Everyone here is before this design decision in GEF.

I think the intention is that it is only that top level figs
(FigNodes, FigEdges and the primitive FigText, FigLine etc) are
considered to be composite parts of a Layer. The child Figs that then
make up the detail of a FigNode of FigEdge are considered composite
parts of that Fig and so the layer is redundant.

It's a shame 2 different types of composition are being used where
only one is relevant at a time. I think I'd have preferred to have a
common interface for both FigGroup and Layer and base a parent/child
composition on that.

However at the moment we are more likely to just enforce that child
figs have the same layer setting as their parent. Note though that
that should be a one way reference only for child figs. The Layer
should only contain a list of Figs at the top level of the fig
composition hierarchy.

Bob.



2008/11/15 Dave Thompson <[EMAIL PROTECTED]>:
> Hi,
>
> I've been working on issue 1048, and during this, I found that some of the
> Fig elements on the display have getLayer() == null.  This causes some
> problems when trying to do things with text labels, such as move them.  When
> I do start to move them, I get stdout filling up with:
>
> SelectionManager: layer is null
>
> Which is coming from org.tigris.gef.base.SelectionManager.drag(int,int)
>
> I think I've found the reason why this situation occurs, but I'm just
> wondering whether it is intentional or not.
>
> Regards,
>
> Dave
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to