> I love guestures! The delete/cancel guesture and the scribble guestures are the only one's i use (know about?)
> However, how about mapping pretty much every major feature to a guesture?
I know this sounds like a good idea, but trust me, it really isn't. People have tried this before. It doesn't
work. You won't be able to remember all the gestures, and the system will never be very good at recognizing
them. And new users have a really hard time learning to use systems that are gesture-based. Basically I
avoid gestures as much as possible - sorry =). Although, there is some new stuff in development that does
sort of use gestures, which you might like...
> The most awful thing at the moment in Shapeshop is the fact you can't see what you've drawn!
> I have it running on my laptop, and a trackpad is very hard to draw with, so i'd like to be able to see a trail of what i'm drawing.
Umm...you don't see a blue line as you draw? You should. It should look just like in the demo videos. It might
be your laptop, I turned OpenGL line smoothing on by default, and it seems like some embedded graphics
cards just don't draw lines when this is enabled....
> Well it can export, why not import?
Heh. Although ShapeShop can export a mesh, ShapeShop modeling is not mesh modeling. It's implicit modeling,
like metaballs. Basically this means that there is no "mesh", there is a functional surface (kinda like a nurbs patch)
that gets meshed for rendering. That's why you can turn the mesh resolution up and down arbitrarily. So,
to "import" an object, it would have to be converted from the input mesh to a scalar function over R^3 that
has an iso-surface that approximates the mesh. This is basically still an open research problem, the ways to
do it take minutes-to-hours of processing, and would be brutally slow to edit.
That said, I do want to support this. I have ideas about how to make it fast, but it's outside my current
research, so it won't happen soon...
> Highlight to see what something does:
Yea that's a good idea. I'm moving away from using MFC, to my own custom OpenGL window library,
so that I can port to OS X and Linux more easily. I'll add tooltips to the list of stuff to re-implement...
> I've read your other statement about open source, and this is part of your PhD research. Once that's out of the way (timeframe anyone?) i'd love for you to re-build it open source.
Unfortunately I'll be doing my PhD for about 3-4 more years =\.
> For God's sake, you must keep developing this program. It has the potential to blow most 3D applications away.
> ADVERTISE!!! Think about what would happen if you got a larger userbase?
See, these two are kind of related in my mind. The problem with a large userbase is that you have to support it or it goes away. I'm not sure I want ShapeShop to "get big" until I can commit a reasonable amount of time to it. Also, once it gets big, you have kind of committed to a feature set. It's really hard to take features away when thousands of people are using them. Right now it's in a sort of Alpha-Tech-Demo stage, where I can still completely change the interface if I think there is a better way to do it. I couldn't do this if thousands of artists were using those features - they would hunt me down and have my head. This is why most 3D modeling software interfaces are so hard to use - the developers get locked in by the userbase that is paying their salaries. That's also why I don't ask for money - once someone is paying me, I kind of have to listen to them.
(Also, I would be embarassed taking money for ShapeShop in the state it's in now. Maybe in a year or two, once I get some of (what I consider to be) the critical stuff implemented, like proper interoperability, texture import/export, large model support, etc...)
Anyway, thanks for your feedback!
-RMS
