On Sat, Jan 28, 2023 at 05:26:24PM -0500, John C Klensin wrote:
Thanks for the clear and detailed exposition of the status quo.
Just one nit:
> (ii) Special character interpretations given by IDNA2003 but
> removed by IDNA2008, notably including the mapping of Eszett
> (Sharp S, U+00DF) to "ss" and the treatment of Dotless I
> (U+0131) remained as given in IDNA2003. See Section 1.3.2 of
> version 15.0.0 of UTS#46.
This is if "transitional" mappings are enabled. Applications should no
longer enable transitional mappings, and I expect most don't.
$ dig -t mx +noall +qu +ans +noidnout straße.de
;xn--strae-oqa.de. IN MX
xn--strae-oqa.de. 79 IN MX 5 smtpin.rzone.de.
$ posttls-finger -Lsummary -c straße.de
posttls-finger: straße.de asciified to xn--strae-oqa.de
posttls-finger: Untrusted TLS connection established
to smtpin.rzone.de[2a01:238:20a:202:50f0::1097]:25:
TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
key-exchange X25519
server-signature RSA-PSS (2048 bits)
server-digest SHA256
In other respects, the situation is, best I know, as you described.
You're also quite right application developers are typically not expert
in the fine-details, and this is where library defaults matter. Postfix
flags when initialising LibICU are just:
- UIDNA_NONTRANSITIONAL_TO_ASCII
Additional possibilities are:
https://unicode-org.github.io/icu-docs/apidoc/dev/icu4c/uidna_8h.html#a726ca809ffd3d67ab4b8476646f26635aa1eb63014cdaf41c7ea6cf3abecf1169
- UIDNA_ALLOW_UNASSIGNED
- UIDNA_USE_STD3_RULES
- UIDNA_CHECK_BIDI
- UIDNA_CHECK_CONTEXTJ
- UIDNA_CHECK_CONTEXTO
and perhaps we should also be setting some of the last three...
--
Viktor.
_______________________________________________
Uta mailing list
Uta@ietf.org
https://www.ietf.org/mailman/listinfo/uta
