Gary Wade wrote:
>
> I agree, the PUA should be up to the application developer as to what is
> in there.
>
> Now, for a PRACTICAL discussion about the means of determining what
> those characters are, there SHOULD be added (if it isn't already there,
> that is) some means to tag those characters in a particular file as
> being "pua/microsoft" or "pua/apple" or "pua/DoofusCompanyInc" characters.
>
> This would allow applications that want to operate upon Unicode-based
> files to correctly interpret and use that PUA area as originally
> designed by the developer (sort of like an application developer's code
> page, as it were). Of course, versioning information probably should be
> added, too, so you might have tags like "pua/microsoft/1.0" or
> "pua/apple/1.2".
>
Oh my, here we go again.
Who will be the registration authority for these names? Oh, right, they
can just be MIME types, since anything and everything can be a MIME type.
I hope there will be millions of them. This way, every program that wants to
"interoperate" with other programs will have to understand every single PUA
registration, and include gigabytes of conversion code.
The other choice would be to insist on using only vendor-neutral
international standards for data interchange. But nobody wants to take the
trouble of creating a standard for glyph registration, so instead Let The
Million Flowers Bloom, and call each one a standard.
I realize there has been some talk here of the demise of the AFII (of which I
know nothing) but maybe a little thought up front, and perhaps reconsideration
of an AFII-like body, would save untold amounts of grief down the road.
- Frank
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:59 EDT