Time we switched to unicode? (was Explanation of this Python language feature?)
Rustom Mody
rustompmody at gmail.com
Tue Mar 25 01:00:23 EDT 2014
On Tuesday, March 25, 2014 9:59:48 AM UTC+5:30, Chris Angelico wrote:
> On Tue, Mar 25, 2014 at 3:08 PM, Rustom Mody wrote:
> > And Chris is right in (rephrasing) we may have unicode-happy OSes and
> > languages. We cant reasonably have unicode-happy keyboards.
> > [What would a million-key keyboard look like? Lets leave the cost aside...]
> Actually, it wouldn't be that bad. Unicode allows for only 1114112
> characters (thanks to UTF-16), of which only the first three planes
> have any actual characters on them (so, a maximum of about 200K
> characters). All you'd need would be a system that organizes them
> (using their hex codepoints isn't exactly useful), and you could type
> any character with a maximum of, say, 6-8 keystrokes; Huffman coded,
> of course, so the average would be 1.5 keystrokes per character
> actually used. Add one meta key: Charset. Hold that and press L and
> you get (say) λ, U+03BB; Charset+Shift+L is Λ, U+039B. Ctrl+Charset+L
> might give you a Cyrillic л (U+043B), and Ctrl+Charset+Shift+L would
> then be Л (U+041B), the upper-case version of that.
> Emacs users would love it.
Its already there -- and even easier
Switch to cyrillic-jis-russian (whatever that is!)
and I get л from k Л from K
Full layout
+----------------------------------------------------------------+
|1 !|2 @|3 #|4 ” |5 :|6 ,|7 .|8 *|9 (|0 )| − _|= +| ё Ё |
+----------------------------------------------------------------+
| й Й | ц Ц | у У | к К | е Е | н Н | г Г | ш Ш | щ Щ | з З | х Х | ъ Ъ |
+------------------------------------------------------------+
| ф Ф | ы Ы | в В | а А | п П | р Р | о О | л Л | д Д | ж Ж | э Э |\ ||
+-----------------------------------------------------------+
| я Я | ч Ч | с С | м М | и И | т Т | ь Ь | б Б | ю Ю |/ ?|
+-------------------------------------------------+
+-----------------------------+
| space bar |
+-----------------------------+
The catch is in the claim to huffman coding
Huffman coding requires a statistical distribution
Whose shall we take
And now the discussion is political not technical
More information about the Python-list
mailing list