Re: simple 3d modeler

[Gregor Mückl <GregorMueckl@xxxxxx>]

On Thursday 14 April 2005 20:49, Stephen J Baker wrote:
> Gregor Mückl wrote:
> > It's a shame that there's no free modeller that can be called "easy to
> > use" yet. My own work on a modelling tool hasn't come to any usable
> > results yet, either. Hey, I'd *really* love to put a plug here, but not
> > being able to do so after more than two years is quite frustrating :-/.
> > Oh well, I'm working on it ;-)
>
> It is a *terrible* shame - and one I tried hard to rectify with the
> PrettyPoly project (http://prettypoly.sf.net).
>
> Unfortunately, writing a good 3D modeller is an INSANE amount of work.
> Certainly more than writing (say) a paint program.
>
> The world needs another OpenSource 3D modeller - but it's hard to imagine a
> new team making much of an impact.
>

Well, I know how much effort lies before me before I can call the stuff I'm 
working on "usable". However, making it easy to use is an another, entirely 
different matter. And to be honest: that's the part I fear most. I don't want 
to produce another obscure UI, but I don't feel that I'll be getting it 
right, either.

> FWIW:  PrettyPoly got to the point of loading, displaying and navigating
> around 3D models - with Python acting as the 'glue' between GUI, core
> editing functions and viewer (kinda like Maya uses 'Mel' scripting). 
> Things kinda bogged down over the actual editing part and somehow the
> project just fizzled out with developers leaving in ones and twos until no
> more work happened.
>
> It was just flat out too much effort for a part-time team of five or six
> guys.
>

It's almost the same here, except that I would consider 5 or 6 devs one hell 
of a team to go with. We're 2 active developers at the moment and the program 
is in a state where presenting our work would most likely do us far more harm 
than good.

Anyway, I'm just about to get started on basic mesh modelling operations. 
Maybe I can show off an early version by the end of the year: loading, 
saving, basic mesh modelling and possibly an obj exporter or basic material 
editing and with very much luck even a simple yafray integration.

> > Even worse, a good portion of the blender community is very reluctant to
> > change. As it is the UI provides a very streamlined workflow to those who
> > truely have mastered this program and with every change to the interface
> > these masters show a fear that their productivity could be hurt - and
> > they do so quite loudly in my oppinion. And I haven't yet talked about
> > how the UI structure directly provides the program structure for the
> > internals...
>
> Yes.  Those who 'get it' are often quite intolerant of those of us who are
> genetically incapable of getting it.  They always say that I should try the
> tutorials, read the book or watch the video's - never understanding just
> how hard I've TRIED all of those things.  Sometimes they accuse me of
> laziness and an unwillingness to learn new things.
>
> There might perhaps be a case for 'forking' Blender to make a second
> version that has a more conventional (even if less ergonomic to some) user
> interface.
>
> I've explored that option - but the blender source code is an utter mess
> with GUI stuff spread all over the sources.
>

The state of the source code is one of the reasons that made me turn my back 
on blender as a developer. Otherwise I'd most likely have done a couple of 
bigger patches right now. Once we had worked on a patch to add a "native" XML 
format to blender, but that didn't get very far because blender's internals 
are as obscure as the interface.

> > Anyway, blender *is* the most powerful 3D tool there is at the moment.
>
> Well, the most powerful OpenSourced 3D modeller at least.  Maya is without
> doubt more powerful (and runs on Linux) - but it costs several thousand
> dollars and is heavily node-locked so you can't even run it on your laptop
> when you travel and your desktop machine when at home.
>

Sorry, I was a bit sloppy when I wrote this. I actually meant "most powerful 
open source 3D tool". And Maya is just way too expensive. Well, student 
editions start at around 400 Euros for Maya Complete, but that's still quite 
expensive for students. 

Softimage has reduced the price for XSI a lot. Softimage|XSI Foundation 4.2 
officially costs 549 US-$. Didn't they have 5 digit prices once? I remember 
having seen prices at around 20.000 Euros for Softimage quite some time ago, 
if I'm not totally mistaken. Unfortunately, I don't know anything about how 
that software handles because I couldn't get the demo running on Linux so 
far :-(.

Regards,
Gregor