[ODE] some quick notes on 0.6...

[Terry L. Triplett]

No comment on the rest of the critical commentary which seems a little
strong.

On 6/22/06, Rodrigo Hernandez <kwizatz at aeongames.com> wrote:
>
> Although its true that the primitive is work in progress, ODE is still
> on a beta stage, hell, its not even suposed to be part of a Linux
> distribution, if we were to make 0.6 a completelly stable full-feature
> library, we would still be trying to get it out of the door.
>

Just a comment on this particular bit.  ODE *is* out the door - "beta" or
not (though I'm not seeing much indication that ODE is "beta", considering
it's well respected, used in commercial games and has a solid community).

Linux distributions support the distinction between mutltiple levels of
stability, so the fact that ODE might be made available as a package for
linux distributions should not be a major point of contention.  If assigned
the appropriate status (unstable at present), ODE can be made available to
Linux users via their distributions while still giving some idea of the
developmental stage of the library.  Expanding publicity by creating
packages for Linux distros is not necessarily a bad thing - it may attract
more ODE developers that can bring focus on the areas that need work.

In the larger sense, maybe ODE needs to rethink how development goals are
viewed and communicated,  because clearly some parts are very stable while
other areas are experimental/unstable.  Is the development strategy really
to target a magical "1.0" release where everything is perfect and complete?
Most people these days understand this to be more of a marketing technique
rather than a true assessment of maturity and completeness.  Real libraries
that are actually used by actual developers constantly evolve to meet the
needs of the times.

Seems to me that a more reasonable approach is to take the long view,
realizing that development is never really finished, and give clear
indication with each ODE release which parts are stable and which parts are
not.  Forget about 1.0 - just release updates as major milestones are
reached and be clear about the areas that need attention.  Target
features/areas of functionality that need stabilizing, not magic numbers.
The former provides a way to communicate to potential contributors which
areas they could contribute to, the latter doesn't communicate much at all.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://q12.org/pipermail/ode/attachments/20060622/7be983e8/attachment.htm