closures as Tk commands?

From: Lars Thomas Hansen <lth_at_cs.uoregon.edu>
Date: Mon, 17 Jul 1995 11:37:59 -0700

Are there any plans to allow true closures as Tk commands, or anyone
looking into implementing these? Having to use lists for the commands
tends to cause a very undesirable proliferation of global variables and
procedures for implementing Tk commands, and is very un-Scheme-like.

There are clearly some problems with allowing true closures as commands
-- Tk presumably only allows strings for the commands, so STk must
maintain some system-internal table which maps string identifiers to
closures, and garbage collecting this table entails discovering which of
the commands are still being used by live windows -- but I think it is
doable. It would be _very_ nice to have. If there is interest, I would
even volunteer to discover how it can be done. Is there interest?

--lars
Received on Mon Jul 17 1995 - 20:43:37 CEST