|
| Re: relative ordering in graphs [message #137061 is a reply to message #137034] |
Tue, 08 June 2004 18:00   |
Eclipse User |
|
|
|
Originally posted by: none.us.ibm.com
There is a simple change going into 3.0.0 which will address crossings when
the attachment point is slightly different:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=65743
So far no one has opened a feature request for constrained ordering of nodes
in the directed graph layout. I'm not sure that we could introduce an API
and/or a behavioral change for a 3.0.x release. I know we introduced graph
layout in 2.1.2, but it was completely new code, not changes to existing
classes.
"Ray Tsui" <Raymond_Tsui@ca.ibm.com> wrote in message
news:8prbc0tlfn02lferbqa07c1esd2n0ahu2q@4ax.com...
> I see that you mentioned that you have relative ordering implemented
> but not released. This sounds exactly like what I need. We are working
> with nodes that have input terminals, and thus there is an input
> terminal 1,2...n. And thus it would be nice if the autolayout
> respected this ordering when positioning the source nodes to those
> target input terminals.
>
> Is there something we can use now? Will this be officially supported
> in the future (next 6-10 months?)
>
> Ray
|
|
|
|
| Re: relative ordering in graphs [message #137162 is a reply to message #137150] |
Tue, 08 June 2004 20:49 |
| Eclipse User |
|
|
|
Originally posted by: none.us.ibm.com
"Ray Tsui" <Raymond_Tsui@ca.ibm.com> wrote in message
news:t73cc0t5eh66em79gqamnt9se3igrbstrb@4ax.com...
> Hmmm, maybe it's time for somebody to open up that feature request :)
>
> But if this change is indeed going to allow me to predetermine the
> order by setting the targetOffset, then that is exactly what I need.
The change will swap the location of two nodes as shown in the bugzilla.
This will only occur if the number of connection crosses is reduced, so in
some cases it will not occur.
|
|
|
Powered by
FUDForum. Page generated in 0.03317 seconds
] 
Search
Help
Register
Login
Home
