WG Review: Internationalized Domain Names in Applications (Revised) (idnabis)

Kenneth Whistler kenw at sybase.com
Tue Apr 8 23:59:37 CEST 2008

> Could someone remind me why IDNA2003 couldn't be updated in a minimal
> fashion to reference Unicode 5.1 instead of 3.2?  It's not like it is a
> huge amount of work for every new Unicode version.

Unicode 4.0.0 (April, 2003) added 1226 new characters.

Unicode 4.1.0 (May, 2005)   added 1273 new characters.

Unicode 5.0.0 (July, 2006)  added 1369 new characters.

Unicode 5.1.0 (April, 2008) added 1624 new characters.

And there are some thousands more in various stages in ISO
ballotting for ISO/IEC 10646, which will also have to be
added to Unicode in the next couple of years, to keep it
in synch.

And hidden underneath that were a small but significant
number of character property assignment changes which could (or
could not, depending on your point of view) have a bearing
on decisions about which characters are appropriate for
IDN labels.

> Remember, the IDNA WG in 2003 chose consciously to reference a
> particular Unicode version, for reasons that still are valid.  It didn't
> happen by accident.

No doubt. But the IETF could not (or at least did not)
provide an update for Unicode 4.0 or any of the other
releases since then, so IDNA2003 has stayed stuck at
Unicode 3.2, despite very significant additions to Unicode
since then. I have trouble envisioning how the IETF gets
unstuck for maintaining IDNA, unless it disconnects that from
having to review table contents for thousands of new characters
and property stability for tens of thousands of existing
characters each time it tries to update. Hence the need
for building in Unicode version independence moving forward.


More information about the Idna-update mailing list