Re: cross manual references in html manuals, second

[Karl Berry]

    If you look closely at my proposal, the files used only [a-z], as somebody
    said that on windows the case didn't matter. 

It is true that case is "preserved" (not significant) on Windows.
However, that is no reason why we should not generate file/target names
using uppercase letters.  I think we should, just because it's closer to
the actual name of the code.

    >     @today today

Let's just disallow @today in node names.  It doesn't make sense either way.

    Argh. This is quite annoying, as we then have to use an encoding of
    special chars like _xxxx (if I'm not wrong about the number of
    hexadecimal characters used to describe an utf-8 character)

Yes, four hex digits for unicode.

I know there are people who would like to use Texinfo with both Latin 1
and Latin 2, at least.  It seems we have no choice but to go to utf-8.
Argh, indeed.

    @node a @equiv b 
    but rather write
    @node a == b

Well, until we support @ commands in node names, it's a moot point :).
But when we do, then there would be no harm in using @equiv{}.

    by the way don't hesitate to correct me, I am not a
    native english speaker

Not a problem, your English is a million times better than my French :).
I'll rewrite all the English when I include this in one of the Texinfo
manuals anyway :).  What's important now is that we agree on the
procedure.

Thanks,
karl