[plt-scheme] flowing text into two text% objects

On Nov 13, 2007 8:29 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> Here's one quick thought:
>
> Maybe you could have a single text% object that is displayed in two
> editor-canvas%es, similar to what happens when you split the
> definitions window in DrScheme. The scrollbars on each canvas would be
> hidden, but the right canvas would be scrolled down one page relative
> to tge left canvas.
>
> Matthew

Hi Matthew,

I think what they're after is the ability to fill, but not overfill,
an editor% with text. (/me thinks a bit more...)

Hm. I see. They could break the entire document up into a list of
editor%s, where each editor% is simply scrolled down one page. So,
"turning a page" would actually swap editor%s around in the GUI. Well,
there's only one editor% that is split... but it feels like there's a
whole bunch running around.

OK. I'll play with that, and see if I can provide some guidance as to
how they might tackle that in code. They've been wrestling with how
they might fill (but not overfill) an editor% with text when the font
size is variable.

I will say, it seems inconsistent that 'get-start-position' returns a
value, while 'get-visible-position-range' expects boxes to be passed
in by reference as opposed to returning some kind of value. I would
personally like to see the API consistent---either everything goes in
and out boxed, or everything comes back from method calls as values.
Perhaps, though, there was a reason for the mixed approach to
interacting with editor%s?

Cheers,
Matt