Re: Outlook & the Euro

From: Alain LaBont\i\ (alb@sct.gouv.qc.ca)
Date: Mon Jun 22 1998 - 16:18:20 EDT


A 11:25 98-06-22 -0700, Erik van der Poel a écrit :
>Hi Alain,
>
>Alain wrote:
>>
>> It would be nice if for email, in order to communicate with standard 8-bit
>> environments, there be also an *option* so that translation from and to
>> Latin-9 (ISO/IEC 8859-15) be done by Microsoft products.
>>
>> This would allow communicating in a standard way in the 8-bit realm outside
>> of Windows environments. Otherwise there is indeed a character loss in
>> other environments, and the characters lost are less useful than those in
>> the C1 space used by 1252

[Erik] :
>I don't understand this. Are you saying that there would be a character
>loss in the C1 area (0x80-0x9f)? If so, under which circumstances would
>these be lost?
>
>Or are you saying that even if a user agent sends mail with a charset
>label that says "windows-1252", the receiver will not be able to
>understand it?

[Alain] :
At least one person was attentive (; . I myself saw that I had said the
contrary of what I meant just after the message was sent and I waited for
such a question in order to rectify. I am often a distracted person very
dyskinetic at the keyboard (my fingers sometimes seem to make decisions out
of the thoughts I transmit them, as if they were making automatic writing
(: !), and I was extremely tired when I wrote this, after a month of travel
and jet-lag suffering on 3 continents almost without rest. I should have said:

"Otherwise there is indeed a character loss in other environments, and 8 of
the characters lost in the C1 space used by 1252 are *more* useful than
those 8 in the Latin-1-G1-space that Latin 9 reallocates. [Therefore it is
preferable in the interchange to lose the Latin 1 characters that Latin 9
replaces to the profit of the otherwise unfortunately lost ones that are in
the nice Microsoft 1252 code page]."

[Alain] :
>> which is, btw, a nice and avant-gardiste
>> character set since the very time it was available (but it remains private,
>> with interchange problems for the external word -- other private
>> environments also take advantage, in general, of standard interchange).
>>
>> What I suggest is not a revolution, just a sensible option for those who
>> wish so (and it should perhaps be a default in French, Finnish and
>> euro-currency aware 8-bit environments which are concerned all the time).

[Erik] :
>ISO 8859-15 will probably be implemented by a number of vendors, but it
>will take some time until a large percentage of the users start using
>those versions. Until then, it might be wise *not* to make 8859-15 the
>default when sending mail.

[Alain] :
On the contrary, if you look carefully at what is lost of the Latin 1 space
when exchanging Latin 9 data, it is not only wise but clever to use latin 9
as a default for most of Europe [and I would say the two Americas, at least
in Québec (; ] , as even most users of Latin 1 will not even remark the
difference (that said Latin 9 should be tagged as Latin 9 and not as Latin
1, that will be honest).

So far a lot of false data is tagged Latin 1 while it is not, when it comes
from the Windows environment (I understand Microsoft-platform-based
environment developers, they had no much choice as they could not get their
character set blessed while it remained useful and most Latin 1 users did
not make a difference anyway, since in most cases it did not contain for
them the C1 *graphic* characters -- only French and Finnish users were
affected, plus a few other with fancy needs for special quotes). It would
be a way to correct this practice, making it more standard, without
touching Windows systems, just hacking some of the characters of 1252 back
and forth for email, rescpecting round-trip integrity for the full
repertoire of latin 9.

The main character that will really make a *visual* difference in Europe
(outside of French-and Finnish countries) is the one allocated to the EURO
SIGN, and that is the one that matters for all Europe.

The character of Latin 9 that are not in Latin 1 (in this zero-sum game)
are as follows:

Code space allocation Latin 1 'meaning' Latin 9 meaning

  A4 ¤ (CURRENCY) EURO SIGN
  A6 ¦ (BROKEN BAR: not ASCII one!) CAPITAL S CARON
  A8 ¨ (STDALONE DIAERESIS) SMALL S CARON
  B4 ´ (STDALONE ACUTE) CAPITAL Z CARON
  B8 ¸ (STDALONE CEDILLA) SMALL Z CARON
  BC ¼ (ONE QUARTER: system incomplete) CAPITAL LIGATURE OE
  BD ½ (ONE HALF: same comment) SMALL LIGATURE OE
  BE ¾ (THREE QUARTERS: same comment) CAPITAL Y DIAERESIS

Those who would even like to minimize changes could even, as Eudora Pro
does between ASCII and Latin 1, analyze the message sent, and if it is the
common intersection subset of Latin 1 and Latin 9, could even tag it as
Latin 1 in such cases, the default remaining Latin 9 (Latin 9 could be
tagged in the same way without any difference).

Alain LaBonté
Québec



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:40 EDT