Kaixo! On Sat, Jan 05, 2002 at 09:29:55PM +0100, Erika Pacholleck wrote:
> Assumption: > A diacritic is the small part added to a letter to form a > digraph, like acute would be the diacritic for e with acute. > This is at least what I mean with it. Also, a diacritic has no existence of its own (I mean, it is never used alone, but on top, bottom or middle of a base character). And it genreally doesn't change the widtth of the resulting di or trigraph. That is a difference with a digraph or trigraph formed by several characters, like 'th', 'ae', 'tch', etc. > Question: > Is there a difference between the diacritics which dumpkeys > shows as being synonyma? How do I have to understand synomym? > > Example: > My linux/kbd combination says in dumpkeys that caron is a > synonym for circumflex. So as one would generally understand > synonym it means flatly spoken the same. But to my knowledge > the caron is an inverted circumflex and all example fonts > show z with caron as z with a little v on top. The reason linux loadkeys works like that is only for simplicity and due to technical cosntraints. loadkeys has only a limited range of dead keys. then a combiantion of a dead key and another key can produce a given char. THere is no real link with accents or digraphs here; of course that is for what dead keys are generally used, but I mean the computer doesn't know about it; and there are also cmbiantions unrelated with diacritics; for example circumflex + 2 to give upper2, it is just a convenient way to type it, that's all. Now, linux kernel only defines a few dead keys suitable for western languages, and that's almost all. So, instead of modifying the linux kernel, dead keys have been used in a creative way; for 8 bit encodings, like iso-8859-* it is enough, as generellay there aren't simultaneous c-caron and c-circumflex, for example. Now, it doesn't work for utf-8, for sure. > Would anyone please comment on this, links wellcome, > my list is just at the beginning. Thank you. I think I have been one of the people (ab)using the loadkeys system in such way; that was several years ago, when even using iso-8859-1 accents was difficult on the console. I noticed that it would be possible to allow use of iso-8859-2 accents by cheating with the meaning of the dead keys. An ugly hack, but that did it job for that time. Now, for utf-8 support, that needs to be changed, but the whole console keyboard support needs to be changed, as currently there is hardwired limit on number of dead keys definable, but also on number of compose combinations (only 256 I think) -- Ki �a vos v�ye b�n, Pablo Saratxaga http://www.srtxg.easynet.be/ PGP Key available, key ID: 0x8F0E4975 -- Linux-UTF8: i18n of Linux on all levels Archive: http://mail.nl.linux.org/linux-utf8/
