I support this document. I have a few suggestions: Section 1: Is it worth mentioning that the reference code includes an implementation of the ISO/IEC 9899 C and IEEE 1003.1 POSIX time functions? Section 3: The term "TZ name" should be used instead of "key" in the update criteria, to avoid confusion with the cryptographic key mentioned earlier in the section. Criterion 1 could perhaps be clearer. Do I understand correctly that the intent is something like "New keys are only to be created when it is found that the region a TZ name was envisioned to cover does not all follow the same set of timezone rules between 1970 and the present day." That is, the thing that must be "accurately reflected" is the scope of the TZ rule, not its name or anything else. Is it worth documenting the naming scheme here, or would that be unhelpful? Typo: "policy policy". Section 5: Should the document explicitly mention that rough consensus of the TZ list should also be used when considering an appeal to the IESG? Section 6 and 8: The term "identity" appears whereas section 3 used "well known key". Is "identity" supposed to imply something more complicated than a key, such as an x.509 certificate? Tony. -- f.anthony.n.finch <dot@xxxxxxxx> http://dotat.at/ Viking, North Utsire: Northwest backing southwest 4 or 5, occasionally 6. Moderate or rough. Mainly fair. Good. _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf