[mcclim-devel] Reinstate #\: as command-dispatcher
Hi, The description of *command-dispatchers* at the bottom of http://www.stud.uni-karlsruhe.de/~unk6/clim-spec/27-6.html#_1436 is quite clear: #\: must be included. I'd like to commit the attached patch, but based on my recent track record I figured it would be better to ask first. Cheers, Rudi
Rudi Schlatte <rudi@constantly.at> writes:
The description of *command-dispatchers* at the bottom of http://www.stud.uni-karlsruhe.de/~unk6/clim-spec/27-6.html#_1436 is quite clear: #\: must be included.
... in the standard set.
I'd like to commit the attached patch, but based on my recent track record I figured it would be better to ask first.
I don't think that the passage you refer to is strong enough to mandate #\:'s presence in *command-dispatchers* for all applications, of which the Listener is one; instead, I think it means that programs can rely on the global value of *command-dispatchers* having #\:. I can see that there might be a problem in launching (without threads) another CLIM application from the listener, which would then have a surprising value of *command-dispatchers*; is this the case you're trying to address? Cheers, Christophe
Christophe Rhodes <csr21@cantab.net> writes:
I can see that there might be a problem in launching (without threads) another CLIM application from the listener, which would then have a surprising value of *command-dispatchers*; is this the case you're trying to address?
This problem should be solved by using the listener-view to define a more specific accept method for command-or-form that rebinds *command-dispatchers*. Any command or form will be evaluated outside of the accept method, and hence the setting will not be visible in any subprograms. -- \ Troels /\ Henriksen
participants (3)
-
Christophe Rhodes -
Rudi Schlatte -
Troels Henriksen