Examples of complex checkbox treeviews

15 Apr 2008 - 2:17pm
6 years ago
2 replies
710 reads
Pawson, Mark
2007

We have a thick desktop app that uses a checkbox treeview to show data
layers that are available to display in a separate window. However,
although Checkboxes look nice and consistent in this display they are
the wrong control to use in some cases. A checkbox implies that more
than one data layer can be chosen, however some data layers are mutually
exclusive, others are not. On first pass it struck me that either the
checkboxes should only appear next to those data layers that are
inclusive and disappear when a mutually exclusive one is chosen. Or have
a treeview that mixes radio buttons and checkboxes.
Is there any good examples of such complexity available for public
viewing? I have checked Vista guidelines and a few pattern sites with no
luck.

Thanks

Mark

Comments

15 Apr 2008 - 3:21pm
SemanticWill
2007

The problem, I think, you are going to run into with check-boxed
tree-views is the expected behavior when multiple sub-item "views"
are selected. It could be an [OR], [AND], [XOR] -- so what is the
actual behavior when selecting multiple things? Are you [OR]ing the
data views, or traveling down a tree of exceptions - for instance:
All Fruit > Apples >> Only Grannysmith OR All Fruit > Apples >>
McIntosh which would join those two sets?

Is this what you are asking?

Functionally - tree views details pane is a design pattern you
could use, but it's broken for this purpose. This kind of data
really is best served with a dynamic visualization - thingk of
building a semantically rich graph (data nodes/sets of data are
visually represented, and not dumb graphics, but semantically
meaningful so you can do analytics against them) so users can
visualize the data sets and how their interactions
include/exclude/change the outcome dataset.

One other thing, is that you may end up - as I have - having to spend
a decent amount of time working through all the set theory edge cases
for how this works, what is displayed, and what the user expected to
be displayed. Visualizing set theory is not an easy thing if your
target is not quantitatively inclined.

- Will

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=28087

24 Apr 2008 - 11:05am
Pawson, Mark
2007

Yes Will, I think you have nailed it. Sorry I took a long time to
reply. It seems to get more complex everytime I think I have an idea.
The data is geological in nature, geological surfaces, seismic
sections, well logs, faults etc etc. These can be displayed in
various windows either 2D or 3D. The problem is how to communicate
the control of what can be displayed in these windows in the tree
view. For some of the data nodes the treeview pattern works, but
others have functionality attached to them that causes the tree view
Parent Child relationship to break. Any examples of these rich
visualizations you talk about to stir the creative thinking would be
appreciated.
Thanks

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=28087

Syndicate content Get the feed