At 04:51 PM 2/17/2004 -0500, Ray Denenberg, Library of Congress wrote:
>To be on the safe side I'll wait until tomorrow morning to announce 1.1.
>I'll post an announcement if I don't hear from anyone. (But I haven't heard
>from enough of you to be comfortable posting tonight.) Attached is a
>revised announcement, very similar to the earlier draft. I'll post about 10
>US Eastern time, so if you have comments on the announcement get them to me
>before then.
Ray,
What you drafted labels SRW/SRU as a "successor" to Z39.50 but
SRW/SRU needs to be positioned as a profile of Z39.50. You may
recall that we discussed this at a meeting at LoC. You asserted
that SRW/SRU can be regarded as a profile of Z39.50 in the same
way as Cliff's RFC 1729 "Using Z39.50-1992 Directly over TCP".
I recommend just two small changes to your text. In paragraph 3,
change "its predecessor protocol" to "Z39.50" and in paragraph 4,
delete "Z39.50" as an example query language. (You may also want
to delete the other three examples as pointlessly provocative.)
I am very keen on pushing uptake of SRW/SRU but I can only do so
if it is labeled as a Z39.50 profile (ideally, it should be an
ISO 23950 profile, but we can defer that for now). My current
work in that respect includes the soon-to-be-publicized Statement
of Requirements for Search Interoperability, available in "sneak
preview" at http://www.search.gov/interop/requirements.html
BTW, your Microsoft e-mail includes characters that do not
translate into e-mail ASCII. You could save the message content
as text (without line breaks other than paragraphs) and then
reinsert into the message body before sending.
Eliot
