Ghost Jumper

[Janik Schneider]

hey jussi, what do you think about a ghost jumper, you know, for example in racing-games when you have a very good round you can drive against your ghost, sorry for bad englisch. so in dsj 4 you can jump vs. your best flight. i think it´s a good idea and i hope that i am the first with this suggestion, if not, sorry.

[SuperAdo111]

Niceeeeeeeeeeeeeee idea^ I want it too

[alexander]

this would be pretty cool. if it's possible to use the replays or whatever.

then we could enforce replay submission on a top 10 world record. that way we could use the world record as a ghost jumper, and thus have the option of selecting a WR-ghost or a PR-ghost. (and we'd also be more aware of cheaters.)

the ghost jumper should be toggled when you start the jump (i.e. click the mouse).

[skiflyer11]

hey jussi, what do you think about a ghost jumper, you know, for example in racing-games when you have a very good round you can drive against your ghost, sorry for bad englisch. so in dsj 4 you can jump vs. your best flight. i think it´s a good idea and i hope that i am the first with this suggestion, if not, sorry.

as option nice idea

[Jussi Koskela]

This would require double the cpu power (=too much).

[alexander]

how much CPU power does DSJ4 require?

[Jussi Koskela]

how much CPU power does DSJ4 require?

About 1.5 GHz. Maybe 90% is physics calculation.

[Janik Schneider]

i´m happy about that my idea is not only good in my opinion
Jussi, can´t you manage it this way, that you can turn the ghost jumper on and off and when your cpu power is not
good enough you jump alone^^
i think for beginning, for example in beta 3 you can make, that we can take two replays and watch them as one
--> two different jumpers/jumps in one replay (but only on the same hill)
maybe you can look you personal record togheter with the worldrecord....

[alexander]

About 1.5 GHz. Maybe 90% is physics calculation.

well that's nothing. the ghost jumper would obviously be optional. "everyone" has at least a duo-core 3.0GHz these days, so for us non-laptop users it could be a cool option. but if 90% is physics calculation, why would using the ghost jumper (which essentially is a more advanced way of playing a replay) result in 2x CPU demand?

[aslan97]

or like to be able to let go of two replays and compare them together

[piter_ski]

Yeah, a kind of replays comparison would be great, somewhen in the future, but it will be probably extremely hard to create, I guess.

[Jussi Koskela]

About 1.5 GHz. Maybe 90% is physics calculation.

well that's nothing. the ghost jumper would obviously be optional. "everyone" has at least a duo-core 3.0GHz these days, so for us non-laptop users it could be a cool option. but if 90% is physics calculation, why would using the ghost jumper (which essentially is a more advanced way of playing a replay) result in 2x CPU demand?

The physics must be re-simulated every time that the replay is played. Otherwise the replay files and the memory requirement would be huge.

[alexander]

The physics must be re-simulated every time that the replay is played. Otherwise the replay files and the memory requirement would be huge.

define huge plz. if it's anything "anyone" can afford in 2011 it should be memory.

[Jussi Koskela]

The physics must be re-simulated every time that the replay is played. Otherwise the replay files and the memory requirement would be huge.

define huge plz. if it's anything "anyone" can afford in 2011 it should be memory.

I haven't done the exact calculation. At least for the replay files it would be too much. Second option would be to precalculate the data before every jump, which would cause ~10 second loading times.

[Ascot]

The physics must be re-simulated every time that the replay is played. Otherwise the replay files and the memory requirement would be huge.

define huge plz. if it's anything "anyone" can afford in 2011 it should be memory.

I haven't done the exact calculation. At least for the replay files it would be too much. Second option would be to precalculate the data before every jump, which would cause ~10 second loading times.

10 sek. are to long .
I think .