[Mailman-Users] ISO-8859-1/Latin1 vs UTF-8

[Bernd Petrovitsch]

Sorry for wrong threading but I accidentally deleted the last email
here:
> One reason is that the server may very well translate the encoding
> based on negotiation with the client.  (I guess you could argue that

Yes, but *if* the encoding is negotatied, a default value makes not that
much sense (apart from the situation that absolutely no negotiation takes
place).

> it should remove the charset attribute from the META tag if it does.)

Technically: *If* the webserver hands out a file and tells explicitly
it's encoding, it must have it right. If that means "parse the .html
file, find appropriate headers, convert the file and rewrite all
relevant headers etc." than he absolutely must do it.
At least in Apache (1 and 2) there is ATM no mechanism in there which
tries to do that AFAIK.

And you are right with the unquoted part: A default makes here absolutely
no sense.

> A second reason is that admins will occasionally translate encodings
> and not even be aware that some users who are too smart for their own
> good have used META tags.

Well, but the real reason for the problem is somewhere else - namely in
between the admin and the users. And we have probably now the situation
that most of the .html producing users and tools are dumb ebough to
actually need such crazy options ....

BTW I wonder why there are no "check the files on a webspace" scripts
out there which simply check this ...

	Bernd
-- 
Firmix Software GmbH                   http://www.firmix.at/
mobil: +43 664 4416156                 fax: +43 1 7890849-55
          Embedded Linux Development and Services