Present: Bob Alan (HOTS); Pat French (UCD); Crystal Graham (UCSD); Lynne 
Hayman (UCSB); Rebecca Doherty (CDL); Rebecca Utz (recorder) 

1. APPROVAL OF MINUTES 12/18/98 and 12/23/98. 
The Minutes from 12/18 have been approved.  The minutes from 12/23 have 
been approved, pending one modification. 
 

2. BRIEF REVIEW OF "REPORT OUTLINE (IN PROGRESS)". 

Pat French has added the disclaimer to the Website.  She is also adding 
decisions as we make them, even if they are tentative, with the 
understanding that they will be revised. 

One of the specific application points on the outline which has caused 
some concern is "If a database or website contains the content of 10 or 
fewer print sources, add the electronic access fields to each of the 
related print records."  There needs to be a further distinction made 
that this statement is meant to apply only to databases/websites which are 
also separately cataloged.  Input is also needed from SOPAG and Public 
Services on the extent of analysis desired and the level of funding needed 
to support this analysis. 

ACTION: French will amend this statement in the outline.  FRENCH 

This led to a larger discussion of whether there is a need to 
differentiate aggregretor databases, publisher's home 
pages and "combination" databases for the purposes of cataloging. 

A publisher's homepage is a site that functions as a gateway to journals 
to which the user has a subscription.  In this situation, we might include 
the URL of that home page, but not catalog it separately (even though it 
contains more than 10 titles) because it does not have a chief source or a 
meaningful title in its own right.   A "combination" database (ex: WEB of 
Science) brings together journals from different publishers and has a 
unique recognizable title/identity of its own.  Aggregretor databases 
function as an index to contents but do not have a chief source  title 
page.  There can also be differences of scale involved (i.e.,  number 
of journal titles contained or represented by the database.  Is this a 
significant variable in terms of cataloging approach?  It is unclear how 
to draw distinctions by types of databases.  Graham will prepare a set 
of more complete distinctions to present to the group. 

ACTION: Graham will prepare and present write-up.  GRAHAM 
 

3. REVIEW OF OPEN ACTION ITEMS 

1. "TF members will make personal contact with an appropriate 
representative from the other campuses to discuss the issues raised by the 
group and to make certain that everyone had seen the initial TFER report. 
Graham will contact Santa Cruz, Alan will contact Riverside and San 
Francisco, Hixson will contact Irvine, Hayman will contact Berkeley." 
(Minutes 11/20 Everyone). 

Lynne Hayman has contacted UCB; Esther Fulsaas was identified (by Amanda 
Barrone at UCB) as the UCB contact for input into developing guidelines. 
Lynne sent Esther Fulsaas  materials. 

Graham has been in contact with Hui-Yee Chang of UCSC, and will 
continue to be in contact with her when Hui-Yee returns from an absence. 

Bob Alan will contact UCR and UCSF soon.  He also sent updates to HOTS 
and notified them about the website. 

ACTION: Maintain contact with "buddies"- EVERYONE 
 

2. "Some discussion of matching keys ensued, the OCLC number vs. the ISSN 
vs. the LCCN, with the LCCN (ISSN?) seeming most promising."  (Minutes 
12/10 -needs clarification from Doherty). 

There are actually two separate topics/discussions in this item: 

        1. What is the best method for campuses to compare records 
existing in their local databases with those in the proposed database of 
cataloging records for CDL titles (i.e. OCLC nos.? RLIN nos.?). 
        Options for distributing cataloging records for CDL resources need 
to be developed further before specific match keys can be identified.  The 
technical capabilities of local systems to load and dedup incoming 
records is a critical element to explore.  There will be a need to 
protect local campus data if record loading is based on an overlay 
process; each campus will need to develop its own field protection 
profile. 
        There will also be a body of tasks requiring human intervention 
needed to complete the process of adding CDL records to local databases, 
regardless of the specific record loading process adopted.  Although the 
intellectual work involved with initial analysis and creating CDL records 
can be done centrally, the process of integrating these records locally 
will vary on each campus and will require staff.  The nature of the local 
tasks and the level of staffing needed (classification and time 
percentage) is not yet clear.  The cost value of the record sharing 
options also still must be explored. 
        Another question is how CDL records loaded by campuses will be 
handled when update tapes of campus records are sent to Melvyl.  How would 
these duplicate records be handled in the Melvyl database? 

        2.  What would be the most reliable match key(s) for clustering 
CDL records with existing records in Melvyl? 
        This question will be examined carefully as proposals for record 
distribution options and possibilities for a CDL location reporting scheme 
are developed.  Rebecca Doherty is working on technical alternatives for 
record distribution and will report progress at the 1/15 meeting. 

ACTION: Doherty will report progress on technical alternatives for 
record distribution at the 1/15 meeting.  DOHERTY 
 

3. "Users should be able to link from a CDL Directory entry to a record in 
the Melvyl database.  Hixson observed this approach had already  been 
adopted by several systems and volunteered to investigate and bring 
information to the group."  (Minutes 12/10 Hixson). 

ACTION: Carol Hixson will report on this action at a future meeting. 
HIXSON 
 

4. "Cost and feasibility and of establishing an OCLC UC FTP site." 
(Alan). 

Alan has received information from OCLC on cost and feasibility of 
establishing an OCLC UC FTP site.  He will study it and report at a future 
meeting. 

ACTION: Alan will report.   ALAN 
 

5. "Rebecca reported that Brian Warling would be happy to 
participate in one of our calls next year.  Pat suggested that she speak 
one-on-one with Brian prior to the phone call to give him some background 
and prepare him for our questions.  Action: Pat will speak with Brian 
Warling about plans for the CDL Directory."  (Minutes 12/18 French). 

French spoke with Brian Warling about coordinating task force plans 
with CDL directory planning.  Areas of mutual interest include examining 
the possibilities for sharing descriptive data contained in MARC records 
to provide description for resources listed in the directory, 
investigating technical alternatives for sharing this data and 
clarifying/developing the optimum relationship between subject vocabulary 
used in cataloging records (primarily LCSH and MESH) and the subject 
browsing scheme in the CDL Directory.  Warling will participate in the 
1/15 meeting.  French will draft questions and topics in advance. 

ACTION: French will send out a list of topics before the next meeting. 
FRENCH 
 

6. "Rebecca will talk to MelWeb programmers about displaying 856 $3." 
(Minutes 12/18 Doherty). 

Doherty reported that programming to display the 856 $3 materials 
specified data in Melvyl may be possible to do soon.  She will consult 
further at CDL and report back. 

ACTION: Doherty will report progress at a future meeting.  DOHERTY 
 

7. "Carol offered to consult Eric Jul for more information on PURL 
servers. In the meantime, all were encouraged to look at the PURL 
website  http://purl.oclc.org.  Action: Carol will consult with Eric Jul on PURL 
server models."  (Minutes 12/18 Hixson). 

Hixson has contacted Erik Jul at OCLC and she has referred to OCLC's 
website on PURL servers.  She will report further at a future meeting. 

ACTION: Hixson will report at a future meeting.  HIXSON 
 

8. "Rebecca suggested we might make a recommendation about PURL servers in 
our final report or we might write to Joan Gargano to formally request 
that CDL investigate the possibility while we continue our work.  The task 
force agreed we should pursue both options.  Lynne offered to write to 
Joan after the holidays."  (Minutes 12/18 Hayman). 

Hayman has received a reply from Joan Gargano indicating that CDL 
Management has recommended the deployment of PURL servers and that the 
future expansion of that service would be "linked to the functional 
requirements of services as they are implemented."  The PURL service could 
be either centralized or distributed, depending upon application.  The 
task force should identify the need for the service and outline how we 
would like a central PURL server to work.  The task force would want to articulate such things as whether it is advisable to update PURLs in OCLC or just 
locally, and who would have access to the PURL server for the purposes of 
contributing corrected URLs.  Once there is a go-ahead from the SOPAG and 
the ULs, the details of deployment can then be addressed. 

ACTION: Lynne Hayman will rewrite this action item. HAYMAN 
 

9. "Crystal will ask Karen Cargille or Terry Vrable about whether the same 
URL is used by all campuses for CDL-licensed resources."  (Minutes 12/18 
Graham). 

ACTION: Graham will report on this at a future meeting.  GRAHAM 
 

10. "Rebecca reported that she's been thinking more about different 
central cataloging agency scenarios and their ramifications for CDL and 
the campuses.  She will send a summary sometime after the holidays. 
Action: Rebecca will send the group her thoughts on central cataloging 
scenarios." (Minutes 12/18 Doherty). 

ACTION: Doherty will report at a future meeting. DOHERTY 
 

11. "Concern was expressed about separate records coming in from CONSER 
and other input sources.  Crystal volunteered to write up the problem for 
further discussion with Rebecca or Laine.  Action: Crystal will write up 
concerns about inputting separate records from outside sources."  (Minutes 
12/18 Graham). 

ACTION: Graham will report on this at a future meeting. GRAHAM 
 

12. "Pending Rebecca's agreement, all agreed that continuing to meet on 
Fridays 10:00-12:00 in January was satisfactory.  Action: Ask Rebecca 
whether she can meet at this time in January." (Minutes 12/18 Doherty). 

ACTION: The Group agreed that meetings in January will be on Fridays from 
9-12 (1/15 and 1/22).  The meeting in the last week on January will be on 
Wednesday the 27th from 10-12.  EVERYONE 
 

13. "Regarding attaching holdings code for electronic titles in OCLC: OCLC 
has made conflicting statements regarding  holdings in OCLC.  The group 
agreed that we should not add holdings because it would be confusing for 
ILL and the electronic publishers generally prohibit lending in their 
licenses.  Action: UCSD will write a formal letter to OCLC to persuade 
against requiring holdings byte."  (Minutes 12/23 Graham). 

UCSD has identified a group of staff to draft a letter to OCLC against 
adding holdings for remote resources in OCLC.  Graham will report 
back at a future meeting. 

ACTION: Graham will report back at a future meeting.  GRAHAM 
 

14. "The group agreed that there need to be notes added for various 
purposes: - to express that these topics are under active discussion and 
still negotiable, - to direct questions and comments to Task force 
members, - to state that this page is considered to be a communication 
mechanism, - to indicate that minutes from previous meetings are available 
for reference.  Action: Pat will draft a disclaimer and add it to the web 
page at the top of the outline."  (Minutes 12/23 French). 

Pat French has added a disclaimer to the report outline on the webpage. 

15. "How should the records be linked when the electronic form needs to be 
cataloged as a database?  a. Do cataloging rules allow using 78X fields 
to connect the two?  b. All agreed that it is important to lead the user 
to the other record somehow.  Action: Crystal will investigate with LC and 
CONSER."  (Minutes 12/23 Graham). 

Graham reviewed guidelines for using linking records for 
monographs and serials using 7xx fields in the LC and CONSER 
documentation.  Her findings (reported also via email) were that such 
links are not directly prohibited, but that the documentation indirectly 
indicates that formal links between the two formats are not used.  Crystal 
contacted Jean Hirons at CONSER, who indicated that the time may be right 
to review this practice and amend it to expressly permit these links. 
Graham will post this question to the CONSER list for discussion and 
report back at a future meeting. 

ACTION: Graham will report back at a future meeting.  GRAHAM 
 

16. "When a record is closed would we keep the 856 and the CDL location 
symbol on the closed record?  There are several changes possible: a. 
retain the 856  b.  close out 856 $3  c.  close out the fixed fields and 
publication date  d.  make linking entries  Action: All agreed that it 
was important to consult with their respective staffs before making a 
decision."  (Minutes 12/23 Everyone). 

Graham reported that UCSD cataloging staff who currently work with 
electronic resources were amenable to closing out the print record (ie. 
the fixed fields and publication date) when a print title ceases as well 
as retaining the 856 field(s) on the closed record and possibly adding a 
linking entry to the new record.  There may be technical considerations 
related to other kinds of record linking in MELVYL raised if multiple 
records carry the 856 links to a single website.  These potential 
complications were further discussed in Action Item 17 below. 
 

17. "We discussed the possible complications for Melvyl arising from 
having the same 856 on multiple bib records or multiple 856s on the same 
bib record or series of bib records.  There are still technical issues to 
be uncovered.   We need to understand more about the implications for 
linking with the CDL directory and with A&I databases in Melvyl.  What are 
the problems associated with "One to Many" and "Many to One" linking 
relationships?  Action:  All decided this was an issue to consult about 
with Rebecca Doherty or Brian Warling."  (Minutes 12/23). 

The overall complexity of linking between databases in Melvyl (A&I 
hooks-to-holdings, linking with the CDL directory) are complex and need to 
be better understood.  The problems associated with "One to Many" and 
"Many to One" linking relationshiips need to be explored before deciding 
on guidelines for carrying the same 856 fields on multiple bib records. 
What are the implications for linking between records using unique numbers 
when the numbers themselves are used in multiple records and therefore are 
no longer unique? 

ACTION: Hayman will write up a clearer description of the question 
and resubmit it to the group. HAYMAN 
 

18. "Crystal will write up a description of the problem of 
aggregretor databases for future consideration."  (Minutes 12/23 Graham). 

How to handlle aggregretor databases remains unclear.  Graham will write 
up a description of the issues for future consideration. 

ACTION: Graham will write up a description of the issues of aggregretor 
databases for future consideration.  GRAHAM 
 

4.  FEEDBACK REGARDING QUESTIONS/ISSUES REFERRED OUT FOR COMMENT BY TASK FORCE MEMBERS. 

UCLA and UCSD have both reported their reactions to task force work via 
email.  UCSD staff is strongly supportive of the overall strategy being 
proposed.  Carol Hixson has presented the concept to UCLA staff through an 
email summary but has not received a response yet. 

UCSB has discussed it in two separate library committees (Task Force on 
MARC Cataloging on Electronic Resources and Committee on Access to 
Electronic Resources).  Hayman expressed the desire for more public 
services input in response to a task force member's assertion that 
public services librarians favored the single record 
approach. Cecily Johns, UCSB's representative to SOPAG, suggested asking 
SOPAG for additional formal input from other UC groups on the overall 
plan.  UCSB public service librarians expressed concerns about ways public 
displays of single records could be confusing.  UCSB feedback 
in these two meetings expressed support for using separate records.  There 
is also concern about the number and complexity of decision points raised 
by the single record approach.  No action item was identified. 
 

5.  DISCUSSION OF DECISION POINT 2: PAPER VERSION CHANGES TITLE 
REQUIRING SUCCESSIVE ENTRIES; EJOURNAL AT ONE SITE. 

The draft minutes of 12/23 indicated that there was a concensus that a 
successive entry record would be prepared when a print journal changes 
title and 856 links for the website would also be added to the new 
record.  As Hayman disagreed that decision point 2 had yet been 
directly addressed, the minutes will be corrected to document the 
concensus on constitution of 856 fields without a statement regarding a 
decision.  Hayman proposed creating a new separate record for the 
electronic site when changes occur to the print source.  This would create 
a gradual shift to separate records over time. Graham said a study by UCSD 
indicated this situation would rarely arise.  Hayman thought that it would 
be a disservice to users to have a number of successive entry records 
pointing to a single digital resource, particularly when the title of the 
resource is a single title. 

This question served to reopen the discussion of the overall merits of 
using single records to represent electronic resources in the catalog. 
No conclusion was reached on this decision point and it will be considered 
again at a future meeting.