Decide how to handle empty 100a field in MARC records

[Edward Betts]

Some MARC records have an 100a field which is blank. All such records have been added to the same author. This is a mistake.

See http://openlibrary.org/a/OL5631454A/noname

[George (george-archive)]

If we decided that Title, Author, Publisher and Publish Date were the absolute minimum metadata we'd accept for any imports, would that help? We could simply throw any records that didn't meet those requirements on the floor.

[Edward Betts (edwardbetts)]

We're missing the author for aroundaboutolden00lond:

http://www.archive.org/details/aroundaboutolden00lond

The MARC XML has a blank author (tag 100):

http://openlibrary.org/show-records/ia:aroundaboutolden00lond

The MARC Binary shows tag 100 with odd characters:

http://openlibrary.org/show-records/aroundaboutolden00lond/aroundaboutolden00lond_meta.mrc:0:705

Here is the MARC record on the New York Public library site:

http://catalog.nypl.org/search~S1?/.b12559691/.b12559691/1,1,1,B/marc~b12559691

There is a similar problem with these ones:

lemanoirde00zier ninette00pari strasbourg00pari tranquilleettour00pari boylifeofnapoleo00d185

All of them are from NYPL.

[Ben (bencompanjen)]

Hopefully anyone will read this - perhaps now that I sent a message to the tech list someone will.

I proposed in the mail to remove the author reference from all 2870 works. The only advantage of keeping the reference to an author "bad author" compared with deleting the references is, I think, that one can point to these "records imported from bad MARC data" using the particular author key. There are other ways for doing that, of course.