Hi Michael,
thanks a lot for your answer.
On Mon, Aug 18, 2008 at 6:56 AM, Antje <[EMAIL PROTECTED]
<mailto:[EMAIL PROTECTED]>> wrote:
Hello,
I try to develop a GUI with gWidgets (TclTk) but I guess my little
application already starts to be messy...
First question: How shall I handle updates of widgets triggered by
event handling? So far, I simply took the widget-object I want to
update as a global variable but that's not a good style. Any
suggestions how to handle this?
It's possible to specify a "user" object when registering the callbacks,
I think. The "user data" is then passed as the last parameter to the
handler. Another good way to control the scoping is nesting environments
(e.g. through closures). That is, your object does not need to be global
(in the workspace) but it can be in a parent environment. In fact,
global variables won't even work when you put your code into a
namespaced pacakge, because the namespace will be locked.
Can you give me a simple example, please? I'm not sure if I understand
correctly how this can work. (Or maybe there is already some kind of
documentation I can check for these possibilities...)
Second question: What are my possibilities to stay flexible? For
example: I though it makes sense to use a gpanedgroup for a
checkboxgroup with variable number of items (to make it scrollable
if it's getting too large). But this does not work like I thought
and the items simply disappear at the end of the panel.
I don't think the paned group is what you want here. Split paned widgets
are to allow the user to control the allocation of space in a GUI,
usually in either the horizontal or vertical directions. If you're GUI
is growing too large, you'll need to make it so that only a portion of
the GUI is visible at once. One way to do that is with scrolled windows.
See the use.scrollwindow parameter for ggroup().
Oh, thanks a lot. I was not aware of the scrolling opportunity if the ggroup.
Third question: How to structure the GUI? Builing, Sizing, Arraging,
filling with content (not every object has some initial content).
For example, I realized that I have to set the svalue for the
gpanedgroup after adding the objects.
Well this is an art and depends on your GUI. Split panes are difficult
in that they require one to specify a numeric value for the split
position. This is a similar problem to specifying the initial size of a
window.
Is there anything I can do beside, setting a fixed size with pixel values?
For example if I'd like to initially maximze the window... (I know that it's
possible with TclTk).
The simplest way is just to guess a good initial position. This
is usually not too bad. Adjusting the split automatically might not be a
good idea, since it should be under the control of the user.
Of course, I just set it initially.
Ciao,
Antje
I don't know, but I have the impression to create my GUI quite
unstructured and I'm afraid to loose overview soon...
The small examples for the usage of different components are very
useful but I'm looking for some hints on how to create a more
complex application (what to do when and how).
Thanks a lot for every tip!
Antje
_______________________________________________
R-SIG-GUI mailing list
R-SIG-GUI@stat.math.ethz.ch <mailto:R-SIG-GUI@stat.math.ethz.ch>
https://stat.ethz.ch/mailman/listinfo/r-sig-gui
_______________________________________________
R-SIG-GUI mailing list
R-SIG-GUI@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-gui
