How the ground boost and teleporter jump works

[setup]

I work on a small article about OB. Unfortunately I didn't have much time to progress on it lately. Maybe I'll be able to finish it this summer (and maybe not :<).

[AL!EN]

I was planing to make an Article section. Posting about the groundBoost was just a test balloon(trial run). I dropped the idea because it would require a considerable amount of time to make such a site. nobody cared to an extend I was looking for, so I shifted priorities.

Well, there is my motivation again
That Article section could also host a set of HOWTOs for installing defrag and training. Let's see. This is going to take a while.

It can be made as simple link to forum category:

| Home | How to play | Rules | Guides | Servers | Records | Board | Pinboard | Contact |

I requested it several times.

[<hk>]

I just stumbled over the code being responsible for teleporter jumping.


In TeleportPlayer() in g_misc.c

Code: Select all

  // spit the player out
  AngleVectors( angles, player->client->ps.velocity, NULL, NULL );
  VectorScale( player->client->ps.velocity, 400, player->client->ps.velocity );
  player->client->ps.pm_time = 160;   // hold time
  player->client->ps.pm_flags |= PMF_TIME_KNOCKBACK;


[TittenIgnition]

explain it? o.O

i was gonna ask how teleporter jumps worked, but i kept forgetting.

[<hk>]

That's just the function that handles the teleport itself.
The interesting part is this Line:

Code: Select all

player->client->ps.pm_flags |= PMF_TIME_KNOCKBACK;

It says that the player is set to a state where he would be after getting hit.
I posted it into this thread because the GB works the same. Just check the 1st post.

[Le another black niger]

i made gb timings measurement ingame with script + bot

teleport: 160ms
pgb (simple knockback, no pre jump): 56 ms
pgb with pre jump: 264ms from wall, 272 ms from floor
pgb with pre with early shot from wall (have to fit attack to 5 frames before ground): 264ms of slick
rgb (no pre jump): 200ms
rgb with pre jump: 264 ms
rgb with pre with early shot (most early is about 200 ms before landing for horizontal shot, more for knocking down one): 264 ms of actual slick

nt.run pointed out to skim code, which actually produces such long and smart gbs
it enables with <-200 ups on z axis (and sets some dmg flag i guess)
flags, gotten with this flag enabled, seem to keep working during its operation time
measured skim time is 248ms

[Le another black niger]

well, i see now, this flag actually extends different flags timing, even enabled before
basically you have to get pg knockback for <50ms before landing or rl knockback for <200ms before landing

[dqopb]

I am either right or nobody cares at all.

Sorry, I joined this forum too late =)

I guess that they removed the PMF_TIME_KNOCKBACK from this if() statement in CPM which would make the effect 10 times stronger than in VQ3.

Ice in VQ3 has same strafing physics as airmove. But in CPM ice is like ground without friction. Setup said the same in other words:

They removed the complete condition
...
removing that first condition causes the cpm ice behaviour

But there's another inaccuracy in your explanation. This code is only for VQ3.

Code: Select all

float pm_accelerate = 10.0f;
float pm_airaccelerate = 1.0f;


In CPM there's pm_accelerate = 15.0f for ground/ice (and pm_airaccelerate has different values in different cases). So it is 15 times stronger in CPM. And this is why the yaw turning rate for a CJ in CPM is faster than in VQ3 (because 15 > 10), and the max possible strafe speed on ground in CPM is greater for the same reason. The turning rate for a groundboost in CPM is even faster than for a CJ (roughly: 15 > 15_minus_friction).

is there any practical use for a vq3 ground boost? Can it generate a faster start than regular strafing?

You cannot actually boost in vq3. However, you can continue strafing as when midair. Same as on a slick surface. So you can delay your jump. Khetti told it in other words:

The main use is to change jump timing, allowing you to land in different places or skim walls differently.
It is probably easier to just do a plasma hop, but if for some reason that isn't practical you can use a GB.

There are such situations, indeed. Seen in runrs:

runrs[df.vq3]00.20.712(G.Ukraine).dm_68

There are awkward pads right after plasma. PJ is too strong there, but a short slide is just fine.
What for telejump, the advantage (in vq3) is well seen when you need 410+ ups at jump time right after tele, e.g. to cover a subsequent gap.

fdsj2[df.vq3]00.10.552(FM.Russia).dm_68

FM continues strafing as when midair after teleporters, for a fraction of a second (not longer than 160ms). This allows to gain 420-440 ups on the ground. Without it quite a longer ground time required to cover the subsequent gaps.
(I hope it is OK to upload these demos, because they are public in defrag.ru tables).

i made gb timings measurement ingame with script + bot

Good job! PGB from floor is the longest, no surprise though =)

[Le another black niger]

shouldn't they all have the same 248 ms slick time?
different numbers could be a result of later landing

just made test map with 0.5s telejump

[bunny]

what about le water? i mean we all noes yer only gets 0.5 times le air acceleration theres. but can yer gb under watur?