> Why won't they just be z39.50 diagnostics? We've got no interoperability
> otherwise. If we need to add new ones to z39.50, then that's fine. But
> let's not just start over with a new set.
1.We're not planning to accompany an srw diagnostic with an oid, are we? So
would SRW simply assume bib-1? Is that a good idea?
2. Supposed we need a new diagnostic that doesn't make sense for Z39.50 (e.g.
invalid session id, session id expired, result set ttl expired, etc). We
normally seek ZIG approval before adding a diagnostic to bib-1. Ususally
there's no objection, or even discussion. On occasion though a diagnostic has
been rejected. But at least, the diagnostics proposed usually make sense.
What's the ZIG going to say when we propose one that doesn't?