Print

Print


intentionally sent or not - it is very informative to see these things being discussed.

thank you Gary and Kevin



From: Program for Cooperative Cataloging <[log in to unmask]> on behalf of Gary L Strawn <[log in to unmask]>
Sent: Thursday, August 13, 2020 13:38
To: [log in to unmask] <[log in to unmask]>
Subject: Re: [PCCLIST] New authority toolkit version available
 

Kevin is exactly correct in all points. (I stupidly sent this message to the list—again!—instead of to the intended recipient.) I was describing how things look in my blinkered world, not how they may actually be.

 

Gary L. Strawn

Northwestern University Libraries

Northwestern University

1970 Campus Drive, Evanston IL 60208-2300

e-mail: [log in to unmask]    

office: 847/467-7240

storage facility: 847/467-4619

authority toolkit documentation: http://bit.ly/1Hl1jST

Forsan et haec olim meminisse iuvabit

 

From: Program for Cooperative Cataloging <[log in to unmask]> On Behalf Of Ford, Kevin
Sent: Thursday, August 13, 2020 1:36 PM
To: [log in to unmask]
Subject: Re: [PCCLIST] New authority toolkit version available

 

I want to offer some clarification here.  From some off-list conversation with Gary, I understand what Gary means when he says “format” and “some fields had inconsistent results” (or at least I think I do) but I don’t want this audience to confuse Gary’s words, which refer to technical serialization differences, with the fact so many of us live in the world of fields and subfields of the various MARC formats.

 

In short, Gary is referring to syntactical differences in the XML that have no semantic import.  The record format remains MARC Authority.  And I can think of no reason why the data in the fields (or subfields) of the MARC/XML would be inconsistent with their MARC 2709 representations and/or the very same records viewed elsewhere (for example, LCCN’s permalink and ID have the same Names and Subjects and there should be no daylight between the two sets). 

 

What Gary is encountering are artifacts left behind from our pre-processing workflow, which recently underwent significant change.  Tools that process the XML as XML, which converts the XML into some tree of information that can be navigated programmatically, will completely ignore these artifacts.  If, however, the XML is treated as a string, which I believe is what Gary’s tool is doing, then it is more susceptible to these non-semantic changes.

 

All the best,

Kevin

 

--

Kevin Ford

Library of Congress

Washington, DC

 

 

From: Program for Cooperative Cataloging <[log in to unmask]> On Behalf Of Gary L Strawn
Sent: Thursday, August 13, 2020 1:10 PM
To: [log in to unmask]
Subject: Re: [PCCLIST] New authority toolkit version available

 

I think I know. Now.

 

In addition to changing the way records need to be requested from LC, LC also changed the format of those records. But I believe that LC also has a cache of records that exists to speed up requests, and records in that cache used the old format. That explains why the terms I was using for testing were all successful—they were in the cache; why some fields had inconsistent results (one heading verified, one failed); and other records were complete failures. The complete failures were for things not in the cache, which came through in the unexpected new format.

 

I have made allowance for the new format in my code, and I have then had success with all of the examples people have sent me.

 

So I need to find a USB drive (long story) and then I can make a new installer.

 

Gary L. Strawn

Northwestern University Libraries

Northwestern University

1970 Campus Drive, Evanston IL 60208-2300

e-mail: [log in to unmask]    

office: 847/467-7240

storage facility: 847/467-4619

authority toolkit documentation: http://bit.ly/1Hl1jST

Forsan et haec olim meminisse iuvabit

 

From: Program for Cooperative Cataloging <[log in to unmask]> On Behalf Of Shorten, Jay
Sent: Wednesday, August 12, 2020 4:44 PM
To: [log in to unmask]
Subject: Re: [PCCLIST] New authority toolkit version available

 

My 370 verification is still not working, but I’m not sure why.