On Wed, 6 Sep 2000 22:33:00 -0500 "Eric Arseneau" <eat at huv.com> wrote:
>Can someone explain to me why delete'ing a morph, does not pass this on to
>the submorphs ?  Is this an expected feature ?  The problem I ran into was
>that a submorph down in the morph tree needs to release some resources when
>it is no longer in need.  I had assume that delete would trickle down the
>morph tree, but it does not seem to.  It does disconnect itself from its
>parent, but the children are left out in the cold.

Eric,

A couple of reasons come to mind:
1. It's extra work that's rarely necessary.
2. It would make certain tasks more difficult. Presently it is possible to delete a morph and later add it somewhere (where it was originally or someplace quite different) and all the submorphs are still attached.

Possible alternatives:
- Implement the method you want with a different name (#destroy, #deleteCompletely, #deleteRecursively or somesuch). There was some discussion about this a few months ago.
- Implement some sort of registry where morphs that need post-delete notification can get it. The registry could simply check its contents periodically and inform morphs that were no longer in the world of that fact.

Cheers,
Bob