RE: [uddi-spec] Basic Profile Version 1.0 - Working Group Draft

[Matthew Dovey <matthew.dovey@las.ox.ac.uk>]

> I understand.  But, is there any implementation problem in 
> adding UTF-16 as mandated?  I presume adding UTF-16 is right 
> thing to do for consistency, if there is no real world impact.

The real world impacts for UDDI are:

i) existing implementations which need to be changed (this impact
decreases as we go up from v2 to v3 and vanishes when we hit v4 ;-)
hence the debate on which version we introduce it, *IF* we do)

ii) test scenarios will potential quadruple if we need to test all
combinations - but since UTF-8/16 support should depend on the XML
processor and the actual Unicode character set doesn't change regardless
of the encoding (i.e. if the database engine can't support Korean
characters it will break when it gets one regardless of UTF-8/UTF-16
encoding issues), it is debatable whether this testing it completely
necessary for UDDI *if* you trust or have already tested your XML
processor.

Matthew Dovey
Oxford University