Re: Z39.50 character encoding

Mark Reichert wrote:

> Doesn't character set negotiation currently apply to all
> InternationalStrings sent over Z39.50?

Most.

A list of  parameters of type InternationalString, within Z39.50 and associated
definitions, designating them as either name or message strings (a "name" string
is, for example, a result set; a message string is, for example, a diagnostic),
is at:
http://lcweb.loc.gov/z3950/agency/defns/namestr.html
The character set negotiation records points to this page.


> Both SUTRS and GRS make use of
> InternationalStrings.
>

So theoretically, yes, by roundabout (and perhaps convoluted)  inference,
character set negotiation applies to sutrs, and to a lesser extent, grs.

Z39.50 character set negotiation history dates back to about 1994 or 1995, when
character sets were a much bigger problem than today as unicode is more readily
accepted.  We could not then come to agreement about how to negotiate character
sets for records, only for name and message strings.

You'll note (on the above-referenced page) that SutrsRecord, the main parameter
in SUTRS, is designated a "message" string.  (The thinking back then was that
SUTRS was for carrying messages.)  So this means that whatever is negotiated for
message strings applies to sutrs.

grs is worse, but bottom line, character set negotiation applies here too (you
need to read the notes on the above-referenced page).

--Ray

Received on Tuesday, 26 February 2002 12:05:24 UTC