Still trying to kill that pesky morph......
Karl Ramberg
karl.ramberg at chello.se
Mon Oct 22 05:03:36 UTC 2001
John Hinsley wrote:
>
> Andreas Raab wrote:
> >
> > John,
> >
> > > I think what I was missing was that WorldTimeMorph (at least, the
> > > initializeLocally part of it) is all class side and thus won't behave
> > > like "normal" morphs in this respect.
> >
> > No, no, no. I would be _very_ surprised if #initializeLocally would not
> > return the instance it just created. If it does, you can remember it (e.g.,
> > just assign it to some variable). If it does not - then it needs to be
> > fixed. A class method that creates an instance should almost always return
> > the instance it created (what's the point in return values otherwise?).
>
> Well, I can only speak as regards what happens (I'm not up to speed on
> theory and am always ready to be corrected!):
>
> WorldTimeMorph initializeLocally (creates a wtm)
>
> WorldTimeMorph delete (brings up the debugger -- MessageNotUnderstood:
> delete).
>
> Turning Ned's code into a method "kill" deletes it.
>
> But here it is, if you want to take a peek. Note that the morph is
> Karl's work. I just did a (literally) one word debugging exercise on the
> initializeLocally method to get it to run (that method was kindly put
> together by Karl just so I could bring it up without using a port which
> I envisaged using for something else with the Rolodex) and popped Ned's
> code in as a kill method.
I think it's about time Karl cleans up the code so the
morph behaves the way it should
I'll do some work on this asap and mail it
Karl
More information about the Squeak-dev
mailing list
|