Hi, the 2606bis draft proposes to modify the following three paragraphs among others:
=== old === The ".localhost" TLD has traditionally been statically defined in host DNS implementations as having an A record pointing to the loop back IP address and is reserved for such use. Any other use would conflict with widely deployed code which assumes this use. === new === The ".localhost" TLD has traditionally been statically defined in host DNS implementations as having an address record pointing to the loop back IP address and is reserved for such use. Any other use would conflict with widely deployed code which assumes this use. See [RFC1122] for IPv4 and [RFC4291] for IPv6 loop back addresses. === end === Is that as it should be ? Another modifications affects the use of the various test pseudo-TLDs (.test + 11 new IDN test TLDs): === old === ".test" is recommended for use in testing of current or new DNS related code. === new === ".test" is RECOMMENDED for use in testing of current or new DNS related code. Applications SHOULD treat ".test" like any other TLD; a special handling could defeat the purpose of a test. === end === The interesting bit is the SHOULD, the RECOMMENDED is only a side-effect of using RFC 2119 key words. For .invalid it's a MAY: === old === ".invalid" is intended for use in online construction of domain names that are sure to be invalid and which it is obvious at a glance are invalid. === new === ".invalid" is intended for use in online construction of domain names that are sure to be invalid, and for which it is obvious at a glance that they are invalid. Applications MAY treat ".invalid" as what the name says. === end === See <http://tools.ietf.org/html/draft-ellermann-idnabis-tests-tlds> for the added IDN test TLDs, and extended security considerations. Frank _______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
