Re: Poll: consensus to change the encoded-character extension

[Ned Freed <ned.freed@xxxxxxxxxxx>]

> On Sun, 2007-04-15 at 13:17 -0700, Ned Freed wrote:
> > I'm fine with this. I do note, however, that error handling here is kinda
> > tricky - since you cannot be sure until you reach the } that the syntax
> > rules are met, you have to flag any errors you see in unicode values and
> > only report them at the end. Perhaps a note about this would be in order...

> yes, I guess it's easy to trip over this, causing a potential
> interoperability issue.  suggested text:

>         It is an error for a script to use a hexadecimal value that
>         isn't in either the range 0 to D7FF or the range E000 to
>         10FFFF.[...]

>         Note: Implementations MUST NOT raise an error for an out of
>         range Unicode value unless the sequence containing it is
>         well-formed according to the grammar.

Nice wording! I think thie definitely belongs in the specification.

				Ned