1. I think that the "editability" is a property of the model which contains
the elements not of the elements

2. I think that graying up the property panels (and perhaps adding a
"[ReadOnly]" tag to the profile in the explory and somewhere in the
propertypanels is enough)

maas

2008/4/8, Bob Tarling <[EMAIL PROTECTED]>:
>
> I was thinking that when the user selects a model element that we
> immediately call the model subsystem to determine if it is editable.
> If not then we call a setEnabled(false) method on the prop panel that
> will iterate through all controls and set their enablement status.
> That would immediately indicate to the user that the element is not
> editable by the greyed out fields.
>
> I'd need to build something into FigNodeModelElement and
> FigEdgeModelElement to do similar there.
>
> A model subsystem method such as isEditable(Object modelElement) would
> be useful.
>
> It may be that the model subsystem should also reject changes in some
> way should some other system try to call its mutators but I was hoping
> that the GUI would indicate this to the user.
>
> Let me know if you have any ideas and I'm happy to try the GUI
> implementation.
>
>
> Bob.
>
>
>
> On 08/04/2008, Tom Morris <[EMAIL PROTECTED]> wrote:
> > On Mon, Apr 7, 2008 at 7:52 PM, Bob Tarling <[EMAIL PROTECTED]>
> wrote:
> >  > http://argouml.tigris.org/issues/show_bug.cgi?id=4994
> >  >
> >  >  Any suggestions on the best way of testing whether a model element
> is
> >  >  part of profile or the user editable model?
> >
> >
> > I would suggest that the test should be against whether something is
> >  part of the primary model or not.  This would cover not only ArgoUML
> >  profiles, but XMI files which contain external references to other XMI
> >  files.  I don't think we want to require the application to handle
> >  this.  Instead there should be a way to load things "read-only" that
> >  the Model subsystem enforces.  The trick will be to find something
> >  which is not too expensive.
> >
> >  The alternative way to do this, which is less satisfactory from a user
> >  point of view, is to handle notifying the user when it's time to save
> >  the file and ask if they want the other linked files to be written out
> >  as well.  This would be pretty straightforward to handle, but isn't as
> >  nice for the user.
> >
> >  Tom
> >
> >  ---------------------------------------------------------------------
> >  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]
>
>


-- 
-
http://www.marcosaurelio.com
*1984 +2057

"It is slightly better to be simple than correct."

Timothy Finin

Reply via email to