Quoting Lee Passey <[email protected]>: > On 5/11/2010 11:24 AM, Karen Coyle wrote: > >> Quoting Jonathan Rochkind<[email protected]>: >> >>> OL can and will modify the data (not just automated transformation, but >>> human editing, yeah?) after the original import, right? >> >> OL is wiki-based, so yes, anyone can go in and edit, some some >> automated edits (adding data from Wikipedia, for example) does happen. >> In history you can see and compare the edits a record has gone through. > > This answer, while no doubt completely accurate, does not address Mr. > Rochkind's core issue: how do you get OL's current data exported in MARC > format? The answer, it seems, is that you can't.
Lee, I believe I answered this in my previous post. You can put anything you want into MARC tags and subfields (as we have all seen folks do ;-)). That does not make it useful data to import into a library catalog. OL is not an "AACR2-based" catalog. If folks would like non-AACR2 data exported in MARC tags, that can be done, but I suspect it will not fulfill the needs mentioned here. Remember, "MARC" means many things. For import into library ILS's, the content must follow AACR2 rules. Encoding the data in MARC tags and subfields is not the sole criterion. kc > > /If/ the OL record was derived from a MARC record, you can use the OL > web interface to find the original data from which the OL record was > derived. If you examine the URL closely, from there you can get to the > dump files stored at Internet Archive. You now have the MARC data dump, > and OpenLibrary becomes irrelevant. If, on the other hand, the OL record > was /not/ derived from a MARC record, such as one of the hundreds of > "vanity" records added by authors promoting their own books, there is no > way to get /that/ record out in MARC format. And if the record was > derived from a MARC record and subsequently /corrected/ (which hopefully > will occur frequently), there's no way to get the /corrected/ data in > MARC format. > > It seems to me that if I were a librarian, I would want /lots/ of data > that I could import into my own database, or I would want data for a > single record delivered to me in a structured format that an automated > process could consume. In either case, I would want that data in a > standard format so I don't have to create new, custom computer > applications; neither the OL web interface or the OL RDF or JSON APIs > satisfy these requirements. > > On the other hand, as a "patron" most of the data contained in the OL > catalog is of little interest to me unless it leads me to a /book/. > Knowing that Farrar, Straus and Giroux published _The World is Flat_ in > 2005 is not useful unless that information helps me actually get my > hands on a copy. > > So, except as an entry into Internet Archive's text archive, I'm having > a hard time coming up with a use case in which OpenLibrary is actually > useful. I think it's time for an answer to the questions, "what is the > intended use for OpenLibrary?" and "how well does the current interface > address the documented intended use?" > >> >> kc >> >>> >>> I'm personally less interested in seeing the original MARC than I am in >>> seeing OL's current data exported in MARC format. > _______________________________________________ > Ol-tech mailing list > [email protected] > http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech > To unsubscribe from this mailing list, send email to > [email protected] > -- Karen Coyle [email protected] http://kcoyle.net ph: 1-510-540-7596 m: 1-510-435-8234 skype: kcoylenet _______________________________________________ Ol-tech mailing list [email protected] http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech To unsubscribe from this mailing list, send email to [email protected]
