2012.03.30 21:34, John Scipione rašė:
On Fri, Mar 30, 2012 at 11:53 AM, Rimas Kudelis<rq@xxxxxx> wrote:2012.03.30 05:10, François Revol rašė:What you are trying to use here is a 2-bytes character + 2 2-bytes diacritics as a single entity. Isn't there precomposed versions instead ??No, there aren't. These bytes represent Unicode named sequences. And even if there were precomposed versions, would it make any difference? You mean we're OK with keeping Control and Opt layers for "advanced users" yet don't feel like entering more than one codepoint with one key is something an advanced user might want? :)There is a 4 byte limit on UTF-8 chars in keymap and that is not easy to change. Furthermore, I don't know how much benefit adding the extra bytes would be. If there is any way that you can make the codepoint fit in 4 bytes, do so, if not, then we'll have to think about re-architecting once again.
Well I said that there is no precomposed form for those characters, so no, I can't fit them into 4 bytes. WRT re-architecting again, I'd just like to see the X-byte limit removed altogether, or to make sure it can only be reached with an intention of reaching it. I had analyzed and compared the possibilities of Windows, Mac and Linux keyboard layouts in my BSc paper. Both Mac and Linux allow more than one (and more than I tested) codepoints to be entered with one key (or deadkey+key combo) Haiku, along with Windows, is currently way behind those two in this regard...
Rimas