Bug#99933: Default charset should be UTF-8

[Raul Miller <moth@debian.org>]

On Tue, Jun 12, 2001 at 12:41:02PM +0200, Radovan Garabik wrote:
> > How can this person tell which kind of boldness is in use?
>
> mathematical symbols could use different typesetting convention
> (see latex)

But this should depend on mathematical context, not code point.

Or are you suggesting that latex shouldn't render ascii characters
using mathematical typesetting conventions?

> > Let's imagine that a person is actually reading this document. What
> > difference does it make to this person that the Unicode Consortium
> > has named the code point using the word MATHEMATICAL? How would the
> > person even find out about this?
>
> decent html browser would render mathematical symbols differently.
> But, of course, it need not, depeding on font used.

Again, this should depend on whether a mathematical context is
in use (e.g.  mathematical equation).  Unless (to use the same
example again) you wish to prohibit the use of ascii characters
in mathematical equations.

> > expression?  If there was surrounding text describing the character of the
> > variables x, y, and z, would you insist on this contrived intepretation
> > of yours?
> 
> I do not insist on it... but as you can see, without a context
> anything can be misinterpreted, and special symbols are just a tad
> helpful.

I'll agree that context is important.  I'll agree that special symbols
are important.

I disagree with the idea that special symbols may only be used in 
certain contexts.  That's like saying that HTML should only be used
to describe the structure of a document and not its appearance --
fine language from a standards body, but with little to do with
how the standard is actually used.

> > If we assume that the user is using debian software which merely
> > displays the characters (and doesn't actually inform the user of
> > the unicode names for the underlying code points), would there be
> > any particular reason for the user to interpret some characters as
> > algebraic variables and others as word forming characters in some
> > unknown programming language (for some reason other than knowledge
> > of the unicode code point numbers)?
>
> this is just nitpicking... unicode is full of characters having the
> same glyphs how do you distinguish between LATIN CAPITAL LETTER A,
> CYRILLIC CAPITAL LETTER A, GREEK CAPITAL LETTER ALPHA ? they look
> the same in upright font, but if you select cursive font to view the
> document, they will look differently. The same with mathematical
> symbols... they might look the same with one font, but if you prefer
> slanted font, you suddenly see a difference... (or vice versa).

Exactly.

Thanks,

-- 
Raul