From: Mark Davis ☕ (mark@macchiato.com)
Date: Tue Jan 25 2011 - 17:36:00 CST
This is not different in kind from what has happened with other cases.
Generally the actions are:
Add alternate CMAPs to fonts
Add roundtrip mappings to encoding tables (leaving former PUA characters as
one-ways)
For databases, consider whether to convert old data or not (probably not
necessary in this case)
...
Mark
*— Il meglio è l’inimico del bene —*
On Tue, Jan 25, 2011 at 09:22, Peter Constable <petercon@microsoft.com>wrote:
> Your suggestion entails putting both a symbol-encoding and a
> Unicode-encoding cmap into a font. I have no idea what different software
> would do with that: some might recognize the Unicode cmap, other software
> might pick the one that's listed first in the font, other software might
> just reject the font.
>
>
>
> That's the font issue. I was thinking more about the data. For example I
> can search for the symbol-encoded Windings symbol, but now it would have two
> different representations.
>
>
>
> For legacy encoding of Emoji characters, most of that data is transient in
> nature (SMS messages) or is email sitting on a server belonging to the
> mobile provide who defines that proprietary extension to Shift-JIS. They
> already manage all the data coming into or going out of their environment.
>
>
>
>
>
>
>
> Peter
>
>
>
> -----Original Message-----
> From: Christopher Fynn [mailto:chris.fynn@gmail.com]
> Sent: Tuesday, January 25, 2011 5:13 AM
> To: Unicode List
> Cc: Peter Constable
> Subject: Re: square bullets added to unicode.
>
>
>
> How is the mapping from the old (phone company) Emoji character encoding to
> the new official Unicode encoding handled? Couldn't something similar be
> done?
>
>
>
> A nasty (but probably effective) way of handling it would be to simply map
> the glyphs in an updated Wingdings font to both the PUA codepoint and an
> official codepoint.
>
>
>
> - C
>
>
>
> On 25/01/2011, Peter Constable <petercon@microsoft.com> wrote:
>
>
>
> > Various people have indicated at one time or another that they would
>
> > work on a proposal to encode wingdings, etc. But I haven't seen any
>
> > such proposals submitted yet.
>
> >
>
> > Of course, if an when such symbols get encoded in Unicode, there will
>
> > just be a different problem to sort out: today, you can interchange
>
> > text meaningfully with someone else so long as the Wingdings (or
> Webdings, etc.
>
> > as appropriate) font is used. It doesn't work for plain text, but it
>
> > evidently works reasonably well with rich text (since the font
>
> > formatting is retained). But once in Unicode, data using the new
>
> > character will not be interchangeable with the large amount of legacy
> data out there.
>
> >
>
> >
>
> > Peter
>
>
>
This archive was generated by hypermail 2.1.5 : Tue Jan 25 2011 - 17:40:12 CST