"Component" Information associated with URN processing versus to be passed to objects, URLs, etc

[klensin]

There appears to be agreement in the WG that we need some sort of namespace-independent lexical mechanism to distinguish between information that is intended to be part of the URN processing (possibly including resolution) process and information that is intended to be processed by, or at part of, whatever the URN points to (such as an object or URL).

2141bis-11 lays the foundation for making that distinction on a "component" basis, thereby tying it to URN equality and persistence, which may or may not be desirable. There was a mailing list discussion of defining and reserving some keywords in the q-component .a year or so ago, but it led nowhere. A proposal is now under discussion to subdivide the q-component into two parts. using "??" as a delimiter to separate the two. That proposal would pass the f-component through (there have been separate arguments that RFC 3986 will not permit anything lelse). p-component still seems to be up in the air.

Whatever is decided upon, we need to not only document it but recheck euqivalence and persistence text to be sure everything is consistent.