[precis] Category changes with Unicode 6.3

"Martin J. Dürst" duerst at it.aoyama.ac.jp
Thu Oct 17 00:04:03 CEST 2013


On 2013/10/16 22:49, Patrik Fältström wrote:
> On 16 okt 2013, at 16:46, John C Klensin<klensin at jck.com>  wrote:
>
>> This is great.   To a considerable extent, it reinforces the
>> utility and importance of the IDNA2008 approach.
>
> Absolutely!
>
> Having the original IDNA locked with one version of Unicode is so weird I ask myself quite often how I could come up with that idea :-P

I think there are simple explanations, but they are non-technical.

When we worked on IDNA2003, there were so many people asking for so many 
things (many of them rather impossible, dangerous, or otherwise of 
doubtful value, but never mind), that any idea that was able to restrict 
the solution space looked like a good idea.

Also, there seems to be a psychological value in the newest version. In 
economic terms, no difference between 10 cents and 5 cents, and 5 cents 
and free, but because of psychological reasons, people behave quite 
differently once something is free. Likewise, with versions, even though 
the differences from version 3.1 to version 3.2 and from version 3.2 to 
3.3 or 4.0 may be of the same nature or magnitude, if version 3.2 is the 
newest, that's seen as more stable and permanent that the others.

Thinking about many standardization efforts I have seen over time, there 
also seems to be a tendency to ignore versioning in the first version, 
because it's seen as the only one, and people are really concerned with 
getting the basics right. Occasionally, one sees a lot of effort being 
put into versioning and extensibility, but it usually doesn't work out 
exactly as intended.

So in summary, versioning is hard, and maybe it's easier to get right in 
the second version than in the first.

Regards,   Martin.


More information about the Idna-update mailing list