Mapping (was: Issues lists and the "preprocessing" topic)

[Andrew Sullivan]

On Tue, Aug 19, 2008 at 02:31:22PM -0400, John C Klensin wrote:

> A-label form.  I've got a device here whose "keyboard" contains
> specific keys for "www." and for ".com" and several other TLDs.
> My using those keys is a UI convention: I press the single keys,
> but the four-letter strings go into the file (or browser
> location bar, or whatever).  These provisions for local mapping
> are not really different from those specialized multi-character
> keys if the conversions are done immediately.

I don't find this example especially compelling, because it seems to
be an example of a device doing special preparation before doing
anything impinging on what I'd think of as "protocol space".  That is,
it's taking your special key and using that to prepare certain
combinations of characters that are often used in labels.  If that's
all the local-mapping step is supposed to be, why can't we just be
completely silent about it?  It's a user interface problem, and
completely outside of the protocol.  If people want to encode their
stuff in Revised Augmented EBCDIC, and write some preprocessor to
change it all before it gets to the point where we are talking about a
label (of any variety) in an IDNA-aware application, what business is
it of ours?

A

-- 
Andrew Sullivan
ajs at commandprompt.com
+1 503 667 4564 x104
http://www.commandprompt.com/