Re: [K3d-development] CVS release notes - K-3D 0.3.0.6

[Romain B. <rom...@ya...>]

> Won't try to hide it, gang - it's been a damned
> exciting week, another good one for the scripting
> buffs:
> 
> * Added property (i.e. name/value pair) support to
> k3d::icommand_node. 
> This allows an object to "advertise" all of its
> public properties, and as you'll recall, is an
> integral part of our future gui-separation work.
> 
> * Added support for easily advertising any k3d::data
> object as a property.  That's right - any k3d::data
> object that supports the name() concept can become a
> public property in one line of code.

Very interesting, do you have any example script? :)

> * Completely refactored the JavaScript object model.
>  Got rid of the many "detail" classes so that
> 
> 	MyDocument.Objects.Create("Mesh");
> 
> becomes
> 
> 	MyDocument.CreateObject("Mesh");
> 
> ... similarly,
> 
> 	point.Location.Local = [1, 2, 3];
> 
> becomes
> 
> 	point.local = [1, 2, 3];

Got rid of ObjectCollection in Python engine too, but
I'm not sure about Location. I have trouble to read
paths in Python, since point pointers can't be
compared within the scripting engine as it's done in
C++: every pointer is wrapped and real object pointer
is private information. export_geometry.python gets
point list, but I can't figure out how to assign a
point index to each pathpoint. May I wrap
k3dIPoint::PointID ?

> FWIW, things were the way they were to support the
> ugly, practically-gone k3dIEditable way of
> supporting undo/redo.  I'm happy to be rid of both.
> 
> * Even better, the new JavaScript code "attaches"
> methods and properties to objects at runtime, so
> that each object is automatically the union of
> all the underlying interface capabilities.  This is
> gonna scale significantly better than the old set
> of rigid classes, and is (finally) taking advantage
> of the flexibility of an interpreted environment.

Is this why you removed the ShowViewport() and
HideViewport() functions from CPythonCamera ?

> * Best of all, the new properties in
> k3d::icommand_node are automatically inserted into
> the JavaScript objects - so the JavaScript object
> capabilities will automatically grow as new
> features are added to the C++ objects, without any
> coding.
>
> * In practical terms, what all this means is that
> with these changes we get about an
> order-of-magnitude more things that can be scripted,
> with another order-of-magnitude in the short term as
> I finish the k3dVariable-to-k3d::data conversion.

Hope this doesn't imply "without any documentation
work" ;)

 Regards,

  Romain


__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/