Re: [idn] An open letter to the IDN WG (long)

["Eric A. Hall" <ehall@ehsco.com>]

John C Klensin wrote:

> these may differ even between countries or regions in which
> the same language is being used.  And people are very attached
> to their languages as keynotes for their cultures -- some
> group of engineers are not going to successfully change
> language-use rules worldwide, no matter how convenient that
> would be for computer systems.

Isn't this the case for directories as well? Would a structured directory
solve this problem (whatever it is) better than a structured lookup? I
mean, by this argument the end-goal is unstructured, free-form meta-data
about a person, place or entity. And even then we would need structure in
order to make sense of the data (hair: brown, eyes: brown, area: X-Y).

Stepping back a bit, I think historically that we have seen people will
abide by rules and mechs which (A) address their needs in the networking
space, and (B) which are workable towards the objective of A. To wit,
nobody likes the current DNS naming system but everybody uses it anyway
because it facilitates networking and connectivity. As long as whatever
solution emerges actually works to facilitate better networking among the
Internet's users then they will adopt it. They will complain (no matter
what) but they will use it as long as it addresses some pressing need and
it works well enough.

Cumulatively, to me this means that DNS and some directory approach are
equally problematic and viable. We can argue as to whether one of them has
more long-term functionality and promise than the other (I am in the
directory camp on that one) but for solving some given problem I don't
think directories are "the default" better choice.

Furthermore, I'm just not convinced that DNS can't be made to work.
However, I am convinced that people will not embrace directories until
there is significant value from doing so (apps that use it *AND* data
stored in it).

Of course, the latter is the tricky part. Not only is it chicken-and-egg
problem, but it also opens up a HUGE political can. People who see it as
being DNSv2 will jump on it with all their mass, and it will become
immobilized. There are some ways around this problem (I have a couple of
ideas but I don't know where to take them), but just outright designating
directories as "DNSv2" would be fatal to the effort.

For all of these reasons, I think it is better to extend DNS as best as
possible. We can build up the directory architecture in the background.
Once there is enough value in the directory people should be able to
switch over to it without changing anything (or at the most will change
some workflow habits). In the meantime, DNS should be extended.

Overview:

   current DNS (no international capabilities, no directory functions)
   internationalized DNS (but still no directory functions)
   directory (with features)

The people that want the features will move to the directory model when it
is operational and has valuable services associated with it.

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/