Message-Id: <9212031536.AA27509@sc.zib-berlin.dbp.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
In-Reply-To:  Paul Taylor's message of 03 Dec 92 13:35:10 +0000.
Date:         Thu, 3 Dec 92 15:34:47 +0000
From: Robin Fairbairns <Robin.Fairbairns@CL.CAM.AC.UK>
Sender: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
To: Multiple Recipients of <LATEX-L@vm.urz.Uni-Heidelberg.de>
Subject:      Re: accents
Status: R


|> There may well be standard uses of 8-bit ascii for the small number of accent
|> letters in French and German - standard, that is, in the French-speaking and
|> German-speaking countries - but the international standard (for whatever acce
|> on whatever letter) must remain Knuth's. It is needed, for example, where
|> accented languages are quoted in an English context.

Paul evinces some confusion about the usage of the word "standard"
here.  On a point of fact, ASCII (an American - i.e., ANSI, now -
standard) is a 7-bit standard.  It can be regarded as the USA national
specialisation of ISO 646 (though in fact ASCII came first).  The fact
that we all use ASCII most of the time (unless we're IBM-mainframe
based) reflects the curious fact that for much of the history of the
manufacture of computers, the USA was the best country at that game.

There are French and German (and for that matter, British)
specialisations of ISO 646.  These sacrifice various characters for
appropriately accentuated or other characters of interest to people
who think in French or German (or British).

But ISO 646 isn't the only standard: there's ISO 8859 (which has
several different parts) and ISO 10646 (aka, now, as Unicode, which is
a multi-byte character set).  ISO 8859-1 specifies a "Latin" character
code; in it, you can specify texts in a large number of Latin-alphabet-
based languages (including, nominally, English, American, French _and_
German).

What Knuth has defined may be thought of as a standard, but it's only
such an object within the TeX world.  It's not going to metamorphose
into an 8-bit character set any time soon, since Knuth has (quite
rightly) decided to go back to work on The Art.

|> I don't know what these 8-bit conventions are. Now you may put that down to
|> my ignorance, but it is some evidence that they are not as "standard" as
|> Gaulle Bernard would like.

An interesting assertion.  There are masses of standards whose content
I wot not of.  I would never claim that the fact _I_ don't know about
the content of the standard for grommets means that grommets are
somehow "less standard".

|>                            It also means that if you come to me (or the TeX
|> implementation I maintain at Imperial College) with your 8-bit documents,
|> they won't work unless Don Knuth, Karl Berry, Frank Mittelbach or Rainer
|> Sch\"opf has put them in to (Unix-C) (La)TeX without my knowing it.

Quite so.  Standards need to be implemented.  But the standard isn't
less of a standard while Frank or Bernard or whoever is writing its
first ever TeX implementation.

|> Of course French and German speakers will use their national 8-bit standards
|> without problems (I hope) within their own language communities, and I have
|> no wish to force them to do otherwise, but please preserve the international
|> standard.

ISO 8859-1 _includes_ ASCII.  ASCII mostly maps directly to Knuth's
curious code table.

|> PS I never use tabbing anyway.

Me neither.  I'm horrified to discover its present behaviour!  I'm not
at all surprised that the non-native-English speakers are upset about
it.

Robin

PS: those who know me will observe that someone's touched my
"standards sore spot" again...   :-)