--- In
cybalist@yahoogroups.com, Miguel Carrasquer Vidal <miguelc@...>
wrote:
>
> On Sun, 16 Mar 2008 23:57:11 -0000, "tgpedersen"
> <tgpedersen@...> wrote:
>
> Just to complain about Yahoo Groups...
>
> What I see is:
>
> >pres.
> >*dúbo:
> >*dúges
> >*dúget
> >...
> >*dungént/dúmbont
>
> The headers say:
>
> User-Agent: eGroups-EW/0.82
> X-Mailer: Yahoo Groups Message Poster
>
> so I'm assuming Torsten is using the Yahoo web interface for
> posting.
>
>
> The message is sent as multipart/alternative (text + HTML).
> The MIME-headers of the first part state:
>
> Content-Type: text/plain; charset=ISO-8859-1
> Content-Transfer-Encoding: quoted-printable
>
> The first line is a lie, as Torsten clearly was posting in
> UTF-8.
It's worse than that. Yahoo actually assumes the message is in
Windows-1252, and procedes to convert the curly single and double
quotes to straight single and double quotes, corrupting 6% of
continuation byte patterns, to convert it to ISO-8859-1. The effects
on Latin-A can be seen at
http://tech.groups.yahoo.com/group/JRW_test/message/16 . There are
only two ways to get round this:
1) Submit the post from an e-mailer. That will break up threads if
you did not receive the message you are replying to by e-mail.
2) Use an encoding that does not generate control characters with
high-bit set, e.g. UTF-7. UTF-7, however, make the archives
unreadable for those who rely on Internet Explorer, as it does not
support UTF-7.
It is not nothing that the group's recommended encoding is Latin-1
(ISO-8859-1).
Richard.
P.S. Solutions to cybalist_admin @ yahoogroups.com , not cybalist @
yahoogroups.com, please.