Hyperjumping calculation (server part)

[HD_Morpheus]

(12-02-2010, 09:49 PM)Bayamos Wrote: Dunno if you're using a library method or not, and I doubt it really matters since it's 3x3, but Gauss-Jordan is slower than Gauss/backsubstitution; I think it ends up introducing slightly less accurate results as well if you're using floating point. /shrug

Oh well i'm not a math boy, just tested with my knowledged, as matlab used Gauss/Jordan for my tests, then passed to python code to simulate in custom 3d. Did in Matlab, then checked if same a,b,c in pyth code.
Anyway, as long as you are able to pass and receive a matrix, code could work with the math function as a blackbox.
I would like to do benchmark with the one you mentioned though

(12-02-2010, 09:57 PM)Sixxth Wrote: So, got the math - Im an engineer FFS, so no issues there for me

SO, will you use the existing animation? Actually, it seemed like there were a couple of animations depending on how high you were/how long the fall was, but whatever.

Would be awesome to have hyper jump back for sure - but how about little jumps over things, jumping off roofs etc? Same calculation I'm guessing? This is fascinating for me, prio to the MxOEmu, I always assumed that all the jumps were client side things, not server....

Client calculates if there is a building or crap like that. That's the red cross when you hyperjump or so targeting. I've never done it on real server but did on emu. Saw the flips and so.
Animations and that thing is what's sent by server, along the coords explained above.

Anyway, any question related or critics are welcome .

PS: at last! college maths serve for a purpose! XD