Re: Enhancements for HTML 2.1

Gavin Nicol (gtn@ebt.com)
Mon, 20 Mar 1995 12:08:20 -0500

>> As such, the infrastructure is largely in place, but the real
>> remaining problem is making sure that non-Latin1 encoded HTML is also
>> legal according to the SGML declaration etc. that HTML uses.
>
>This would seem on the surface to be a straightforward extension of the 8-bit
>DTD (cf. the ERCS proposals).

To a large degree that is true. For your browser, this is certainly
the case.

>> A question: was it really extremely difficult to implement Unicode
>> support, as many naysayers claim?
>
>No. Unicode support is in fact *much* cleaner than support of the status
>quo (EUC, KOI8, ISO 8859/x, etc., often all labeled as ISO 8859/1). On
>platforms which do not support Unicode directly, it is necessary to create
>tables to map Unicode to local encodings for display, but these tables
>are available via FTP from unicode.org.

Thank you. Having worked with Unicode, I share the same opinion.
Does your client also handle things like EUC and whatnot?
In the next release of my "paper" (considerably larger now!), I have a
fair amount of text devoted to how one can implement truly
multilingual browsers.

>I am strongly in favor of Unicode support in both HTML and MIME precisely
>because it is easier to implement than other multilingual encodings, and
>gives better results when combined with language tags.

And it is certainly easier to support a single "core" encoding than
supporting a myriad. I discuss this in my paper too.

If you have any comments on my earlier proposals, I'd very much like
to hear them (privately might be best, so we can avoid cluttering up
the list).