I suspect that most users will only need multiple keyboards, as Michael suggests, to ENTER TEXT DIRECTLY, not to type in some magic Unicode mode.
For those of us who need access to characters outside the range of currently installed keyboards (or who are not "touch typists" in that language---I do a poor job typing Japanese, for example, and might want the characters directly), typing in the hex value is not that much better than using something like CharMap or another equally crummy, hard-to-find-when-you-need-it utility.
As you point out, there should be a magic Unicode mode associated with every keyboard/IME, rather than having to switch to an L-1 keyboard or its surrogate to type the hex value. The decimal value is probably better than hex for human entry anyway: anyone with 10-key experience can master it and I know many translators who ONLY type extended Latin characters that way (they are quite speedy about it) and refuse to use compose utilities or other crutches that you would think touch typists who are paid by the word would embrace.
The other alternative is a keyboard controlled pop-up (IME style) that can let us get a the characters we're interested in in an organized fashion based on various properties (one of which is the code point: probably the default).
Hex entry is better for a number of reasons, maybe...
Another option would be to use the arrow keys (or the decimal separator key) to increment values above 9 when typing in the code. Using arrow keys seems valuable, since you could use the left arrow to decrement from F (one keystroke for F, one for E, etc.) and the right arrow to increment from A (one for A, two for B, etc.)
The most distant code points would be U+CCCC and U+DDDD with 13 keystrokes each (magic key plus four groups of three keystrokes). The least keystrokes (5) would be those composed entirely of 0-9,A,F, which make up 75% of the potential combinations. Assuming equal distribution of characters (which is a poor assumption today) the average number of keystrokes is probably just over six.
Maybe decimal isn't so bad after all... the minimum AND maximum is always six.
Just a thought,
From: John Cowan [mailto:email@example.com]
Sent: Saturday, June 05, 1999 11:07 AM
To: Unicode List
Subject: Re: Hexadecimal in many scripts (ISO 14755)
Michael Everson scripsit:
> Ar 09:12 -0700 1999-06-04, scríobh John Cowan:
> >It is annoying enough to have to type in random 10646 characters by
> >code number. It is doubly annoying to have to use some escape
> >mechanism on the keyboard to be able to enter the code number
> >at all.
> Apple's way cool WorldScript switched keyboards easily, and I imagine will
> be a model for multiscript implementations of the UCS. :-)
But consider the use cases:
<typing along happily in Arabic>
<need a U+2323>
<switch to Latin keyboard>
<enable magic Unicode mode>
<type "2 3 2 3">
<disable magic Unicode mode>
<switch to Arabic keyboard>
<try to recover train of thought>
John Cowan firstname.lastname@example.org
e'osai ko sarji la lojban.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:46 EDT