Another Diagnostic- or Variant-Set?

Question from: Mike Taylor mirk@system-simulation.co.uk Thu, 7 Aug 1997 18:25:47 +0100

Do we foresee another operative diagnostic set besides the (gradually growing) bib-1 set? Or another variant set besides Variant-1?

A few years ago it was thought that there would be different diagnostic sets for different applications, or for different groupings of diagnostics, but that seems no longer to be the current thinking. Recently several Init diagnostics were added to bib-1; at that time it was suggested that we assign a dedicated diagnostic set, but nobody wanted to; everyone wanted those diagnostics to be put into bib-1.

There is a diag-1, which is a diagnostic format, not a set (the distinction is described in the error diagnostic appendix), and that was developed with the idea that the server would be returning complex, structured diagnostic information, where each diagnostic definition would be tailored to the information. We spent much time developing diag-1, and it seems that nobody has used it.

What may happen is that we would define a bib-2 set (or a more generic-named set) allowing us to completely reorganize the diagnostics and ultimately deprecate the bib-1 set. But that won't happen anytime soon.

For variants, that's much harder to predict. The direction for diagnostics is easier because diagnostics are an integral part of Z39.50, and we have much more experience with them than with variant. Until we have some practical implementation experience with variants, it's difficult to speculate on their direction.