Present: B. Alan (HOTS), R. Doherty (CDL), P. French (UCD), C. Graham (UCSD), 
L. Hayman ( UCSB), C. Hixson (UCLA) 

ANNOUNCEMENTS/UPDATES 

Approval of minutes from March 15 is postponed until next meeting. 

Doherty will be out of town after today, Karen Coyle has agreed to serve as our 
CDL liason as needed for the duration of the project. 

POSTING COMMENTS ON THE WEBSITE 

The group discussed posting comments sent to us from other campuses and our 
responses on the TFER web site as a method of communication.  The group agreed 
that since respondents had not been informed their comments would be publicly 
posted, it would not be advisable to do so.  Instead mention will be made in the 
report that we received feedback and have included changes and corrections in the 
final report.  Graham agreed to write to individual institutions in response to 
some specific questions they raised, e.g., 246 vs. 740.  Persons interested in our 
responses to the comments can find the information in the minutes.  

EMAILS FROM DOHERTY 

Doherty invited the group to feel comfortable to add or delete information 
from the documents she has prepared for the task force.  Her intention in creating 
them was to record the decisions we have made about Melvyl processing and record 
distribution.  Additional text may be needed to further clarify some ideas. 
Doherty suggested that we use her documents as technical appendices to the report. 

ACTION: Hayman will summarize in the report and insert references which will 
direct readers to the appendices. 

We have received four substantial documents from Doherty outlining the technical 
aspects of receiving, loading and distributing CDL records.  They are: 

1. "COST VARIABLES FOR CDL-T PROCESSING, MAINTENANCE, AND DISTRIBUTION OF CCA/CDL RECORDS" (email from Doherty, dated March 22) 

Doherty prepared this information to aid Alan in the creation of the Cost Model. 
Although we do not know exact figures yet, Doherty estimates that implementing 
these recommendations will take much less time and effort than the TFER1 display 
recommendations. 

2. "CDL CATALOGING GUIDELINES- LOCAL DATA REQUIREMENTS FOR CDL-T" (second draft, email from Doherty, dated March 23) 

This document outlines the data elements and tagging required for processing by 
the CDL. 

3. "CDL-T PROCESSING AND DISPLAY GUIDELINES" (third draft, email from Doherty, dated March 23) 

This document describes Melvyl normalization and display processing steps to be 
performed by the CDL. 

4. "MORE MISC. ITEMS" (email from Doherty, dated March 23) 
The following topics were mentioned in this email and discussed: 

   A. Use of 9xx fields for CDL Directory vocabulary terms 

Doherty commented that although the future configuration of Melvyl may not 
accommodate 9xx fields, we can use these fields for now so long as we are 
consistent in their use.  A 9xx is favored for this use because a field can be 
uniquely defined for this purpose.  Existing 69x bibliographic fields are all 
designated for other purposes. 

Doherty brought up an idea Laine Farley had about using acquisitions records 
for getting information needed by the CDL Directory, rather than taking it from 
the Melvyl input record.  Graham reported that they had tried this at UCSD for 
another purpose and it was not successful due to the complexity of the 
programming.  Order records would not work for this purpose because there is only 
one order record per subscription service, whereas the catalog and the CDL 
Directory contain individual records for each title in the service.  For example, 
there is one order record for JSTOR but 59 catalog records. 

   B. Indexing of the 776$x 

We will include a recommendation to index the ISSNs in 776$x in the report 
and state the reason this is desirable. 

   C. "AT CDL" limit implementation 

We will include a recommendation that an AT CDL limit be established.  

Hayman observed notes of Doherty's conversation with Farley raised the 
question of whether such an indication might be confusing for users in those 
instances not all campuses had access to the resources, and this was a point the 
task force had not considered. 

There will be an access restrictions note which may alleviate this concern. 

   D. Making the CCA/CDL record the base record 

We will include a recommendation on this in the report. 

ACTION: French will go over the report and add text to reflect the ideas Doherty 
wanted to be sure were mentioned, then share it with the group. 

   E. 856 display 

 Doherty commented that she felt it was important that our report mention 
that the 856 display is being addressed outside of this committee, and that any 
examples given in the report are only possibilities. 

   F. Adding 856 fields to Melvyl Full Text titles 

 Doherty reported that CDL-T is working on other ways to link to full-text 
from the California Periodical Database.  We will not address this in our report. 

DELETING RECORDS, AND OTHER POST CATALOGING ISSUES 

The group discussed the need to establish a mechanism for notifying local 
libraries about deleting CCA records.  Doherty said if the CCA sent a record to 
CDL to be deleted, CDL would be able to mark records as deletes.  It would then be 
up to individual libraries to recognize the coding, and act accordingly. Doherty 
will provide the coding for identifying delete records to the group so it can be 
included in the report. 

The group discussed the need for a 2-way communication mechanism to share 
information about changes to CCA records.  A listserv and/or a web site are some  
possible options. 

The group also discussed campus-specific dropping of CDL licensed titles.  At 
this point, the logical approach seems to be that the CCA will delete 
campus-specific location/access code and notify the campus.  It will be vital that 
a reliable mechanism be established to inform the CCA of licensing changes.  The 
CDL Acquisitions unit can easily provide this information to the CCA. 

Separating CDL records into categories (update, delete, new records) before 
campuses receive them will not be possible through CDL.  

SURVEY RESPONSES FROM CAMPUSES 

The group reviewed the compiled responses from the heads of cataloging.  The 
compiled responses will be posted on the web site. 
 

1.  Would you please confirm that FTP is the preferred method for 
    retrieving files of records at your campus?   If not, what would be 
    the preferred method of receipt? 

All campuses are able to receive files via FTP.  We will add a statement 
about FTP being the means of transmission in the section on record distribution. 
 

2.  Would a separate campus specific file be preferable to retrieving a 
    complete file of CCA-generated CDL records and separating the records 
    at the campus level? 
3.  Would your campus prefer having the files divided by package, e.g. 
    JSTOR, rather than by campus? 

A majority of campuses desire a campus-specific distribution.  Three campuses  
expressed a preference for package, and three more said files tailored by 
campus and composed by package were desired.  UCSF indicated interest in both 
types of division.  Alan will contact UCSF for clarification. 

It would be ideal for the CDL to subdivide the records by package as well as 
by campus.  We will check with CDL on how feasible this would be.  For now, the 
group agreed that the initial distribution of records would be packaged by campus, 
and the CCA would work through the records package by package.  
 

4.  Would weekly or monthly distribution of CCA-generated records be 
    acceptable to your campus? 

The group agreed that weekly distribution would be preferable.   Although the  
records would be available weekly, campuses who would like to receive their 
records monthly could still pick up their records every month since the records 
will be stored on the server for three months. 
 

5.  There is a cost associated with retention of files of records on a 
    server.    How long would your campus need to have the records 
    retained on the FTP site before being deleted?   Would three 
    months or less be acceptable? 

The consensus was that three months is an acceptable period. 
 

6.  Would your campus be able to process records that include extraneous 
    data appearing in MARC 9xx fields? 

All campuses indicated that they could process records with extraneous information 
in the 9xx fields, with the exception of UCB which did not answer the question. 
Hayman will speak to Armanda Barone for an answer. 
 

7.  Would having the CDL-specific subject headings in your local catalog be 
    problematic? 

All campuses were confident that their catalog would not be adversely affected by 
the CDL-specific subject headings, with the exception of UCSF.  UCSF is 
assuming that there will be authority control by the CCA, which will not  
be the case. Alan will contact UCSF to clarify the situation.  The headings could 
be stripped out by CDL if desired.  This will need to be decided once the CCA 
begins adding them to records.  CDL headings will not be added initially pending 
further development and formalization of the CDL vocabulary scheme. 
 

8.  Does your campus consider it critical to include the LC call number in 
    CCA-generated records? 

All campuses except UCLA were amenable to including call numbers in CCA cataloged 
records but not requiring that they be added if they were not in the existing 
record.  The group decided that the best solution is to include call numbers if 
they are present in cataloging copy but not supply them when they are lacking. 
Libraries who want call numbers in all records could add them locally. 
 

9.  Incoming records will include the OCLC control number as the most 
    reliable matching point.    If many of your records do not have OCLC 
    control numbers this may cause matching problems with machine matches. 
    Other matching points, e.g. ISSN, may be less reliable and therefore 
    problematic.   Comments? 

No campus disagreed with the OCLC number as the most reliable matching point, even 
though most said that all local records did not include an OCLC number.  The group 
agreed that the OCLC number is the best option for a matching point. 
 

10.  If records are loaded and overlay existing bibliographic records,  is 
     there local data, e.g. local notes, call numbers, etc., that will 
     need to be preserved and can that data be protected? 
11.  Most of the CCA-generated records will be for print titles that may 
     have been cataloged or converted under earlier cataloging rules and 
     therefore the choice and form of entry will vary in your local 
     catalog.  Comments? 

The campuses who are planning on overlaying records are prepared to address  
these aspects of CCA records in their local systems.  UCSB is planning to 
employ a student assistant to convert CCA records to separate records and load 
those records, then conduct manual review of print records in the local catalog  
as time permits.  All other libraries are planning to do a manual review of CCA 
records.  
 

12.  The CCA-generated records will not include local bibliographic and/or 
     holdings information that will need to added at the campus level. 

All campuses recognized that they would need to add local information or local 
holdings in conjunction with their location code at the campus level. 
 

13. Other considerations? 

UCLA raised the question of how campuses will be notified about changes to CCA/CDL 
records.  Please see preceding section in these minutes- "Deleting Records, and 
other Post-cataloging issues" for discussion regarding this. 

14.  Would your campus require separate catalog records? 
15.  If so, would your campus want the separate records to complement the 
     CCA-generated records or would you prefer only files of separate records? 

UCLA and UCSF expressed interest in having the separate records for materials for 
which they do not have a print subscription.  All other campuses plan to use 
single records.  

The CCA will be creating records using CONSER's single record approach. UCSB will 
be creating separate records as complements to the CCA records.  If the 
campuses do not consider the need for such records to be sufficient to justify 
making distribution part of the centralized distribution method, UCSB will find  
a way to make those records available to UCLA and UCSF.  The group agreed that it 
is not necessary for the task force to formalize distribution and costs for the 
distribution of UCSB's separate records.  UCSB will be doing so, in any case, and 
will share that information with HOTS. 
 
 

French will send out an email to all campuses thanking them for a timely response 
and informing them that they can see the responses from all campuses on the web. 

The next meeting will be Friday, March 26 from 9-11.  We will continue to discuss 
the comments on the Cataloging Guidelines, and the cost model.