The Flibo Bug

Started by SchlumPF
23/12/2008 18:09
Page 3 of 5
SchlumPF25/12/2008 20:34(edited 25/12/2008 20:40)

thopa you need to measure the jump from the player back on jumpoff to the players front on landing.
you set the first point of your measuring correctly but the second on shouldnt be on the upper left corner of the player, it should be on the upper right.

edit: thats not possible. basically the player owns 2 boxes: one is his model which turns with the players view AND the collision model. the collision model says when a player hits something etc and therefore its much more important than the players modelbox.

edit2: this could also be a small bug in ljstats since ljstats is getting the players center coordinates on jumping and calculating it on landing and adds then 32 as a constant. i think that 32 has to be changed too :/

i could be uber wrong about this all but hey there is a lot of crap in this forums so why shouldnt i post my stuff too :P

thopa25/12/2008 21:11(edited 25/12/2008 21:21)

Originally posted by SchlumPF
thopa you need to measure the jump from the player back on jumpoff to the players front on landing. - That's what I did.

you set the first point of your measuring correctly but the second on shouldnt be on the upper left corner of the player, it should be on the upper right.


This doesn't make sense. If you calculate with Pythagoras using the playerbox (in this case) you should always calculate using the same side of the origin (start/stop). In my case I used the player box left side as my origin/start, that's why I also the left side as stop. You could use the right side as start/stop, but you should never mix them, especially in a case like this.

The third (wrong) picture calculates the diagonal of the jump, which is ~245.41 units. But when you are including/using the size of the player box you can't use the same Pythagoras-calculation since it will only show us 245.41 units again. What we want is to know how long you need to jump if you fully use the player boxes size of 32 units when diagonal-jumping. And that answer is only given using the two left pictures.

Picture

I totally don't think this is wrong, but If I am, ill stay out of this thread. :)

SchlumPF25/12/2008 22:53(edited 25/12/2008 23:02)

the engine uses the very center of the playermodel as its origin (-link-). in ljstats the jumpoff origin is retrived, the landing origin is calculated. to get the full distance 32 units are add (-link-).
simply turn the image and you see why your pic was incorrect ^^

zyllEE25/12/2008 23:30

did it with 238.9........ ;o

thopa25/12/2008 23:38(edited 25/12/2008 23:40)

Originally posted by SchlumPF
the engine uses the very center of the playermodel as its origin (-link-). in ljstats the jumpoff origin is retrived, the landing origin is calculated. to get the full distance 32 units are add (-link-).
simply turn the image and you see why your pic was incorrect ^^


Yeah, I understand that the engine uses the very center of the playermodel, and the plugin for lj stats uses this center as it's origin. But my calculations doesn't actually have anything to do with the lj stats plugin, that's why I don't really calculate using the center of the player model. Even though there is a center, you can stand on the edges on the playermodel, and when calculating the shortest amount of units you need to jump to make this block, you don't need to calculate from it's center - Unless you aim to perfect a plugin or something.

My picture is based on that you don't "stand on the point where the center of the origin is". You can stand on the very edges of the playermodel, and that's how I calculated, and that's how you don't need to make a 246 LJ to make the block.

Maybe it's beyond my understanding, but I still believe in my calculation and that it's enough to make a jump that is +236.91 (but alot less under 245) units to land the block. Reality has also shown it. :)

So it's not really a bug, it's just the way you use the playermodel's size of 32x32 when you are doing diagonal jumps. It's working the exact same way if you play with it irl.

WelcomE25/12/2008 23:48(edited 25/12/2008 23:51)

Distance: 248.102813 MaxSpeed: 334.872619 (62.551) PreStrafe: 272.321350 Strafes: 5 Sync: 81 I can't with less than 240 :X

thopa26/12/2008 00:36

Originally posted by WelcomE~
Distance: 248.102813 MaxSpeed: 334.872619 (62.551) PreStrafe: 272.321350 Strafes: 5 Sync: 81 I can't with less than 240 :X


I tried to get as low stats as possible and got it with 239.582.
http://www.sendspace.com/file/6mljnu

(232.266 units straight forward. 232 units is the distance between the blocks if you don't use Pythagoras..)

SchlumPF26/12/2008 00:42

zylle as i said the ljstats is maybe bugged so my 239.x are correct + no demo no proof ;D

WelcomE26/12/2008 00:52

Strange thing is the CS isn't it?

SchlumPF26/12/2008 01:41

yes it is.
btw if ljstats is really buggy than we can jump further if our jump is performed in a angle to the x/y grid axis' :D

zyllEE26/12/2008 02:07

Originally posted by SchlumPF
zylle as i said the ljstats is maybe bugged so my 239.x are correct + no demo no proof ;D


i didn't demo 238.9 but i got demo of 239.5 i just made

Spaniolu26/12/2008 02:14

@Schlumpf

IMG

SchlumPF26/12/2008 04:08(edited 26/12/2008 04:18)

spaniolu that one should just be right in the case that the jump is turned in an 45° angle which the inferno jump is not
but if you look into my formula at the mainpost angle / 45 * sroot(2048) you should see how you can calculate the thing you wnated tzo calculate (2048 = those 32^2 + 32^2 in your pic)

(( edit.
forget about the formula... looks wrong to me :)

s1gh26/12/2008 06:31

Hmmm, so I'm not really sure how the HL engine works, but this is pretty interesting. How come these collision hitboxes don't disappear after you've stopped... colliding? Once you've landed on these, does it kind of "teleport" you to the actual box? Does it work in the same way that an eb transports you to the side if you land within however many units of the brush you need to land?

I mean, are you actually standing on these colliding boxes when you have landed and stopped?

cC|SpiNe26/12/2008 12:05

Originally posted by thopa
As far as I've heard the playerbox doesn't angle like that when you are strafing. It's always constant (forward/back/left/right, never diagonal).


What about when you press w+d isn't that diagonal? I don't know if you gain enough distance to compensate for the loss in speed. If this theory is ocrrect.

Fliborecord holder Tournament #226/12/2008 14:04

Originally posted by SpiNe
Originally posted by thopa
As far as I've heard the playerbox doesn't angle like that when you are strafing. It's always constant (forward/back/left/right, never diagonal).


What about when you press w+d isn't that diagonal? I don't know if you gain enough distance to compensate for the loss in speed. If this theory is ocrrect.


Yes, you move diagonally, but your player model is always the same way, so you will need a diagonal block also. This "bug" isn't something which you could use for any advantage, it's just interesting... For example, with this "bug", mappers can create 250.5xx blocks etc ^^ (Yeah it's pointless). My original idea was just to use this for /measure plugin, so it would be a bit more accurate.

SchlumPF26/12/2008 14:29

as i said, this has not to be bug for ljstats.
either we can really jump longer cuz of this bug which means ljstats is buggy (and those 245.3 are 100% correct)
or we have to say that those 245 of the enginge are wrong because it feels and is a lot easier.

Spaniolu26/12/2008 14:32(edited 26/12/2008 14:38)

Schlumpf take a look at this

and its correct because if u take the angle 0: the result its 32 and if u take 45 the result is 45,2548..

SchlumPF26/12/2008 14:57

yes that one is right ^^

itachi^26/12/2008 17:46

geeks when ur done plz sum it up i dont understand half of this

Page 3 of 5

Please log in to post replies.