In draft-iab-identifier-comparison-00, Dave Thaler describes the same
"buckets" from draft-ietf-precis-problem-statement. It would be good to
have common terminology here. Dave's terms "Absolute", "Definite", and
"Indefinite" seem fine to me (and preferable to "Type 1", "Type 2", and
"Type 3" currently in draft-ietf-precis-problem-statement).
The PRECIS draft also says:
A subclass of case (3) is one in which, within some constrained
population, the comparison rules are clear even though such rules are
not universally applicable. So, for instance, users of US-ASCII may
all agree on a comparison function, but the set of US-ASCII users and
Turkish users may not all agree about the same comparison function.
For the purposes of the present work, it is not plain whether this
subclass case is relevant, so categorization will include it.
It's not clear to me if we need this "Type 3a" bucket. Dave's draft says:
o Indefinite: identifiers that have no single comparison algorithm
on which all parties agree. For example, human names are in this
class. Everyone might want the comparison to be tailored for
their locale, for some definition of locale. In some cases, there
may be limited subsets of parties that might be able to agree
(e.g., US-ASCII users might all agree on a comparison algorithm
whereas US-ASCII and Turkish users may not), but identifiers often
tend to leak out of such limited environments.
Here again, a common understanding would be beneficial. My first
impression is that we can probably do away with the "Type 3a" bucket,
although we might want to look at different examples of such "subsets of
parties" to see whether the comparison algorithms they use really will
leak out in all cases.
Peter
--
Peter Saint-Andre
https://stpeter.im/
_______________________________________________
precis mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/precis
