emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: ill-advised changes [was key bindings in hideshow.el]


From: thi
Subject: Re: ill-advised changes [was key bindings in hideshow.el]
Date: Sat, 17 Feb 2001 04:16:15 -0800

   From: "Eric M. Ludlam" <address@hidden>
   Date: Mon, 12 Feb 2001 10:33:53 -0500

   >>> Per Abrahamsen <address@hidden> seems to think that:
   >If we assume people only use one overview mode at a time, we could use
   >'M-o' as prefix all of them.  The current prefix for
   >outline-minor-mode 'C-c @' is very hard to type, at least on a Danish
   >keyboard. (@ is AltGr-2).

   I like that idea.  Perhaps outline-mode, and outline-minor-mode could
   become a base behind folding modes the way comint is a base behind
   various interactive shells.  That would help things remain consistent.
   outline-mode already provides a spiff interface for mode-authors for
   building foldable interfaces.  If hideshow has additional requirements
   (beyond the regex already provided) perhaps Thi could offer
   suggestions on what might be required from outline mode to do
   that. Maybe allowing a function in addition to the regexp for finding
   folding boundaries would be good enough?

hideshow uses regexps to define a block start as well as end.  blocks
can nest properly.  outline could be implemented using hideshow since
the latter is more general.  and hideshow can be implemented using
(parts of) imenu since that is even more general.

i like "overview" as a categorization, too.  it's kind of like "meta"
yet more verbose.... :->

from a practical standpoint, i'd love to get a green light from gerd to
change hideshow before next release to use "C-c @" prefix, unwieldy as
it may be, to at least be congruent w/ outline minor mode.  people
interested in playing w/ a copy of hideshow.el with such a change can
grab it (and the change log) from:

  http://www.glug.org/people/ttn/software/hideshow/hideshow.el
  http://www.glug.org/people/ttn/software/hideshow/ChangeLog.txt

in the meantime, i think the category we are looking for is "structure
analysis / presentation".  some major modes go to great lengths to infer
structure from flat text, and some to present structure nicely and
flexibly.  what would be nice to have is a pair of frameworks that
factor out these basic commonalities.  imenu does a good job of covering
analysis.

thi



reply via email to

[Prev in Thread] Current Thread [Next in Thread]