Re: [idn] Dots, and a path to working IDNs

["James Seng/Personal" <James@Seng.cc>]

If this is the vote for the strawpoll, it is too vague to move forward
with. UTF-8 on the wire ... how? tag? untag? negiotated? retrofit into
the packet? what needs to be change? what will break? what wouldnt? what
difficulties would we face.

We could move forward with a strawpoll if only you have a concrete
proposal (or one which you can point to). Currently, all we have is what
you have on your website and honestly, I think it is not a complete
picture nor a detail description of who needs to do what and what have
to be taken care of. And given our discussion, it is apparently your
Fast-Nameprep (on Keyboard) idea is far too different compared to some
other proposal we have.

In order to move this forward, could I ask you document your idea into
an I-D first? Thanks!

ps: If I were to vote now, I would vote for UTF-64 (if there is one). My
guess is 10-20 years later, we going to see 64bits hardware, with 64bits
OS computing in 64bits wide characters. Thus UTF-64 makes more sense for
a long term solution. We wont want our future engineers to scream at us
for limiting to them only 8-bits when they have 64-bits to play with.

-James Seng

> ``The long-term IDN solution will encode Unicode characters as UTF-8
on
> the wire.'' I suspect that we have consensus on this point, even
though
> we don't have consensus on the short-term approach.