[Bf-committers] Need summer of code advice

[Tom M]

Ton,

> It has a lot of benefits that is is
> simple, fast to use, and not bloated.

simple, fast and not bloated are not the inverse of capable.  Those
are issues of how things are designed.

> Designing a real video/audio editor is a different task, and with a
> different focus. For that it would be much better for devs to help out
> jashaka or cinerella or whatever-is-there-now instead.

Hmm what features were listed that weren't also useful for editing
video needing 3D video integration?  Resolution independent edits are
needed due to the required speed.  Ie you mentioned motion tracking -
the tracking is generally done on a low resolution pass first, then
redone on a final resolution.  Resolution independent also impacts
hardware requirements - these changes would make Blender better
scalable for high resolution editing - even with pretty beefy hardware
my understanding is that it still lags a bit.

Jahshaka is generally considered quite a ways off from being ready for
professional usage.  Cinellera is developed by a mystery development
team that does occasional code dumps.

Blender is really the only open source software that is reasonably
close to being a professional level non linear editor or compositor.

> Same goes for the Compositor. It is not the ultimate Shake replacement,
> it is an extension of the Blender rendering pipeline, which happens to
> work as well when not used as such, but that is not the focus.

I don't see how becoming a 'Final Cut' or 'Shake' replacement would
require a 'change of focus'?  About the only two differences I can see
are acquisition (ie I don't think that support for firewire; etc.
would be reasonable)and publishing (ie the publish to web features
that many modern NLEs are promoting).  The primary difference would
likely be what is needed 'first'.  (Ie integration of particles; and
motion tracking are earlier priority for Blender whereas they would
tend to be towards the end of the list for a typical NLE and
Compositor).

> For the rest of it, like the useful quick-blend option for effect
> strips, that sounds really like the regular patch we get in anyway...
> that is maintenance and won't be 2 months of coding work :).

He mentioned earlier that he would like to tackle a number of moderate
sized task rather than one huge task.  I don't think that is an
unreasonable approach, and in my mind gives a much larger chance of
success.

LetterRip