From: Alain LaBonté (alb@sct1.gouv.qc.ca)
Date: Tue Jul 27 2004 - 07:58:33 CDT
À 18:24 2004-07-26, Mike Ayers a écrit:
> > In less pedantic terms:
>
><SNIP/>
> Oddly, that was the pedantic explanation I sought.
[Alain] Am I supposed to find this nice?
> > Any national group is group 1 by definition according to ISO/IEC 9995.
> > Group 2 is a Latin supplementary group to access those
> > Latin-script-written languages not accessible with a national group 1
> > also using Latin script. Other groups are still not numbered and their
> > actual access not standardarized.
>
> I am again baffled here. If "any national group" is group 1,
> then my U.S. keyboard layout, a German keyboard layout on a U.S.
> keyboard, a German layout on a German keyboard , and Michael's Irish
> Unicode setup, are all group 1? Certainly I misunderstand this. More
> pedantry, /si vous plais/.
[Grammar teacher] "si vous plais" should be "s'il vous plaît": literally in
English "if it you pleases", i.e. "if it pleases you" (^8= ), so it should
please me?
[Alain] Group 2 is fixed, but incomplete per se, and it needs a group 1,
for one reason: it contains all accents used in ISO/IEC 6937, for example,
but none of the basic 26 Latin letters to which these accents are supposed
to apply. Now since we were never able to standardize the group numbering
system (beyond 2 groups), and go beyond the Latin script (because of one
American company opposing to this, or perhaps one individual working for an
American company, in 1991), then group 1 is supposed to be *the* group
which contains *the* 26 basic Latin letters and anything else that
"nationals" judged necessary for their needs. This group cantaining the 26
Latin letters is *the* primary group as far as ISO/IEC 9995-3 (which
defines group 2) is concerned.
Now, as we know, even if the Latin script is perhaps used on all
keyboards sold in the world (there might be exceptions, I don't know one),
it is not the end of the world either (we knew it ot once -- I, for one,
would have liked to at least standardize entry of Vietnamese letters since
we were bound to the Latin script, but I was prevented to go beyond the
ISO/IEC 6937 repertoire at the time). National keyboards *typically* need
to support other scripts (at least 85% of this planet's inhabitants use a
script other than the Latin, even if the Latin script is perhaps the script
in which is written the biggest number of unrelated languages -- that
creates other problems - like the problem of multiple groups for only one
alphabet, the least of those problems).
So there is a need to enhance the group selection model (the "layout
switching" mechanism invented by an IBM team, in which was my respected
friend and colleague Dr Umamaheswaran). I invite all those interested to
join ISO/IEC JTC1/SC35/WG1, which will again try to do this (13 years after
the first try).
> > [ISO/IEC 9995-1]
> > 4.13 level select: A function that, if activated, will change the keyboard
> > state to produce characters from a different level.
> >
> > 4.10 group select: A function that, if activated, will change the keyboard
> > state to produce characters from a different group.
>
>[/|/|ike] These definitions, as well as the definitions of "level" and
>"group", don't seem to make particular distinction between the two. Does
>any hard distinction exist?
[Alain] As I said in my previous mail, these definitions are not the best
of definitions. The distinction is but intuitive, you have to see the
diagrams where labeling makes the difference: on a key label, the levels
[in one group] are in the same column [like on an American keyboard], and
each group occupies it own colums (given the small room on one key, one
could but imagine 3 groups labeled: group 1 to the left (like on an
American keyboard [you will see that letters are typicaly not centered]),
group 2 to the right, and group n in the center (like a group for Cyrillic,
Greek, Arabic, Hebrew, Japanese kanas, and so on).
Alain LaBonté
Québec
This archive was generated by hypermail 2.1.5 : Tue Jul 27 2004 - 08:00:33 CDT