home *** CD-ROM | disk | FTP | other *** search
Text File | 1991-03-06 | 1.5 KB | 40 lines | [TEXT/GEOL] |
- Item forwarded by BURBECK.S to MACAPP.TECH$
-
- Item 5157929 14-July-89 13:53
-
- From: KORNHER Scarab Research, C Kornher, PRT
-
- To: BURBECK.S Burbeck, Steve
-
- Sub: Re GC Hacks
-
- Steve,
-
- I agree. The scheme I mentioned does put a burden on the programmer and it can
- be broken easily. It is an inelegant solution and can carry significant size
- and execution time penalties for complex networks of objects as well. I would
- like to see the option of true garbage collection in future versions of
- OP/MacApp.
-
- I was suggesting that if garbage collection caries a significant performance
- burden, then a true garbage collected object class could be created and
- inherited into whatever objects really needed it, without burdening the rest of
- the system with the overhead. This would, of course, require multiple
- inheritance.
-
- The difficulty with cloning my object trees really stems from the need to
- create "deep clone" ( cloning objects that the cloned object references)
- behavior for each object. I am going to show my ignorance and ask: does any
- language have automatic deep cloning? It would be nice to have it in MacApp.
- Of course there is the problem of knowing just what to clone. I do not yet
- have an idea of how to make automatic deep cloning work well, but it ought to
- be possible. Any automatic deep cloning scheme that I can envision would
- require some programmer input.
-
- { NOTE: My AppleLink address has changed }
-
- Chris Kornher
-
-
-
-