bugfix 22 antilag fixes

ReyalP · Post by **ReyalP** » Thu Apr 08, 2004 1:25 pm

Does this include the player origins not get restored after historical trace ( http://www.splashdamage.com/index.php?n ... pic&t=7391 )
Or was that already fixed in etpro ? I'm mostly curious if older versions of ETpro had roughly the same g_antilag.c and headshot detection as etmain, or was what was fixed in bugfix 22 something else ?

I'd also be interested what you did with the crouch/prone/viewangles thing.

Thx in advance for any info.

Post by **bani** » Thu Apr 08, 2004 3:58 pm

no, it's not fixed. thats a new bug for us. that's pretty serious and i've slated it for fix23.

sometimes this et code is just so excellent it makes you

ReyalP · Post by **ReyalP** » Thu Apr 08, 2004 4:17 pm

ahh, crap, I just assumed you had already seen it on the splash forums. Take a careful look at IsHeadshot and G_BuildHead too, or you will go back to the old, broken RTCW way. That is (at least part of) why headshots were always fucked in RTCW w/antilag. In RTCW, the original hit detection is done with a histtrace, but the IsHeadshot uses the most recent position

Not to mention the fact that prone and crouch go backwards in time

p2 is standing
p1 shoots p2 at t0
p2 goes prone at t1
server processes command from p2 (who is an lpb) and then p1 (an hpb)
when calculating p1s shot, the server rewinds p2 to his position at t0, but puts his head in the prone location, since prone/crouch state is not recorded in the history.

w00t!

Badhabit · Post by **Badhabit** » Thu Apr 08, 2004 7:30 pm

not sure if this is related, and I'm still testing this, but it seems that the headshot hit box for a team-mate is different then for an enemy.

ReyalP · Post by **ReyalP** » Thu Apr 08, 2004 8:08 pm

Badhabit wrote:not sure if this is related, and I'm still testing this, but it seems that the headshot hit box for a team-mate is different then for an enemy.

Very unlikely. The hitboxes themselves are in the same place (unless bani has really changed something, or there is some bizarre bug I haven't noticed yet). The *models* for the classes and teams are different. Also, the relationship between the model and the hitbox *changes* depending which way the model is facing, and the state of the animation.

Screenshots might help prove or disprove any of these theories.