geda-developers team mailing list archive

[Ivan Stankovic <pokemon@xxxxxxxxxxxxxx>]

On Mon, Dec 17, 2012 at 03:36:16PM +0000, Peter TB Brett wrote:
> On Sunday 16 December 2012 12:38:02 Edward Hennessy wrote:
> > I did the same thing in the component manager. In the git repository
> > for the component manager, there are the following sources:
> > 
> > geom-* : Boxed types representing all the primitives.
> > sch-*  : GObject types representing all the GOBJECTs in gEDA.
> > 
> > These portions of the code are decoupled well, and anyone is welcome
> > to use them.
> > 
> 
> Hi Ed,
> 
> I had a brief look at the code. This seems to be a good approach, and as 
> you suggested, using the GBoxed geometry types as properties of the 
> actual schematic objects (or would it be better to call them schematic 
> elements?) makes good sense to allow the use of the "notify" event.

Note that in that case you would get notifications on the level of the boxed
object, which might make some operations inconvenient. Because you're changing
the object as a whole, you only have the information that "something" changed,
which might mean just a small part of the boxed object, or every part of it.
But that's just a minor inconvenience and one can emit more specific signals
for those needing them (for example, see sch_arc_set_center_x() and
sch_arc_set_arc() in gparts repo).

> If we adapted the component manager code for Ivan's libeda stuff, how 
> would you suggest naming the boxed types for an "Eda" prefix? 
> "EdaGeomFoo"?
> 
> Ivan, what do you think of this?

It would be great if we could reuse the code, especially since Edward seems
to have written a non-trivial amount of quality GObject code.

-- 
Ivan Stankovic, pokemon@xxxxxxxxxxxxxx

"Protect your digital freedom and privacy, eliminate DRM, 
learn more at http://www.defectivebydesign.org/what_is_drm";