Wire in Logic Editor Example [message #71436] |
Wed, 19 March 2003 11:43  |
Eclipse User |
|
|
|
Originally posted by: oro7d3.netscape.net
In Logic Editor Example, if I use the default connection router, I can bend
the wires in the diagram.
But once I change to Manhattan Connection Router, I can no longer bend the
wires. Could someone please tell me where in the code which determines
whether a wire is bendable based on the connection router?
Thank you.
|
|
|
|
|
|
|
Re: Wire in Logic Editor Example [message #71527 is a reply to message #71489] |
Wed, 19 March 2003 17:47   |
Eclipse User |
|
|
|
Originally posted by: none.us.ibm.com
Manhattan router is very dumb. It will route connections on top of other
nodes in the diagram. We would like to investigate an intelligent
Orthogonal (same as mahattan, but we need to pick a new name) router, but
have not had time to address such issues.
"Sam Cheung" <sy_cheung2@yahoo.com> wrote in message
news:b5apdn$b8i$1@rogue.oti.com...
> Thanks to both of your for the help.
>
> I have a related question: Does the Manhattan Connection Router consider
the
> Bound of the source/target figures when it routes the connection? i.e. It
> will not route the connection THRU the source/target figure, it will route
> AROUND them?
>
> I study the code, it has this when it calculates the start/end direction:
> rect = conn.getTargetAnchor().getOwner().getBounds().getCopy();
>
> conn.getTargetAnchor().getOwner().translateToAbsolute(rect);
>
> So it is consider the bound when it calculate the Direction?
>
> But in my plugin, I still get cases when a connection going THRU the
> source/target figure.
>
> I am wonder it is because of the Connection Router? or the connection
Anchor
> that I have for the source/target figure (I am just some kind of Fix point
> connection anchor).
>
> Thanks in advance.
>
>
>
>
>
>
>
> "Eric Bordeau" <ebordeau@us.ibm.com> wrote in message
> news:b5a74g$s2i$1@rogue.oti.com...
> > It depends on the edit policy. When the router changes in the logic
> > example, we call WireEditPart.refreshBendpointEditPolicy() which removes
> > the edit policy for the CONNECTION_BENDPOINTS_ROLE if the router is a
> > manhattan router. This particular code is specific to the logic
example.
> >
> > Eric
> >
> >
> > oro wrote:
> > > In Logic Editor Example, if I use the default connection router, I can
> bend
> > > the wires in the diagram.
> > > But once I change to Manhattan Connection Router, I can no longer bend
> the
> > > wires. Could someone please tell me where in the code which determines
> > > whether a wire is bendable based on the connection router?
> > >
> > > Thank you.
> > >
> > >
> > >
> >
>
>
|
|
|
Re: Wire in Logic Editor Example [message #71546 is a reply to message #71527] |
Wed, 19 March 2003 18:20  |
Eclipse User |
|
|
|
Originally posted by: jwoods.journee.com
Randy,
What are the chances of getting a better router added to the plan for 2.2?
Connection routers are very reusable and very important in GEF (and like you
said - the Manhattan router is dumb). I've found Manhattan OK for testing
and prototyping - but real world end users need something better to make use
of applications written on top of GEF. Ditto for anchors. I'd contribute
one to the group, but I'm still in the core functionality stage of my
project and haven't got the time yet for usability issues (sound familiar?).
Something like a smarter Manhatten (no routing on top of nodes, etc.) with
moveable line segments (up/down or left/right) would be great.
- Jason
"Randy Hudson" <none@us.ibm.com> wrote in message
news:b5as1d$d0c$1@rogue.oti.com...
> Manhattan router is very dumb. It will route connections on top of other
> nodes in the diagram. We would like to investigate an intelligent
> Orthogonal (same as mahattan, but we need to pick a new name) router, but
> have not had time to address such issues.
>
|
|
|
Powered by
FUDForum. Page generated in 0.03316 seconds