is so very different... implementing wall walking is easy when the engine is made for it, and I'm sure Flayra would have wall walking all figured out if he was making NS for the AvP engine, but as it stands he's working with a very outdated engine that was never meant to do anything that even resembles what NS is right now, so yeah, if it hasn't been done, it probably can't be done...
PS: if you're wondering why it's not as simple as implementing the code for walking over pipes on the ground, just think: gravity... pipes on the ground bump you away from the surface you're walking on (in this case, the ground), and then gravity pulls you back to the ground... if the same code was implemented on the ceiling you'd bump away from the ceiling, and then gravity would bring you away from the ceiling... it's all a little more logical when you think instead of just crapping out your mouth
You might get offended a little too easily. I never meant to disrespect anyone with my post. I believe Flayra has done an incredible job on NS. There are still some problems with it, but all I really was saying is that if you were going to put someone in charge of fixing the problem, it should be the coders. It's a code problem, not a map problem or a player problem. I never said that I thought the coders had to fix this; all I did was identify the proper party to correct it if they did.
That said, I'll address your P.S.: We should never get into a technical debate about how to solve the problem. We can talk all day about how it actually works, that I don't mind, but when you start looking at "what would happen if you tried to solve it" then all we can do is make stuff up. There are lots of "possible" solutions, and we'll never know if they work until they're tried.
moultanoCreator of ns_shiva.Join Date: 2002-12-14Member: 10806Members, NS1 Playtester, Contributor, Constellation, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Gold, NS2 Community Developer, Pistachionauts
I'd really like to see generally how the current wall walking code works. Maybe there ought to be a separate forum for people with some coding background to brainstorm ways to fix things.
<!--QuoteBegin--moultano+May 16 2003, 01:36 PM--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (moultano @ May 16 2003, 01:36 PM)</td></tr><tr><td id='QUOTE'><!--QuoteEBegin--> I'd really like to see generally how the current wall walking code works. Maybe there ought to be a separate forum for people with some coding background to brainstorm ways to fix things. <!--QuoteEnd--> </td></tr></table><span class='postcolor'> <!--QuoteEEnd--> I second that. I've longed speculated on how it might work, but I've often wondered how much CPU time my solution would take up on the client...
the way wall working works now is limited by clip nodes. Xp cagey has writen a lovely article <a href='http://xp-cagey.com/articles/clip_hull/' target='_blank'>here</a>
which explains how the new clip nodes are going to be done.
the new system makes wall walking amazingly easy on 95% of surfaces
it willb eincorperated into new maps and you life will get better.
the mapping community has this one down kiddies:)
it also is adressed in teh sticky in mapping about maxplanes and better compile tools
<!--QuoteBegin--confused!+May 16 2003, 02:18 PM--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>QUOTE</b> (confused! @ May 16 2003, 02:18 PM)</td></tr><tr><td id='QUOTE'><!--QuoteEBegin--> the way wall working works now is limited by clip nodes. <!--QuoteEnd--></td></tr></table><span class='postcolor'><!--QuoteEEnd--> What does that MEAN? (I'm assuming that a clip node is one part of the clip hull, but XP Cagey's information doesn't address that. )
Are you saying that wall clinging works by detecting if the bounding box is sticking through a clip node? If so, doesn't that conflict with XP Cagey's statement that detection of this type would be very intensive? Or is it just that intensive is a relative term and computers of today can handle it?
I suppose this belongs in a different forum, but to be honest I think it's the coding forum that doesn't exist--so this is as good as any.
the essential resone one falls off the ceiling is that the surfaces of the skulk which are toutching the wall come off. in most cases this happes because the edges of the clip hull are calculated wrong. for example
diffuculty with climing the funrance, from left to right, in refinery hive of bast is casued by the clipping hull.which is too inacurate for a skulk to climb the curves nicely.
using the new tools most of thes problems vanish. hopefully the new model pov rotation in 1.1 will fix the rest.
im waiting to see
the way i under stand wall walking is much the same way walking on teh floor works. if your clip hull toutches teh wall 's clip hull you can walk on it. the problems come in with teh famed sticky corners. then you come out of contact and fall off the wall, assuming you try and go over them. the new compile tools try to get rdi of teh corners and thus allows you to move more naturally along the walls.
lemme know if that aint clear of i spelled some thing wrong
moultanoCreator of ns_shiva.Join Date: 2002-12-14Member: 10806Members, NS1 Playtester, Contributor, Constellation, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Gold, NS2 Community Developer, Pistachionauts
edited May 2003
Aiight, having read this, I have an idea, and maybe someone will be able to figure out what Im talking about. How about whenever a skulk leaves a clip face, there is a very short push in the direction opposite the normal vector of the face he was just on.
A small force pushes up keeping him in contact with the surface. <!--c1--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>CODE</b> </td></tr><tr><td id='CODE'><!--ec1--> F_walk ------> /\ F_push | __ .-| F_walk+F_push = .-' <!--c2--></td></tr></table><span class='postcolor'><!--ec2--> but if he wants to get off <!--c1--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>CODE</b> </td></tr><tr><td id='CODE'><!--ec1--> F_walk | | | \/
/\ F_push |
F_walk+F_push = | | \/ <!--c2--></td></tr></table><span class='postcolor'><!--ec2--> so he can still get off normally, and its just a little more difficult, but if he holds down crouch, he'll drop off just like normal.
ph34r my ASCII ART! (hopefully it doesn't get **** up)
regarding my PS up top... that was from a post made by I believe Flay, way back, and was simply a parahprased "reason" that was given to us when we asked this some 6 months ago
I'm really happy to see that they've managed to get the problem fixed on the compile side though, or that they seem to be getting a lot of progress with it
*EDIT* hadn't seen your post at the time of this one
that does seem like it could work, but currently crouch doesn't actually apply any force on the skulk, but simply releases it from the surface; that's just a technicality though, and could be easily corrected (I believe jump actually pushes you away from the wall/ceiling if you're walking on it, though it could simply be a side effect fo something else)
but yeah it seems like an elegant solution, and I love the ASCII art!
Comments
perfectly normal post to begin with of course, but then the idiots crawled out from under their rocks...
if you've followed NS for awhile you will know that the team's tried to solve this many ways but it's simply too difficult... they're already bending every rule of the HL engine, and having what is basically locallized gravity is perhaps a bit too difficult or impossible to code...
asking a "rhetorical" question such as (paraphrased): "should the mappers make lamer maps with no details, should I bother learning how to navigate detailed maps, or should the coders get off their lazy **** and get this working" is down right immature and irresponsible... not to mention ungreatful
*EDIT*
I decided to finish reading the thread and now have a few more things to say...
the AvP wall-walking code was a core component of the game's engine... the engine was designed so that
A: predators could cloak
B: you could implement vision modes
C: aliens could wall walk
that's pretty much all that engine was based around, and that's all it did, and it did that well, to a point...
if you'll try to think way back, remember how an alien could never ever crawl down a flight of stairs, but had to instead stand up to do it, as crawling would make him jerk back and forth as he tried to navigate the tiny 2 inch gap of the first step? didn't think so... trust me, it's there, boot up AvP and try it if you want... everything seems better when you remember it fondly
I'm becoming annoyed with Frahg too, very quickly... he was the one whom I paraphrased up there... just because something has been done, doesn't mean it can be easily reproduced, especially when the setting (in this case the HL engine) is so very different... implementing wall walking is easy when the engine is made for it, and I'm sure Flayra would have wall walking all figured out if he was making NS for the AvP engine, but as it stands he's working with a very outdated engine that was never meant to do anything that even resembles what NS is right now, so yeah, if it hasn't been done, it probably can't be done...
PS: if you're wondering why it's not as simple as implementing the code for walking over pipes on the ground, just think: gravity... pipes on the ground bump you away from the surface you're walking on (in this case, the ground), and then gravity pulls you back to the ground... if the same code was implemented on the ceiling you'd bump away from the ceiling, and then gravity would bring you away from the ceiling... it's all a little more logical when you think instead of just crapping out your mouth
That said, I'll address your P.S.: We should never get into a technical debate about how to solve the problem. We can talk all day about how it actually works, that I don't mind, but when you start looking at "what would happen if you tried to solve it" then all we can do is make stuff up. There are lots of "possible" solutions, and we'll never know if they work until they're tried.
--Frahg
I second that. I've longed speculated on how it might work, but I've often wondered how much CPU time my solution would take up on the client...
--Frahg
Xp cagey has writen a lovely article <a href='http://xp-cagey.com/articles/clip_hull/' target='_blank'>here</a>
which explains how the new clip nodes are going to be done.
the new system makes wall walking amazingly easy on 95% of surfaces
it willb eincorperated into new maps and you life will get better.
the mapping community has this one down kiddies:)
it also is adressed in teh sticky in mapping about maxplanes and better compile tools
<!--QuoteEnd--></td></tr></table><span class='postcolor'><!--QuoteEEnd-->
What does that MEAN? (I'm assuming that a clip node is one part of the clip hull, but XP Cagey's information doesn't address that. )
Are you saying that wall clinging works by detecting if the bounding box is sticking through a clip node? If so, doesn't that conflict with XP Cagey's statement that detection of this type would be very intensive? Or is it just that intensive is a relative term and computers of today can handle it?
I suppose this belongs in a different forum, but to be honest I think it's the coding forum that doesn't exist--so this is as good as any.
--Frahg
in most cases this happes because the edges of the clip hull are calculated wrong.
for example
diffuculty with climing the funrance, from left to right, in refinery hive of bast is casued by the clipping hull.which is too inacurate for a skulk to climb the curves nicely.
using the new tools most of thes problems vanish.
hopefully the new model pov rotation in 1.1 will fix the rest.
im waiting to see
the way i under stand wall walking is much the same way walking on teh floor works.
if your clip hull toutches teh wall 's clip hull you can walk on it.
the problems come in with teh famed sticky corners. then you come out of contact and fall off the wall, assuming you try and go over them. the new compile tools try to get rdi of teh corners and thus allows you to move more naturally along the walls.
lemme know if that aint clear of i spelled some thing wrong
so for example
________,-'
______.-'
<!--emo&::skulk::--><img src='http://www.unknownworlds.com/forums/html/emoticons/skulk.gif' border='0' style='vertical-align:middle' alt='skulk.gif'><!--endemo-->--->
A small force pushes up keeping him in contact with the surface.
<!--c1--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>CODE</b> </td></tr><tr><td id='CODE'><!--ec1-->
F_walk ------>
/\
F_push | __
.-|
F_walk+F_push = .-'
<!--c2--></td></tr></table><span class='postcolor'><!--ec2-->
but if he wants to get off
<!--c1--></span><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><b>CODE</b> </td></tr><tr><td id='CODE'><!--ec1-->
F_walk |
|
|
\/
/\
F_push |
F_walk+F_push = |
|
\/
<!--c2--></td></tr></table><span class='postcolor'><!--ec2-->
so he can still get off normally, and its just a little more difficult, but if he holds down crouch, he'll drop off just like normal.
ph34r my ASCII ART! (hopefully it doesn't get **** up)
I'm really happy to see that they've managed to get the problem fixed on the compile side though, or that they seem to be getting a lot of progress with it
*EDIT* hadn't seen your post at the time of this one
that does seem like it could work, but currently crouch doesn't actually apply any force on the skulk, but simply releases it from the surface; that's just a technicality though, and could be easily corrected (I believe jump actually pushes you away from the wall/ceiling if you're walking on it, though it could simply be a side effect fo something else)
but yeah it seems like an elegant solution, and I love the ASCII art!