On Mon, Nov 26, 2018 at 8:58 PM Paul Hoffman <paul.hoff...@icann.org> wrote:
> On Nov 26, 2018, at 5:44 PM, Warren Kumari <war...@kumari.net> wrote: > > Actually, there is - the tcpdump man pages (and actually all of their > documentation!) lives in github - the version that this document references > is: > > > https://raw.githubusercontent.com/the-tcpdump-group/tcpdump-htdocs/7785b0d834e1f77a1d2ec56c96f51f4bf3bf3de2/manpages/pcap-filter.7.txt > > Note that this URL is basically useless in an RFC that uses the current > format: it is 135 characters long. It's cute that it points to the correct > man page, but really not useful as a reference in an RFC as they are > currently constituted. > > Going back to Warren's original concern: > > > We discussed this on the telechat, and I took the action to try look > into this. > > One of the concerns with a normative reference to the webpage is what > happens if it is updated to add a new primitive - is it allowed? If someone > implements this on Thursday, can they still claim conformance if a new > primitive is added on Friday? > > For the CollectionParameters map, a "filters" item lists the filters that > are used for this capture file. It truly doesn't matter if the reference is > updated later: the capture still has the same filter string. > > Thus, it is quite sufficient for this reference to be: > http://www.tcpdump.org/manpages/pcap-filter.7.html > That is a useful reference for an implementer, and it fits in an RFC. > .... that is where we started. The concern was what happens if there are new filters added, and implementations written don't know how to deal with them. W > > --Paul Hoffman -- I don't think the execution is relevant when it was obviously a bad idea in the first place. This is like putting rabid weasels in your pants, and later expressing regret at having chosen those particular rabid weasels and that pair of pants. ---maf
_______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
