On Sat, May 29, 2010 at 03:55:30PM +0200, Thilo Schulz wrote: > On Saturday, 29. May 2010 15:15:09 Michael Menegakis wrote: > > That still doesn't answer the question, I guess it's a response to the > > guy that didn't know what it was. I mean, if the approximation has any > > sideeffects on the end-game. > > Possibly for movement, I think the inverse square root might be used for > computing normalized motion vectors. As an example on the effect this could > have is that you can (or cannot) jump certain heights that would not (or > would) be possible to jump with an exact calculation. > Otherwise, the effect should really be hardly noticable, as the max. relative > error is 0.175% over all floats. > > -- > Thilo Schulz
We've (referring to a mod community) have noticed that there are some differences in id Software's release of Quake 3 and those executables that are self compiled. Apparently the culprit in that instance was 'fast' vs 'precise' floating point integers, not that it's *entirely* related but still somewhat interesting. > _______________________________________________ > ioquake3 mailing list > [email protected] > http://lists.ioquake.org/listinfo.cgi/ioquake3-ioquake.org > By sending this message I agree to love ioquake3 and libsdl. _______________________________________________ ioquake3 mailing list [email protected] http://lists.ioquake.org/listinfo.cgi/ioquake3-ioquake.org By sending this message I agree to love ioquake3 and libsdl.
