On 18/01/12 15:21, Nat Echols wrote:
On Wed, Jan 18, 2012 at 7:13 AM, Ed Pozharski<[email protected]> wrote:
The first time you call any command that need a target map to be defined
(e.g. real space refinement), the map chooser gui comes up. But then
the command that initiated it does not really get executed, and you have
to call it up again. Now, this is just a minor nuisance as it only
happens once per session. But I wonder, is there something that
fundamentally prevents the initial command from being executed?
+1
This always annoyed me too - since I like to keep clicking buttons
until I get what I want, it's also only a minor nuisance, but I
suspect (based on painful experience) that many novice users will
simply freak out when the interface doesn't behave as expected.
Yes, I agree this is bad gui design.
The problem, if I recall, was that the "OK" button needs to somehow
store whatever function it was that caused the map chooser dialog to
appear. This is doable for scheme (and python) functions, but to store
the function when the dialog is invoked from deep c++ code was
non-trivial. I maybe able to fix this though. Scheduled for 0.8.
So, how *should* this work?
Here's a possible solution: When the refinement map become undefined the
map-using functions down the right are greyed-out (and inactive) and the
tooltips for these icons say how to define the map (press the "Map" button).
Comments?
Paul.
