[Lumiera] regarding the focus of the Lumiera Project

Ichthyostega prg at ichthyostega.de
Sun Nov 18 00:35:04 CET 2012


Hello Farid,

in addition to the quite complete answer by Benny, I'd like to address
your question regarding the (current) focus of the Lumiera project.

The focus of the project hasn't changed, nor has the roadmap, or the general
approach taken. I am well aware that there is something specific, maybe even
peculiar with that approach. Often, OpenSource projects started with (re)-
implementing some kind of functionality or existing solution, and then, while
using that functionality, discover what features could be added.

Personally, my situation is completely different. I do not need to "discover"
that exploring a video camera can be fun. Also, I do not need to "discover"
that editing a feature film or documentary poses quite another set of specific
challenges. Rather, my starting point is the observation, that most, if not
all existing editing solutions, while being basically well suited for the
work, tend to get into your way and impose some quirks, workarounds and
repetitive tasks. And I am convinced this is not due to fundamental
limitations, but due to oversimplified assumptions in the
initial design, which can't be amended after the fact
by any shitload of "features" or fancy plug-ins.

Probably you know already our introductory page in the documentation website,
which summarises the vision and the "fundamental forces" within our specific
approach: http://lumiera.org/documentation/user/intro/intro.html

What I've indeed learned to understand much more sharply is the challenge
when your goal is actually to allow open-ended combinations of basic building
blocks. For example, generating some audio data and making your speakers buzz
is one thing, but coming up with a generic "output sink", which adequately
captures the expected properties of sound output in some typical studio setup
and then just working against that abstraction is another story.

So the difficulties we were facing at times were not so much of a technological
nature -- what we do is pretty much mainstream technology -- but rather on the
level of assuming the right things and capturing the nature of the problem.
Seemingly, it's a constant stream of identifying one very specific problem,
trying to understand it properly, trying to come up with a solution, or even
more often, finding a way to postpone the solution of that problem without
incurring too much collateral damage. I am well aware that it's difficult
to communicate that kind of work to anyone not involved into the details.

If there is anything like a crisis or concern, than it's the fact that the
work on the GUI frontier has slowed down so much. A lot could be done in that
realm, and completely separate from the work on the engine. Unfortunately
I can't push that frontier ahead and simultaneously work on model and towards
engine and system integration. And what's seemingly even more unfortunate in
our current environment, we'd need someone there willing to get really involved
and immersed into that topic, not just coming up with some random nice ideas.

-- Ichthyo






More information about the Lumiera mailing list