With this approach, we can use the current returnArgument: method.
I.e. just make that method not close the window (when window has
been opened as async).
Brilliant.
Two things I am wondering about:
1) should tm_dialog -a really return a plist? because we _always_
want just the token?
We should just return the token for async. I left the plist partly in
case I wanted to return something else, and partly because it was
returned in most other cases.
2) should --wait-for-input only return the arguments given to
returnArgument:… or the full data model? Here I wouldn’t say we
_always_ want the former, but 99% of the time, we would.
Yes, I'd rather have the returnArgument: arguments. We can add
something later if the full data model should be needed.
My TODO list is these changes:
• --async-window return only the token, not a full plist
• --list-windows
• --wait-for-input as described above, returning on returnArguments:
Plus demo scripts and Ruby wrapper.
Chris
_______________________________________________
textmate-dev mailing list
[email protected]
http://lists.macromates.com/mailman/listinfo/textmate-dev
