[dns-operations] xn--

[Andrew Sullivan]

On Thu, Jun 26, 2025 at 08:35:15PM -0500, Gihan Dias wrote:
>
>This clears it up. (though it seems to have been a heavy process for a 
>small decision).

The original IDNA approach with a different (not "Punycode") ASCII-Compatible Encoding (ACE) was well along in the standards process when someone asserted a patent on the algorithm and very nearly derailed the entire effort.  Also, some registries had been selling IDNA-style domain names (though not using the eventual protocol) for some time, and some prefixes had already been used for those cases (hence the disqualified ones).  There was apparently some concern that someone would (if they knew the prefix) just register large numbers of combinations of [[prefix]--[ldh] thereby taking many desirable names out of circulation.  Remember, this is in the 2002-3 period, and speculation in domain names in that period was somewhat less informed by data (like "Domain tasting" and so on) as it came to be.  People were often just betting, and it still felt as though, if you got one really big hit, the speculation would be worth it.  In that context, the elaborate procedures to ensure the appearance that nobody could game anything were probably worth the PR benefit associated.

Best regards,

A

-- 
Andrew Sullivan
ajs at anvilwalrusden.com