🚀 go-pugleaf

>>  I have a simple suggestion.  Create a new prefix character.  It would
>>  certainly be unreasonable to take C-a and make it a prefix character.
>>  (It's bad enough that backspace on some terminals is a help key.)
>>  C-z is bound to suspend-emacs for historical reasons.  I suggest defining:
>>
>>	 C-c	mode prefix character -- these commands depend on the
>>		 mode of the current buffer
>>	 C-z	user prefix character -- reserved for user

I think Bruce lebans proposal is a good one, to reserve a special prefix
character for end user, rather than "stealing" half of the keymap from
the current major mode.

FSF suggests that every major mode command should begin with "C-c C-" . 
This is a severe restriction for the major mode.
In a large major mode you can't afford to give away one half (or two
thirds, uppercase considered) of the keymap to the end user (who rarely
would use it anyway).

If you strive for consistency with already existing major modes, you don't
want to use for instance C-f, C-b, C-a, C-e, C-n, C-p for non-motion commands.
The biggest trouble I have in emacs is remembering this wealth of bindings
across different major modes. Without very rigid conventions here,I think
we will get lost. The standard motion keys I just mentioned is one example of
candidates for standardization.  The non-FSF major modes that is distributed
on the international nets often fails to conform to existing "de facto"
standard found in the official FSF major modes. This limits their usefulness.

If you want to define additional major mode specific help commands, you might
want to use C-c letter to resemble the global help map. Why not use C-h letter,
you might say, but this would shadow the global help map.

You normally want the keybinding to match the initials of the command-name,
to make it easier to remember and possible to guess.
If 'C-a is already occupied, you would want the option to use 'a or 'A.
This is prohibited by the FSF solution.

A common (and good) convention for keybindings is to group related commands
to the same letter, say :

   'C-a for weak command
   'a   for stronger command
   'A   for strongest command

This is prohibited by the FSF solution.

Furthermore, a binding with uppercase letters (as in "C-c A") is a good
way to protect a "strong" or dangerous command from being called by accident.
This is prohibited by the FSF solution.

The proposed prefix C-z is by no means ideal, but they are hard to find.
C-q is maybe a better alternative since (quoted-insert) is a very rare
command compared to C-z. C-z IS in fact a de facto standard in the Unix
world.

All in all, Bruce's proposal is a good one !

Jonas Jaernestroem      PHONE:    +46 8 719 1009            
ERICSSON Telecom AB     FAX:      +46 8 719 5319
126 25    STOCKHOLM     E-MAIL (globally) etxjojm@ericad.ericsson.se
                        E-MAIL (from Cadence) etxjojm%ericad.ericsson.se@murre