Its great to see people starting to work at this!
> - If I have tag display turned off (i.e., WYSIWYG), and I
> place the insertion cursor at the junction between two
> elements, where am I? I could be at the end of the first
> element, between them, or at the beginning of the second.
> The visual display has to inform the author.
This is a problem for existing word processors, such as Word for Windows.
Usually markup (e.g. font selection) doesn't include the white space
following the last word to which the markup applies.
The ability to show paragraph stops is a great help as it provides
visual cues to the presence of the paragraph markup info.
> - What happens if I drag-select across tag boundaries, from
> the middle of one element to the middle of the next? What
> do I get when I paste that?
For simple markup like font changes (e.g. emphasis) the editor is
responsible for inserting the necessary begin and end elements as
needed to make the result legal. More complex selections across
say table boundaries are also patched up by Word for Windows, so
why can't we do the same?
> - How do I sometimes select a complete element
> and other times just select the text content of an
> element (say for replacement) without selecting the element
> markers?
Sound like the job for a menu. The end of the edit menu could show
the current levels of markup, and allow you to select the whole lot
at one go!
> I don't doubt that there are solutions to these (and
> other) interface problems. But the solutions require
> providing more information to the author, and the
> author understanding what that information means.
> Structured editing takes some getting used to.
The basic lesson here is to adopt the solutions used by the
best wysiwyg word processors. Keep thinking up the questions - they
seem an excellent way to develop guidelines for designers.
Dave Raggett
BTW I have been thinking on similar lines, but just don't have the time
right now to implement an HTML+ wysiwyg editor. Maybe in the winter ...