A 05:58 97-10-16 -0700, Lars Henrik Mathiesen a écrit :
> From: "Alain LaBonté - ordi1dgsig" <email@example.com>
> Date: Wed, 15 Oct 1997 17:20:26 -0700 (PDT)
> [About 8859-15:]
> Don't worry, transition will be as harmless as possible. It will be
> a goodie for Europeans, a goodie for French-speakers (who will use
> it outside of Europe immediately), and a goodie for Finnish
> speakers, and will go almost unnoticed by others, who will anyway
> continue to have the choice.
>Alain, does that mean that you will immediately change from posting in
>CP 1252 with a MIME charset of ISO-8859-1, to posting in ISO 8859-15
>with a MIME charset of ISO-8859-1?
I expect to post with a MIME tag indicating ISO-8859-15. It will be cleaner.
I think that even Microsoft would be open to this.
The only thing required from Windows character set will be a proper
translation, without affecting original data in Microsoft proprietary
format indeed in any way.
>Partly I am just being snide here. I haven't heard of any cases where
>the 1252/8859-1 distinction makes any difference for display: If a
>system has any CP1252-compatible fonts at all, it probably uses them
>for documents marked as 8859-1 as well. But documents in 8859-1 and
>8859-15 need different fonts, and confusion here is worse.
So you do not have Internet correspondents in the following environments:
Macintosh (proprietary format)
OS/2 (in general IBM 850 character set)
8-bit enabled UNIX systems (in general a member of the ISO/IEC 8859 series)
EBCDIC 037 or 500 CECP, and many other variants
I do have such correspondents. The easiest common denominator requires an
8-bit ISO standard.
So far *there is* data loss in French even if in some case the repertoire
for French is complete in the recipient's environment (while in others it
is not, for example in EBCDIC, for the only reason that the coding space
fits, maximally, in dimension, with the coding space of ISO 8-bit graphic
character sets [191 characters, no more!], even if the encoding is
different -- and then there is not enough room, unless we have a new ISO
> [From an older mail:]
> Finally it is true that in an ideal world we should avoid for
> banking operations to use anything but 3-letter codes for currency
> identification (I do not count, though, the number of Europeans
> that invent new 3-letter symbols everyday without taking any look
> on international standards, like GBP instead of UKP, SFR instead of
> CHF, and so on -- that has to be said too -- it makes believe that
> these are the true banking codes, while it has nothing to do with
>I fully agree on the sillyness of inventing currency TLAs when anyone
>can pop down to Forex and get a list of exchange rates with more
>banking codes than they are ever likely to need. (And when a complete
>list can be found at http://www.triacom.com/archive/iso4217.en.html .)
>However, whether the UK'nians like it or not, the official code for
>their currency _is_ GBP, and UKP is spurious. Just like the ISO 3166
>codes for that nation are GB and GBR, regardless of popular sentiment.
>Lars Mathiesen (U of Copenhagen CS Dep) <firstname.lastname@example.org> (Humour NOT marked)
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:37 EDT