[fa.editor-p] presentation systems and editing

C70:editor-people (06/03/82)

>From EAK@MIT-MC Thu Jun  3 00:51:41 1982
I think it'd be useful to summarize some ideas that ECC@MIT-AI is
working on for his thesis since he's not on this mailing list,
and becuase they provide a useful way of thinking about some
recent issues.  This message is possibly too terse; if so I
apologize.

First there is the concept of a presentation system.  It's
purpose is to present a reprsentation of an object.  Some
presentation systems are quite simple (e.g. presenting text,
which is what most of our editors do), others more complex.  A
complex presentation system might present a large datbase and be
constantly changing how and what it's presenting, for example to
omit useless detail.

And then there's editing.  Editing an object can be expressed as
operations on the object itself (which will cause the
presentation to update in a real-time editor), or can be
expressed as operations on the presentation, in which case it is
the job of the system to invert the presentation change in order
to change the object.  Both sorts of editing are useful.
Sometimes we want to edit in terms of structure, sometimes in
terms of what we see.

There's actually also the concept of editing the presentation in
order to change the style of presentation but not the object.

In these terms it can be seen that Z is an editor that tends
toward editing the presentation (what's seen on the screen) and
maps that back into changes into the underlying object (a text
file).  EMACS tends toward editing the underlying object (a text
file, sometimes interpreted to have more structure via parsing).
The difference becomes more noticable as the presentation becomes
fancier.