Re: Warnings, RFC 1522, and ISO-8859-1

I'm not much of an expert on i18n, so I can't really comment on
what is "right" or "wrong".  But I can comment on this:

    As for the Warning header: we did not spend days discussing how to
    internationalise the warning text field, this was just a
    micro-decision made by one of the editors along the way.  Maybe it
    was not an optimal decision, but we did not have the time to spend
    days optimising every micro-decision.

Actually, one of my early drafts for the Warning header proposed
this syntax:

   Warning headers are sent with responses using:

       Warning = "Warning" ":" warn-code SP warn-agent SP warn-text
               [SP language-tag [SP charset]]

       warn-code = 2DIGIT

       warn-agent = ( host [ ":" port ] ) | pseudonym
                       ; the name or pseudonym of the server adding
                       ; the Warning header, for use in debugging

       warn-text = quoted-string

I can't remember exactly why or when the language-tag and charset fields
were removed.  However, I do remember that we had a moderately long
discussion of this issue at the Montreal IETF meeting; it was NOT
removed by an editor's micro-decision.

Someone else may be able to reconstruct the argument that was made
in Montreal, but I believe we basically agreed that the charset
field was unnecessary because of availability of the RFC1522 method.

And I think it may be irrelevant that the spec says "the default language
is English" (section 14.45), since there is no way for a user
to explicitly negotiate what language the Warning will arrive in.
The actual default is whatever the implementor wants it to be.

Presumably, either the user understands the language, or the user
doesn't understand the language, but either way, would a language
tag make it any easier for the browser to display the Warning in
a readable form?

Again, I know very little about i18n; is it actually the case that you
cannot always render the text with knowledge of the character set
alone, but must also know the language as well?

If it is, in fact, not possible to correctly render the Warning
text without a language tag, then perhaps this is a serious enough
bug to warrant changing the HTTP/1.1 spec.  (This is, after all,
what the 6-month delay before Draft Standard is meant to allow!)
Otherwise, what actual problem is left unsolved?

-Jeff

Received on Tuesday, 17 December 1996 23:58:14 UTC