Hi--
I rewrote the Nexus spanning tree code to give the spanning tree a
strong bias toward selecting subclass and instantiation relationships in
the graph. Exploring the graph in Walrus makes a *lot* more sense now.
http://netjam.org/spoon/objects.graph.gz
-C
--
Craig Latta
improvisational musical informaticist
www.netjam.org
Smalltalkers do: [:it | All with: Class, (And love: it)]
Hi Jon--
> The big movie requires the narration .AIF file in order to play.
Oops, fixed.
> One thing I found that was really useful when I did a graph analyzer
> is to include the names of the references (inst var name or index).
Yes indeed; just like all the nodes can have labels, the references
can as well. It's certainly essential, I just haven't written them out yet.
Oh: the data file for the movie's graph is available[1]. If anyone
wants to explore around it in Walrus and tell us of interesting stuff,
I'd be delighted. :)
thanks again,
-C
[1] http://netjam.org/spoon/objects.graph.gz
--
Craig Latta
improvisational musical informaticist
www.netjam.org
Smalltalkers do: [:it | All with: Class, (And love: it)]
Hi--
Well, I broke down and wrote some code to calculate a non-cyclic
spanning tree for an object memory reference graph, so now I can use
Walrus[2] for visualization. It's nice! Check out the demo movie I made
(in small[3] and original[4] versions).
This could be the basis of an extremely powerful extension to the
interpreter simulator; I assume it'd be a fun Croquet project. Alan
Lovejoy suggested the name "Nexus"; I like it. More to come...
thanks,
-C
[1] http://netjam.org/spoon
[2] http://www.caida.org/tools/visualization/walrus
[3] http://tinyurl.com/2akd83 (movedigital.com)
[4] http://tinyurl.com/33gah7 (movedigital.com)
--
Craig Latta
improvisational musical informaticist
www.netjam.org
Smalltalkers do: [:it | All with: Class, (And love: it)]