home *** CD-ROM | disk | FTP | other *** search
- Per Abrahamsen wrote:
- >
- > It was also slightly easier to figure out how to redefine oslash, since
- > I could define it interactively and examine what was changed in the keymap.
-
- I don't see how this is any harder in lemacs (C-h b).
-
- > I think about it as remapping text. The text just happens to come
- > from the keyboard in one case, but may as well be in a buffer.
-
- But X doesn't give you text, it gives you keysyms; it's up to the application
- (emacs) to translate keysyms to text, and that's where define-key sits. Emacs
- doesn't really have a second-level hook for doing further translation on that
- text, and this is why I don't think that thinking of it as remapping text is
- the best model (given what you've got to work with right now, at least.)
-
- > I'm not interested in anything from the keyboard that is not text, not
- > even plain-old control characters.
-
- Well let's say you've changed the ` key to generate oslash; I would expect
- C-` to then be interpreted as C-oslash, not C-`, because that's what it
- would do if you were sitting at a keyboard which came from the factory
- with oslash on it, instead of simulating one.
-
- > - Numbers easier for programs than symbols.
-
- No they're not. This is lisp!
-
- > - Symbols easier than numbers for humans.
- >
- > Events may then be either symbols or integers (plus auxiliary
- > information) which might be considered inelegant.
-
- It certainly is by me...
-
- > It is not inelegant from my point of view, since `printable' and `special'
- > keys really are different for my application (they even have different
- > colors on my Sun type-4 keyboard :-).
-
- But for my application (lemacs) they're not different at all; A and F1 and
- oslash and Control-Meta-cokebottle are the same thing: sequences of little
- plastic squares sitting on microswitches which you can bind commands to. Some
- of those commands involve inserting elements of some character set or another
- into buffers, and some don't.
-
- I'm not ruling out use of unicode (I haven't thought too hard about it yet)
- but it's going to take a lot of covincing for me to use integers as events.
- Not that using unicode necessarily implies that.
-
- > For output FSF has included Howard Gayle's code for mapping between the
- > character set in the buffer and the character set being displayed. This
- > means that it is f.x. possible to view (and edit) Latin1 text with a pure
- > ASCII terminal (or font). oslash will be stored as character 248 but
- > displayed as {\o}.
- >
- > I can't figure out if Lucid can do this. It documents
- > `standard-display-table', but then refers to `buffer-display-table'
- > which is not documented.
-
- None of the display table stuff works right now, but it (or something of
- equivalent functionality) will most likely be made to work some time in the
- future (after the new redisplay is in place.)
-
- -- Jamie
-
-