Re: [idn] 7CE

["Adam M. Costello" <idn.amc+0@nicemice.net.RemoveThisWord>]

Martin Duerst <duerst@w3.org> wrote:

> I think it might help quite a few people who don't really like ACE
> to know that the (key) ACE proponents understand that ACE is a layer
> violation in the (character) encoding stack, and will have its serious
> negative consequences like any other layer violation.

I usually think of a "layer violation" as a case of one layer being too
interdependent on the inner working of another layer, instead of knowing
only the interfaces to the layers immediately above and below it.

ACE merely inserts a new layer, but the layers can still be cleanly
separated.

If the old stack looks like this:

    arrangement of glyphs
    sequence of characters
    sequence of code points
    sequence of octets

then the new stack looks like this:

    arrangement of glyphs
    sequence of characters
    sequence of code points
    sequence of basic code points
    sequence of octets

The old transformations between glyphs and characters, and between
characters and code points, are unaffected.  The transformation between
basic code points and octets is exactly like the old transformation
between code points and octets (except that the repertoire is smaller,
so a more primitive serialization is likely to be used, typically
ASCII).  A new translation between code points and basic code points is
introduced, but it is orthogonal to the transformations at the other
layers.

AMC