
4 Sep
2008
4 Sep
'08
6:36 a.m.
Timothy Goddard wrote:
It looks like this code isn't really in fitting with normal Haskell idioms [...]
this is all very much true. Still I understand the application area (for this example) is interactive graph editing and there the "in-place update" perhaps is the right model for what the user of the program does (remove/insert node/edge). So at least the model of the GUI needs some notion of "state"? Of course from the GUI state one should compute the underlying and *immutable* data object for further processing. J.W.