Present: Bob Alan (HOTS), Rebecca Doherty (CDL), Pat French (UCD), Crystal Graham (USCD), Lynne Hayman (UCSB), Carol Hixson (UCLA)

1. MINUTES OF FEBRUARY 19TH
Discussion and approval of last week's minutes has been postponed until next week.

2. ANNOUNCEMENTS AND UPDATE
Re: Transmittal message to SOPAG
French reported that the message has been forwarded to SOPAG liaison, Cynthia
Clark. It has also been posted on the web site, along with the Draft CDL
Cataloging Guidelines and the Preliminary Report and Recommendations.

Re: PURL server
Hayman reported that she had sent an email to Joan Gargano of CDL regarding
the Preliminary Report and Recommendations, and TFER's recommendation for a PURL
server.

Re: Distribution of Survey
The survey that Alan has emailed to the group for review will be distributed to
Heads of Cataloging on March 1.  The group is asking for responses no later than
March 19, so there is time to process the results before April 2, when we will
submit our final report to SOPAG.

Re: Feedback on Guidelines
The group discussed creating a channel for receiving
feedback on the Draft CDL Cataloging Guidelines.  The group agreed to send out
a brief message to Heads of Cataloging, HOTS, and our "buddies" at other UC
libraries, requesting their comment as well as requesting that they forward the
message to other colleagues and staff. In order to include the results of the
feedback in the Report comments need to be received by March 26.

The group also agreed to add a note on the web site requesting comments, directing
respondents to French via email.  French will then distribute the responses to the
group.

ACTION: French will put a request for comment on the web site.

Re: Wording in the Survey
Doherty informed the group that although in the survey it states that campuses
can access files on a weekly basis from an FTP site at CDL/T, the file will remain
at the FTP site for three months.  This brought up the question of what an acceptable
frequency would be for receiving files.

The group also discussed rewording in the section of the survey regarding call
numbers (LC and NLM class numbers).  The need for the CCA to generate records that
do not require customization will be stressed, and the survey will ask if
libraries consider it critical to include the LC call number in each record.

The section of the survey regarding having CDL headings in local catalogs will now
include a question asking if having CDL headings in local catalogs would cause
problems for local systems and what those problems might be.

ACTION: Alan will reword the survey to reflect these changes

3. STATUS REPORTS

Re: NISO Standard
The final version of the NISO Standard for holdings statements will be available
in 6-8 weeks.  According to the NISO publications office there are no substantial
changes from the draft standard we have been working with; the only changes are
editorial.  The standard will cost $48 and, an additional workbook entitled
"Creating Holdings Statements" is available for $20.

This discussion raised the question of how closely we are going to follow the
standard, especially in 856 $3.  The group agreed to put Alan's  summary of the
standard on the web site after the group decides which examples to include.

ACTION: The group will discuss this question either by email or at a future
meeting.

Re: Order of Subfields in the 856
Doherty is still waiting for a definite response from us on the desired display
order of 856 subfields.  We agreed that as long as the $3 is displayed first in
Melvyl Web followed by the $z and $u, the remaining subfields could be displayed
in the order proposed by CDL/T.  Hayman will confirm this with UCSB contacts.

Re: Collection Development Code
Alan reported that Clinton Howard, CDC chair, does not feel that collection
development codes are needed in CDL catalog records.

4. MELVYL PROCESSING FOR DISPLAY

RECORDS FROM THE CCA
Doherty reported that CDL programmers will need to know which 856 field in the CCA
record should be used to create the CDL location/holdings segment.  This is
complicated because records frequently carry more that one 856 for several
reasons.  Some titles have licensed content available from more than one source
(ex: Project Muse and JSTOR) and both may be CDL acquisitions.  It is possible
that there could sometimes be even more than two, when there is content overlap
between digital projects or publishers.  There may also be 856s in a record for
other types of related web sites (organizational homepages, table of contents
sites, sites with important supplementary material such as indexes or abstracts).
Related web sites can be identified by the 856 second indicator "2".

The CDL location/holdings segment would contain the usual 4 elements in the PE
database:
        library-branch-call number-holdings.

These could read:
        CDL Internet [blank]- [856$3 holdings] Access restricted to: UCB, UCD,
                UCI, UCSD

If a location segment were created from more than one 856 field (when more than
one exists), the display could be unclear to users because the project name and
URL are not part of the location segment.

A display for a Project Muse/JSTOR title could be:
        CDL Internet v.1 (1949)-46 (1994) v.47 (1995)- Access restricted to UCB,
                UCD, UCI, UCSD

Users would still need to look at the Electronic Location area in the bib record
(created directly from the 856s) to find the URLSs and to associate the holdings
with the appropriate web site.  Is this useful?

It might also be possible to instruct the program to create one combined
location/holdings segment when more than one 856 field is present.  We would need
to decide how the $3s should be combined together.

For example separate 856s for Project Muse and JSTOR could be concatenated into
one location segment:
        CDL Internet v.1-46 (1949-1994); v.47 (1995)

Alternatively, the #3 holdings area could contain a generic note saying "See
Electronic location" which refers users to the 856 display itself located above.
This approach would work when people are searching directly in the PE file,
because they would have the bibliographic record display containing the Electronic
location area.  But it would be a problem in other databases which have a hooks-
to-holdings link to PE.  The linking program links to and displays only the
locations/holdings segments.  It cannot include the Electronic location (URL)
display because that is only in the bib record.

There are different questions regarding the CDL location segment in the CAT
database.

The usual location segment contains the following elements:
        library-branch-call number-[occasional public note?]

Holdings are not usually part of a CAT location display and the mechanism for
reporting holdings is different in CAT.  Since we will not display a call number,
the basic CDL location segment without holdings could be:
        CDL Internet

Do we also need to display the 856 $3 holdings in CAT as we do in PE?  If so, a
mechanism needs to be developed to support this.  The hooks-to-holdings linking
program does not presently link to CAT holdings so that is not an issue.

No conclusions were reached on these complex questions.  The discussion will be
continued at the next meeting.

Decisions will be needed for all these same questions for CDL holdings reported on
campus tapes.

5. RECORD DISTRIBUTION
Doherty reported that the CDL can distribute CCA records if that is the choice the
group makes.  It would take only a small amount of programming development to
enhance the existing loading program to drop unwanted local fields.  CDL could
separate records by campus without alot of time and effort.  Output of CDL
records could be weekly if needed and the records would remain on the FTP site for
three months.

Graham will report on UCSD record distribution issues via email this week.
 

The next meeting will be on Friday, March 5 from 10-12.  The group will continue
to discuss Melvyl Processing for display.