On Thu, Jul 6, 2017 at 5:17 AM, Jim Hague <[email protected]> wrote: > > All the keys in those maps (and in every map, as far as I can tell) are > > strings, for which "unsigned" is a meaningless concept. > > No. All keys are unsigned ints, with values specified in the CDDL. We > should > make this more explicit in the text. > > String keys would bloat the output file size enormously. >
Oh, wow. Yes, more explicit documentation in the text would be appreciated. I completely missed that, even in the CDDL where the key definitions follow maps referencing them. It seemed like you were using string keys and relying on general compression to eliminate the bloat, but this greatly clarifies your "designate as implementation-specific all key values above a threshold" statement regarding extension fields (and in that case, might I recommend reserving all nonnegative integers and perhaps also all strings lacking an implementation-extension prefix for standard key values? That would suggest negative numbers and prefixed strings for extension keys, a set encompassing one full major type and then some).
_______________________________________________ DNSOP mailing list [email protected] https://www.ietf.org/mailman/listinfo/dnsop
