[Bf-committers] ODE changes
Sun, 3 Nov 2002 23:37:41 +0100
> > Now I have three questions, please comment:
> > 1. Why not move ODE out of the source tree?
> Sure, no problem. I checked it in at the same directory level as the Sumo
> physics library originally was. I'm not sure how Blender handled "external"
> libraries in the past - the only reason I checked it in is so that the
> user doesn't have to download ODE separately to compile.
> Hans said:
> > External code should be checked into CVS on the vendor branch only (for
> > easy updates) anyway.
> So I guess ODE should be moved into "the vendor branch", wherever that
That was the "extern" directory, at the same level as the intern directory. I
hope Hans will lead this kind of stuff (because he is used to doing it :) ).
> > 2. Why can't we make changes to the ODE tree to get it to run? I know
> > there should not be two source trees but if ODE does not work for us,
> > why not make it work and possibly post updates to the author once in a
> > while?
> My only concern was keeping Blender-modifications in sync with the official
> ODE source tree.
Ok, I'll check if the CVS version will build on OSX and wait before submitting
the fix. In the mean time I found some referencences to non-existing sumo stuff
in the tree. Didn't you have problems with that? I can commit the fixes, mostly
simple make file stuff.
> > 3. How active is development on ODE? The 0.03 version I used seems to
> > be the most recent and is dated 25 December 2001.
> The CVS version is quite more recent and has some bugfixes for slider
> some LCP solver fixes, a new version of the triangle mesh collider should
> coming out soon, etc. Development isn't massively fast, but there are a few
> people who do work on it regularly. Check out the mailing list at
> http://q12.org/pipermail/ode to get a feeling for the community activity.
Ok, thanks, I have just downloaded it and played with the examples. Looks
impressive and made me wonder why he left his former employer...