>> My suggestion does not violate cp1252, it only affects a Unicode codepoint >
that is never otherwise used inside a font, and the change does not affect >
anything outside the font. It has *no* bad side effects. The only user that >
would even realise that anything has changed is the one who understands >
Windows machinations with codepages and fonts well enough to recognise that >
something actually *worked* with the font when they expected it to be > broken,
e.g. they can get the euro glyph using 0x80 on Win95 even without > the eurofix
patch installed. I don't think there's anything "yucky" about > that.
>What's yucky about that is that they will start producing Unicode documents
with 0x80 as the euro glyph and spreading them around.
Your response doesn't make any sense to me. I am not talking about documents
that are Unicode-encoded; I'm talking about documents that are cp1252-encoded.
Codepage 1252, as of about one year now, defines 0x80 as the euro symbol. If you
think that's yucky, you're certainly welcome to take it up with MS, of course.
The problem that I'm trying to address is that cp1252, prior to that, mapped
0x80 to 0x0080, which in most fonts would give you the notdef glyph (e.g. a
square box). Users are using a variety of different software. If you're
developing a font with a euro and want it to work when a user enters 0x80, but
don't want to worry about what version of Windows they have, I'm suggesting a
solution which I think beats serious hacks to both cp1252 and Unicode by a long
shot.
My suggestion involved font developers mapping the euro glyph to x0080 in the
cmap, *internally to a font*; this does not mean that they can use codepoints in
cp1252 other than according to definition, and I am not suggesting that font
developers advertise any capability other than they otherwise would. Granted, my
suggestion does mean that if I have an app that stores text in terms of Unicode,
such as Word 97, then you can enter 0x0080 into the document and see a euro,
which would be contrary to the standard. But I'm not suggesting that anybody do
that. If a font developer has the glyph for euro accessible in the cmap from
both x0080 and x20AC, they definitely shouldn't advertise that to their users. I
would not ever expect users to put 0x0080 into a Word 97 doc in order to get the
euro glyph. Where would they get such an idea? My suggestion is not aimed at
users; it's aimed entirely and only at font developers in order to deal with
some issues of Windows internals.
What I find most confusing in your responses to my suggestion is that you did
not react in a tenfold manner to the alternative I was responding to. By that
suggestion, users could get a euro symbol not only using 0x20AC but also using
0x2021, and 0x00A4, both of these being characters that may otherwise be used by
users. To me, that's far worse than being able to get a euro symbol using 0x20AC
and 0x0080, particularly when the latter is not being advertised, and is not
otherwise used by users.
Peter
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:46 EDT