Re: MSKLC restrictions (was: Ways to show Unicode contents on Windows?)

On Wed, 31 Jul 2013 15:30:02 +0200
Otto Stolz <otto.stolz_at_uni-konstanz.de> wrote:

> Another limitation: Apparently you cannot define sequences
> of dead keys. If I am mistaken, I’d appreciate any hint
> on how to define dead key sequences, in the MSKLC framework.

See http://blogs.msdn.com/b/michkap/archive/2011/04/16/10154700.aspx .

> E. g., for a Greek-polytonic keyboard the natural approach
> would be to have just seven dead keys, three for the three
> different accents, two for the two different spiritus, one
> for the subscripted Iota, and one for the diaresis. Then,
> a character with multiple diacritica, such as 1F96 “ᾖ”
> (GREEK SMALL LETTER ETA WITH PSILI AND PERISPOMENI AND
> YPOGEGRAMMENI) would be entered with the corresponding
> multiple dead keys (three, in this example).

> Compare this to the current Greek-Polytonic Windows keyboard
> layout, which is virtually uncomprehensible, because you have
> to memorize one particular dead key for every single combination
> of diacritica (more than 30 different dead keys).

This is one of the reasons I prefer NFD over NFC. I can then type 4
characters and see what I've got at every step. You're also more
likely to get it input correctly. A few years back I coded up NFC
input of IPA characters using a KMFL keyboard, and it was painful to
see application after application leave every intermediate step in the
resulting text. Applications have got better, but one of the problems
with 'just following the API', as Martin Hosken put it, is that the
relevant part of the API isn't easy to knowingly lay one's hands on.

Richard.
Received on Wed Jul 31 2013 - 14:38:32 CDT