Re: parenting (patch)

[Lars Clausen <lrclause cs uiuc edu>]

On Thu, 03 Jul 2003, Vadim Berezniker wrote:
> Lars Clausen wrote:
> > On Fri, 20 Jun 2003, Vadim Berezniker wrote:
> > > I've fixed those now.
> > >
> > > > The standard way to specify object specific behaviour is not to
> > > > recompile but define properties. See "autoroute" for a recently
> > > > added example. [ Search for "autoroute" in the dia sources. ]
> > > I'll look into that shortly  ...
>
> After thinking about this, I can move the children list / parent
> capability attrbiute into individual objects.
>
> As for the 'parent' pointer, should I even bother moving it out of the
> Object struct? After all I cannot think of any object that cannot have
> a parent. Since Object is the 'superclass', an attribute that applies
> to all 'objects' should be inside it.
>
> Should I leave the parent pointer inside Object or move it?

There is every reason to allow any object to be included in other objects.
That's the way you'd make subdiagrams.  So leave the pointer inside Object.
The children list would obviously only be for objects that can have
children (e.g. the package objects in UML).

-Lars

-- 
Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| HÃrdgrim of Numenor
"I do not agree with a word that you say, but I   |----------------------------
will defend to the death your right to say it."   | Where are we going, and
    --Evelyn Beatrice Hall paraphrasing Voltaire  | what's with the handbasket?