SAGE is big, time is limited. Python is easy, but not all students
know it and like writing code.
I'd like SAGE to have what wxMaxima is for Maxima.
That is main actions can be achieved by clicking buttons instead of
writing code.
Html is very flexible to invent all kinds of UI,
but I don't know how this would fit into notebook framework
For example when user selects the syntax (maxima/r/whatever),
it can at once propose the menus
(there could be different collections of menus even for the same
syntax, which could be loaded with notebook command)
I think this could become similar to "macros" (at least in M$ office I
know macro can be called with custom button).
I think this would play very well with interactive input idea
http://groups.google.com/group/sage-edu/browse_thread/thread/1e89fc0a71d2db6e
http://groups.google.com/group/sage-devel/browse_thread/thread/1e89fc0a71d2db6e
I understand Ted Kosan went for http://mathrider.org with similar intention
but back to sage notebook -- for now, I can imagine such macros could
generate notebook code put it in a cell and run it,
but for the math user it might be more convenient not to see the code
by default (with option to look at it when needed)
I am curious when in the roadmap this could happen?
and how good in general nb framework is for this, what should be changed/added?
--
Jurgis Pralgauskis
jabber: jur...@akl.lt; skype: dz0rdzas;
Don't worry, be happy and make things better ;)
Have you seen the interact command? Check out some of the demos here:
Have you seen the interact command? Check out some of the demos here:
>> I'd like SAGE to have what wxMaxima is for Maxima.
> Have you seen the interact command?
> http://wiki.sagemath.org/interact
Yes, I did, it's kind of what I expect (and in some sense - even more),
but with limitation to "once" (one declaration for one cell)
in wxMaxima, you can call the functionality many times in your processing flow.
Roughly what I imagine is menu of buttons by the side of current cell
http://popmokslas.projektas.lt/etc/sagenb-more_gui-buttons.png
or the buttons could be in some right/left toolbar
clicking on the button
1) either just pastes template of code (most easily achieved)
2) or calls interactive input (field for each needed argument),
executes template (or just a function) filled with the input
and gives the result
the input fields mapping to func arguments is already done in
interact, as I see.
so if some common tasks are predefined as functions
with @interact_input code for them
then they could be called many times
(but the collection of buttons should be described only once,
in the beginning or outside notebook session)
by the way, is @interact designed to return some value for further
calculations (in next cells)?
Now I am little confused with my vision...
I see two paradigms:
1) code with GUI *input* helper to speed up writing it
2) extended *interact* way, with ability to build interact sequences
(and no code to be shown)
and changing some input would recalculate the rest stuff which depend on it
Generally I see big future for case 2 in universities and schools,
of course, if interact blocks could pass values between each other
then I could construct the calculation sequence mainly with mouse
clicking (people like it).
maybe even like in http://scratch.mit.edu/ :)
Hope, I am not too unclear...
Maybe I should give more graphical explanation? (but I'm a bit lazy for it)
but first I expect just buttons and input dialogs instead of command
line style..
--
Jurgis Pralgauskis
omni: 8-616 77613; teledema: 8-657 65656;