ISSUE: CHARSET DEFAULT

There is one more issue I would like to be made an ISSUE.

It is the second part of Section 3.7.1, dealing with Text Defaults.

The wording in this part was arived at after careful deliberation,
and I don't want to rehash these deliberations.

However, part of these deliberations was the fact that "some older
HTTP/1.0 clients did not deal properly with an explicit charset
parameter" (last paragraph of 3.7.1). The number of such "older
clients" certainly has decereased since this was written. A adjustment
of the text may therefore be in order.

Also, the text says "Some HTTP/1.0 software has interpreted a
Content-Type header without charset parameter incorrectly to
mean "recipient should guess"". While this indeed may be
"incorrect" according to some wishful specification, it is
not very much up to actual web reality. If current browsers
didn't provide me with an option to guess the encoding, I
would not be able to view more than 99% of Japanese (or you
name it) web pages. In addition to this, HTML (both RFC 2070
and HTML 4.0) defines on its own the priorities for "charset"
information; the same applies to XML (with different priorities).
It would be desirable if the HTTP spec was compatible with
these specs.

If desired, I can draft some new wording.

Regrads,	Martin.

Received on Monday, 28 July 1997 11:05:31 UTC