XForms: problems and suggestions

kuefner (holger.kuefner@FernUni-Hagen.de)
Tue, 03 Mar 1998 11:35:03 +0100

# To subscribers of the xforms list from kuefner <holger.kuefner@FernUni-Hagen.de> :

Hi all together,

I am a xform beginner and first great thanks to T.C. Zhao and Mark
Overmars. Xforms is in my opinion a good written library and it simple
works! No problems by compilation, address paths, unknown error numbers
or something else.
I have "clicked" a medium complex user interface (buttons, sliders, free
objects) for demonstration of a new distributed computer architecture in
about one day and wrote the connections between the GUI and my C
procedures on the next day. That's easy working. (I used lib 0.88.)

Now some slightly problems which occours by working with xforms (maybe
it's of interest for some of you):
> Hardcopy of a xyplot object with fdesign or fd2ps is not supported.
> Hardcopy of a frame with text is not correct. In the hardcopy the text is always fixed above and outside the frame (in the GUI the text is inside the frame).
> A label of a object defined with fdesign has slithliy other coordinates (5-10 points) as in the the program itself. Try it and design a label in a frame (or text) with fdesign and afterwards change the label in the program with "fl_set_object_label".
> fdesign has problems to catch all members of a group correctly. In more than five attempts I tried to group six objects. Only the half of them are defined as group members by fdesign which can be checked in the fd-file. Finally I edited the fd-file t

Now a suggestion to io/to callbacks:
It would be nice to have an object parameter in the io- and timeout-
callbacks as in the "normal" callbacks. Hereby it would be possible to
avoid global pointers for object structures which will be used by the
procedures. Surely it is possible to define a data structure which
covers these pointers and pass the pointer of the structure by the data
parameter of the callback, but this is not straight forward and the
structure is global, too. So an object parameter as in the "normal"
callbacks will be conform and save easy handling.

Thanks for supporting this library.

H. Kuefner
To unsubscribe, send the message "unsubscribe" to
xforms-request@bob.usuf2.usuhs.mil or see
XForms Home Page: http://bloch.phys.uwm.edu/xforms
List Archive: http://bob.usuf2.usuhs.mil/mailserv/list-archives/