Telephone Meeting Minutes
1/16/14
1) Agreed to have weekly calls at 2:00 PM Fridays
2) Discussed set up to receive MARC files from ctsbulk-l mailing list, bulkload Exchange Group account & use of Outlook Express. Workstation configurations quite different (use of OE, for example).
3) Discussed what it would take to get LS tools configured to Columbia specifications.
4) Discussed model of having files processed by LS tools at Cornell, and then an authorization created to retrieve from Cornell and Load at Columbia.
5) Guest authorizations across institutions needed. Net ID's and Unis ultimately are required. See TSI Software and computer setup wiki.
6) GB discussed current work building repository of XML records for Hathi Trust Government documents.
7) Neither institution is currently using or investigating the Global Data Change feature of Voyager.
8) Will create email alias that vendors can send MARC files too rather than FTP. Include Cornell and Columbia recipients
9) Cross-shared Voyager file MARC-21 Repertoire.cfg adding Bengali, Devanagari, Tamil and Thai (Cornell) and left to right display character + Armenian (Columbia)
- will likely add Syriac and Ethiopic soon (also now available as scripts in Worldcat).
1/24/14
1) Continued discussions about having a joint email account for vendor MARC files. GB spoke with Chris Manly about automation of this process to feed into ctsbulk-l routines. Columbia idea was for TS to triage records (to time with material receipts) and drop into a central directory from which cron jobs load what is retrieved.
2) The draft document provided in Phase I of TSI for "Procedure46" using LS-Tools may already be too out of date to be of much use for Columbia's reference. Evelyn has LS-Tools access account for processing POOF! orders, so there is some use in place.
3) Question was asked whether LS-Tools has "versioning" and what version is currently in use? - it does not have version names. New functions and reports are frequently added. In past 1.5 years it has been rebuilt with a complete new architecture... this might be about the 4th time that has happened.
4) We'll set up a Web X display to highlight the capabilities and functions of LS-Tools for the team at Columbia to view. This hopefully would include the developer at the Cornell end to answer questions.
5) Decision made that we need to make telephone meeting notes availble on WIKI.
6) We'll try to develop use cases for investigations of comparing tools such as Pick and Scan, Strawn's location changer, Record reloader, etc. to see if there are any cross institutional areas to make collaborative use of these tools or existing workflows and define best practices.
7) MDW will try to get certain Strawn tools installed at Columbia.
1/31/14
1) Cornell is working on ebook record enhancement using OCLC API and a script with their own (not OCLC's) matching algorithms.
2) OCLC Metadata collection manager not considered helpful if Columbia uploads older Precat holdings to Worldcat, as our records would only get upgraded if there was a change to the master record in OCLC.
3) Human intervention seems a necessary element, but Voyager Select might be used to harvest ISBN's from existing Precat records to batch run in Connexion Client and create a selected record file to load after some MARCEdit work (948, etc.). Selecting one "Precat Month" at a time.
4) Re-examine YBP firm orders and Shelf Ready. Can inability to merge holdings record be worked around ; possilbe use of batchcat.dll with Perl scripting.
5) Peter Martinez has looked at documentation for Global Data Change module of Voyager.
2/28/14
1) Renew determination to set up LStools demo and discussion meeting in which "next steps" are to be discussed. Attendees to be Gary Branch, Chris Manly, Peter Hoyt (Boaz if he wants) from Cornell and Mark Wilson, Gary Bertchume, Evelyn Ocken and Jennifer Chong from Columbia.
Demo topics to cover load of Columbia Herta Berengeur records, Reports that the tool can generate, Show an example of data error correction functionality, like initial articles.
2) Gary mentioned that Cornell has a mandate to clean up data as much as possible before migration to ALMA ; Some other data clean up incentives are blacklight driven.
3) Cornell is sending their "entire catalog" to OCLC for FAST heading application. Including possible use of local headings to do so. (Leadership by Naun of CTS and Steve Folsom from Blacklight project)
4) Cornell original catalogers are addign FAST headings as part of their procedure using a web tool.
5) Discussion on mixed practices of 948 fields - Columbia does not count batch loads or use for acquisitions collection statistics or for record manipulation - At Columbia it is only used for cataloging production replacing the use of manual sheets.
6) POOF is being developed to possibly report lost and missing items and direct these to selectors to decide if re-orders are wanted. The WCS MATRIX would be used to generate orders.
7) Columbia suggested that perhaps ISBN displays in POOF could function as they do in GOBI and that links outward would search the OPAC?
8) A better preliminary interface is being investigated for POOF, including adding RDA values.
(Meeting in Item # 1 is set for 1:00 PM on Thursday, March 13th)
3/13/14 WebEx Demo of LS-Tools
1) Audio quality of WebEx connection was frustratingly poor, though it was very helpful to see its functionality.
2) Record set harvest was demonstrated; the SQL is visible as the definitions are added - Also, the administrative management features were demonstrated.
3) The workstation file structure at Columbia is not set up to work in this method. Cornell would have to visit to see what would be involved in running this locally.
4) Question of a visit to Columbia by an LS-Tools team was discussed, is this desired, helpful? LS-Tools performs identical jobs to local routines that are in place at Columbia.
5) Next steps regarding whether to proceed with LS-Tools would be discussed at Columbia.
3/13/14 - Post meeting phone call
1) Webex quality issues highlight that a site visit would be a better way to demonstrate the LS-Tools
2) Possibly as many as 4 Cornell personell could visit to explore a Columbia instance.
4/9/14 - Meeting (Witte, Bertchume, Ocken, Wilson)
LITO assessment of configuring LS-Tools at Columbia:
There are significant structural differences in the way Cornell and Columbia support technical services processes, specifically record loading. Cornell has chosen to decentralize processes that are centrally managed at Columbia. Staffing in the Library Systems group at Columbia is based in large part on the central management of these processes. It is my strong sense that any efforts to have Cornell convert to a centrally managed process or Columbia to convert to an "LS-Tools" model would entail considerable costs and provide benefits of a very brief duration (assuming a joint Alma implementation some time in the next few years). It is likely that such a conversion might not be complete before Alma implementation starts in earnest.
4/11/14
1) Adam Chandler joined us as the new Liaison to TSI Steering
2) Discussed the assessment of LITO regarding LS-Tools testing at Columbia it was agreed to make a Proof Registry entry to account for the recommendation that it not proceed.
3) Peter Magierski had identified that Cornell had access to enhanced bib records for Kotobarabia ebook collection. Cornell agreed to share these files as the Serials Solutions records were sparse and did not have transliterated versions of the titles.
5/9/14 (unedited)
1) Kotobarabia
"Unicode provides seven characters (U+200E, U+200F, U+202A, U+202B, U+202C, U+202D, U+202E) to help control these embedded bidirectional text levels up to 61 levels deep" - these characters in the Kotobarabia file are problems in Voyager and Marcdump utility.
Thought there might be LCSH in more records, but those OCLC did not have LCSH either...just one Ibn Sina title.
Do they validate in Voyager at Cornell? (no, they do not)
12 or so cornell did not upload, they use them anyways.
Open in raw text to remove invisible characters?
Serials solutions - provider id's and dv ids to link up to license
2) Cornell will send a FAST headings sample file
690 field.
FAST/NIC for local FAST
or $2. local
3) Casalini transition to new system was causing invoicing and MARC record errors in files provided to Columbia, Cornell experienced no such issues.
4) They load approvals from major vendors after selection...
Chroning batch match, what to do with level 3 records over 3 years load?
6/13/14 (unedited)
1) Kotobarabia - Columbia removed the unwanted Unicode characters from the file.
2) Kamar records - Another Eastview collection similar to Kotobarabia. Cornell does not yet have access, will we?
3) Fast heading web/ex for fast headings. Able to use headings that do not exist. /NAF/NIC is a fast $2
With identified headings.
4) Gary will ak Sarah For examples of locally created Fast headings.
Had been speaking with OCLC for some time to convert existing catalog. OCLC encouraged local.
Extract of local created ones for upcoming weeks, and compare with what they create facets for blacklight.
What is interesting about FAST is that it does not require human intervention.
A lot of effort to maintain with batch work.
5) Cornell will have some involvement with Bibframe testing..
Cornell Batch Processing leaves 066 in even if it is in UTF-8 and import the records in MARC-8.
Given the choice they would ask for MARC-8 because they catch the error doing their own conversion.
They were getting SS records in UTF-8 and will now ask for MARC-8.
856 subfi ss ID Data ID all to link to license terms. Not sure if DLF or pdf.
899 codes will switch to SS provided.
Baby Steps - Global Data Change module... Configure for use... Chris Manly
Do we have a test instance? (Yes, now the FY close testing instance will be used for Global Data Change testing)
Theirs refreshes from production several times a years... use it for FYC.
ebook enhancement... SS ID, link string.. .
They are revising their paramaters... for this, their programmer made an assumption... considered 035 $z's errors, only ran with $a
Extract of ISBN's via OCLC API like Yankee approval matrix to pick best record for person to decide whether to merge or not.
LC with Call numbers
other full LC
PCC with Call #
anything PCC wo call numbers
K or M with call
k or M without Call
Cornell will share YBP matrix
6/27/14 (unedited)
What kind of things can be tested in the Global Data Change module?
Language code errors or CP
GMD or RDA 3xx
040 $b eng (good test for inserting a string)
Identify changes to make pre ALMA or other system migration.
Something testing for Holdings data would be very useful.
MARC fixed fields from NOTIS migration
Older loaded recon records Date 1 field is blank, bars, uuuus compared to 260
948 fields?
100 10 Relationship designator ; restricted to AACR2 and RDA records.
Tag tables of language codes and compared against voyager.
ISBN's!
Lost a vendor used for smaller press / Booska is now out of business, using Ambassador, EDI set up.
Response time could be improved.
Last used them 9 years ago...
Usually matrix assigned, for smaller press or not current year, videos. GB used to be head of orders, we considered "BookHouse" they had a OP service, but ultimately went to Busca.
Matrix used to alleviate staff time commitments
Cornell has GDC configured. SysAdmin profiles need to be developed ; Peter Martinez and Gary will try.
They use theirs (refreshed every few month for Batch processing) refreshed by CM.
Gary has read through the manual - Mark has not yet.
GUI is a lot like vger select ; kind of clunky.
Store files on the server. So if you screw something up you can restore immediately. You can see befores and afters ; side by side marc views to be fairly sure of your results. They are thinking about batch deleting. Do entire database or selection records.
Build a query in scan or do full database.
Does no MARC validation, so you could delete every 245 and it would let you ; that is the danger of it.
Perhaps, let people identify records, let someone else do the manipulation.
FAST/NIC fast headings returns look good.
2 records with FAST/NIC
/NAF
/LCSH
OCLC will harvest them each month and will generate FAST.
ANother meeting when everyone gets back from ALA.
Lois did a study, perhaps with Karen Calhoun, about what needs to be in a preliminary record for it to match in OCLC (should be considered for Precat).
What happens with adding holdings and OCLC ID's#
7/11/14 (unedited)
1) Polycom Team Lead Meeting Tuesday
2) Our page on the wiki
3) GDC
Gary got into the GDC this week. Can get some pieces to work. Cannot yet effect a change. Working through the nomenclature worked about half and hour.
They use a desktop service person does an .ini switcher ; care will have to be taken in using the correct .ini file, a test record to guarantee database as a further safety.
440 field text extract for testing. (records with a testing series to use for various tests)
Sat with manual and still not figuring it out. Not intuitive.
Acts a lot like VGER select in setting query language.
Lot of database cleaning. 1,000 records that need work every day. Historical mistakes in 948 fields
They get daily reports. Out of heavy stuff in 10 days or so....
Updates MFHD's and items.
Second indicator in 852 fix would be desirable.
Perhaps that can be an enhancement request. No ideas as to what will be available in Voyager 10, nothing exciting in voyager 9. GDC might be worthwhile for enhancement requests as it is still new. We can identify needs.
MARC Serials and AACR2 - Gary has a prevent for the 33x from sersol. USE OCLC API and take that sheet into
API is on the OCLC website for downloading, it is what POOF uses. Is configurable, OCLC Master, ISBN, LCCN, Restrict by Holdings Library (Like you can do in CONNEXION) LI: NJP (will append to search if you put a space between them for multiple qualifiers).
9/05/14 (unedited)
Loading of new Serial Solutions records in with the newly formatted 899 fields $i ; Gary more involved with the MARC aspects of it..
Got back their fourth sample of FAST headings and they look ready to move forward. The entire database will be done. For Blacklight facets. Plus FAST/NIC
Some authority fallout and clean up results from what they learn, older headings like Journeys.
Moving two collections to Serials Solutions for Knovel. Records seem to have improved, fairly full. October 1 load, so maybe take a look.
JSTOR records from their old WebInterface. 65,000 NAXOS records from NAXOS site, but very dirty. What Options from OCLC
HathiTrust API. - Google API. Michelle Palillo.
CSV to XML transformation - template file from Calturra video files, convert to XML ; A streaming service. Locally created CSV file by those who are making corrections including course reserves.
Peter Martinez and Hannah Marshall will develop script.
Lydia Pettis is the one leading the pre ALMA cleanup effort.(Suppression, item types, Continuation orders) ExLibris Documentation.
Annex materials with order locations, reference materials with catalog happening locations.