Sparx Systems Forum

Enterprise Architect => Suggestions and Requests => Topic started by: Paolo F Cantoni on October 18, 2006, 12:49:41 pm

Title: Label rendering in Tree Style
Post by: Paolo F Cantoni on October 18, 2006, 12:49:41 pm

Selecting tree style imposes constraints on the rendering of the labels.

EA currently (799) has 8 labels (identified on Context Menu|Set Visibility>|Set Label Visibility...)

However, if you use Tree View the Middle and Source Labels (both top and bottom) will overlap.  Once you start naming the connectors, the diagram rapidly becomes unreadable.

I suggest (in Tree view) moving all the labels so that they are on the Source "Leg" of the connector, but still in their relative order (so the source labels would be at the source end of the source leg and the target labels at the target end of the source leg).

This would make the diagrams much more readable.

Thoughts, Votes?
Paolo

[size=0]©2006 Paolo Cantoni, -Semantica-[/size]

Title: Re: Label rendering in Tree Style
Post by: «Midnight» on October 18, 2006, 03:10:42 pm

Now that's a good idea. [Not that I'm surprised paolo.]

Add my vote!

Title: Re: Label rendering in Tree Style
Post by: Paolo F Cantoni on October 18, 2006, 07:06:09 pm

Thanx David.

I've had some more thoughts on the matter.  I originally wrote it with non-Associations in mind.  Particularly since EA only supports one of the rendering methods for Generalization Sets mentioned in the [size=13]UML 2.1 Superstructure (interim)[/size] (http://www.omg.org/cgi-bin/doc?ptc/2006-04-02) Specification.

Certainly, it seems to me that Tree Style is more appropriate for non-Associations.  However, there may be place for it in the case of Associations where the target metadata is the same.  However, I suspect this is unlikely...

Thoughts?

Paolo

Title: Re: Label rendering in Tree Style
Post by: «Midnight» on October 19, 2006, 03:42:06 am

Paolo,

SPEM comes to mind (that sounds a bit too strange) immediately. We'll doubtless see this kind of thing as diagrams get more complex.

At the very least, complex dependency and realization networks will benefit from tree style. If users want to show stereotypes on such links then your suggestion - with a tweak - would be a life saver.

The tweak would be an option - or perhaps EA could 'sense' this case automatically - to show a common stereotype label on the joined section of the tree. [Still, I'd hapily give that up to get the rest of the labels working as you describe.]

Just another 0.02 CAD,
David