Re: Accept-Charset support

Erik van der Poel wrote:
> > What I was trying to convey was that we had talked ourselves round to
> > agreeing that Accept-Charset was essential otherwise how does the server
> > know what to send?
> Well, I see your point. I will neither agree nor disagree at this point.
> We are currently discussing Accept-Charset here at Netscape. Some of our
> team members are also reading this email. Hopefully, some of them will
> respond.
> In response to other mail on this list, I am completely unconvinced that
> UI is needed for Accept-Charset. I am also concerned about "q". For
> Unicode, it might mean giving a low "q" since most environments cannot
> support *all* of Unicode. Also, the bit mask idea was just that: an
> idea. I'm not religiously bound to it.

It might not be needed for NSN, but I am not knowledgeable enough about
all platforms you support. If you can't be 100% sure about code pages
you should leave a way for user to define this. Environment variable on
Unix is good enough. Or something to put in .Xdefaults.

As for q value... Xterm can switch fonts, but there can only be one at a
time (unless you use very dirty tricks which will disable something else).
For a browser running on xterm with accessable fonts in several code
pages, it would be quite desirable to send low q value for UTF-8,
and leave it out (defaults to 1.0) for all other listed code pages.

Life is a sexually transmitted disease.