Document: draft-ietf-geopriv-dhcp-civil-04-halpern.txt From: Joel M. Halpern Date: 13 januari 2005 A General Area review during last call. This document appears to me to be almost ready for publication as a proposed standard. However, I have found some minor items that should be addressed when the document is revised after last call completes. Minor comments: The country code occurs before the "what" value in the DHCP options. Is there any reason for this placement? All other CA information is in type-length-value tuples. I think this is because the country code qualifies the meaning of the CA types. The reason this struck me is that it lead me to wonder about the corner case of a wireless access near a country border, where possibly the DHCP server, the network equipment, and the subscriber could be in three different countries. This can be dealt with by filling in the country code after selecting the "what", even though they appear in the other order in the message. Personally, I would simply swap the two fields around. The Element Format (3.3) description is slightly confusing about casing. The CAlength field indicates that it is the length of the CAvalue. The description of the CAlength then includes the sentence "Data SHOULD be encoded in mixed case, following customary selling." My guess is that this sentence belongs somewhere else. Also, there is confusion because the CAvalue description includes text that say "and written in uppercase letters where applicable." So is it mixed case or upper case? The CAValue text also has a description of the format of the "script indication". This presumably really belongs in the description of whichever CAtype is used to carry a script indication (CAtype=128) Shouldn't the last sentence of the Securiyt Considerations (section 5) say more than "one should consider the potential security risks." Is this intended to refer only to the risks in the previous three paragraphs, or some undefined set of risks? Is this supposed to drive some stronger actions than those mentioned in the previous three paragraphs? For CAtypes 1-6, where should the mappings to nations be documented? 5 specific cases are listed in the document. Is the expectation that the registry will contain multiple entries for CAtypes 1-6, each entry providing the definition of a given country? If so, the description in the IANA considerations seems to indicate that an RFC will be needed to register the interpretation for a (set of) country (or countries). Is there some other way we will document the meanings for other countries? Whatever is intended ought to be called out explicitly rather than leaving it implicit in the document.