We are also still using the plugin. I special-requested a 32-bit OS when
my machine was last upgraded because the plugin is so helpful in a
number of our workflows.
One thing I have noticed with the plugin is that the coding of 007
fields often gets mangled somewhere in the
translation-to-MARCedit/back-to-OCLC process, even if you don't touch
that field in MARCedit. Then, large numbers of your records will not
validate and update in OCLC.
thanks,
-Kristina
-=-
Kristina M. Spurgin
E-RESOURCES CATALOGER
E-Resources & Serials Management, Davis Library
University of North Carolina at Chapel Hill
CB#3938, Davis Library -- Chapel Hill, NC 27514-8890
919-962-2050 -- [log in to unmask]
On 11/25/2013 13:34, Terry Reese wrote:
> On a side note, I've been thinking about taking another crack and updating
> the oclc connexion plugin because on windows 64 bit machines, it stops
> working. I'd kind of thought maybe folks weren't using it anymore and had
> been planning to send a note to ask the community about it -- so the fact
> that you are using the plugin as well makes me think that others may be
> too, so taking another look at this might not be a bad idea.
>
>
> --tr
>
>
> On Mon, Nov 25, 2013 at 12:51 PM, Susan Wynne <[log in to unmask]> wrote:
>
>> John and/or Terry,
>>
>>
>>
>> I’m a MarcEdit novice and completely new to APIs. Like Mary-Frances, I’ve
>> been trying to use the Metadata API to replace my own institution’s ETD
>> records. My process is a little different because I created a local save
>> file of WorldCat records as a starting point. Then I brought them into
>> MarcEdit (using the bib file reader plug-in), made one change to each
>> record, and then tried to replace the master records in WorldCat using the
>> API. (I’ve got a file of 1500+ to change, but I’ve only tried to replace a
>> test file of 5 records.) The replace fails and the error message simply
>> tells me that the record is not valid. I did all of this in the same day.
>> I tried it both with the file extension as .mrk and as .mrc. (We also
>> tried removing the 005 from the edited records.)
>>
>>
>>
>> The change I’m making to these records is removing the ending slash from
>> each ETD URL, so that the format of the URL will match what is harvested
>> from our IR to prevent duplication. I have made this same change manually
>> to a master record, so I don’t think that would cause the records to fail
>> validation.
>>
>>
>>
>> It seems unlikely to me that anyone else would be editing those master
>> records during the relatively short time I was fiddling with them (about an
>> hour, maybe?). So I’m trying to figure out whether the API should allow me
>> to replace these master records and I’m just making some kind of error
>> during editing or with using the API itself, or if the API does not allow
>> replacing master records at all because they can’t be locked. Please let
>> me know if you need any further information.
>>
>>
>>
>> Many thanks,
>>
>> Susan
>>
>>
>>
>> Susan C. Wynne
>>
>> Cataloging & Metadata Librarian
>>
>> Georgia State University
>>
>> 100 Decatur Street SE
>>
>> Atlanta, GA 30303
>>
>> 404-413-2729
>>
>> [log in to unmask]
>>
>>
>>
>> *From:* MarcEdit support in technical and instructional matters [mailto:
>> [log in to unmask]] *On Behalf Of *Terry Reese
>> *Sent:* Friday, November 22, 2013 5:50 PM
>> *To:* [log in to unmask]
>> *Subject:* Re: OCLC's Metadata API
>>
>>
>>
>> Thanks John,
>>
>>
>>
>> I'd planned on expanding my thought after I got off my phone. As John
>> says, the idea behind the API is that you are interacting with the WorldCat
>> metadata live (liveish). This actually also underlies a challenge with the
>> API -- because of the 005 requirement, it would make sense if users editing
>> a record could lock them and at this point, the API doesn't support that
>> functionality (though John can correct me if I'm wrong).
>>
>>
>>
>> --tr
>>
>>
>>
>> On Fri, Nov 22, 2013 at 4:55 PM, John Chapman <[log in to unmask]> wrote:
>>
>> Hello Mary-Frances,
>>
>> The Metadata API is designed to support an online-cataloging model, where
>> you
>> are reading, enhancing, and then replacing WorldCat records. The workflow
>> you
>> are attempting is closer to the model supported by our batchload services,
>> which
>> do not actually directly replace but instead merge records that have been
>> edited
>> in a local system.
>>
>> The 005 check that is preventing this replacement is the way that we check
>> on
>> whether the master record being replaced has changed since the incoming
>> "replace version" was read from the master record. If it doesn't match,
>> but we
>> allow the replace, then we have the potential of record edits being
>> stomped on.
>> Normally, the scenario that we are trying to prevent would involve an 005
>> coming in that was an older timestamp than the one on the master record; in
>> this case it is newer.
>>
>> Overall, we are trying to balance a web service design that is as close as
>> possible to the "RESTful" ideal of get/put - read/update/replace - with the
>> "versioning" capability of the MARC 21 standard, which is sadly lacking.
>> When
>> two catalogers are working on the same record, we need a way of knowing
>> which changes to accept.
>>
>> If you have further questions on the API, please let me know.
>>
>> -John
>>
>> John Chapman
>> Product Manager, WorldShare Metadata
>> OCLC, Inc.
>> [log in to unmask]
>>
>>
>> ________________________________________________________________________
>>
>> This message comes to you via MARCEDIT-L, a Listserv(R) list for technical
>> and instructional support in MarcEdit. If you wish to communicate directly
>> with the list owners, write to [log in to unmask] To
>> unsubscribe, send a message "SIGNOFF MARCEDIT-L" to
>> [log in to unmask]
>>
>>
>>
>> ________________________________________________________________________
>>
>> This message comes to you via MARCEDIT-L, a Listserv(R) list for technical
>> and instructional support in MarcEdit. If you wish to communicate directly
>> with the list owners, write to [log in to unmask] To
>> unsubscribe, send a message "SIGNOFF MARCEDIT-L" to
>> [log in to unmask]
>>
>
> ________________________________________________________________________
>
> This message comes to you via MARCEDIT-L, a Listserv(R) list for technical and instructional support in MarcEdit. If you wish to communicate directly with the list owners, write to [log in to unmask] To unsubscribe, send a message "SIGNOFF MARCEDIT-L" to [log in to unmask]
>
________________________________________________________________________
This message comes to you via MARCEDIT-L, a Listserv(R) list for technical and instructional support in MarcEdit. If you wish to communicate directly with the list owners, write to [log in to unmask] To unsubscribe, send a message "SIGNOFF MARCEDIT-L" to [log in to unmask]
|