Re: OCLC recommends Open Data Commons Attribution License

From: Jonathan Rochkind <rochkind_at_nyob>
Date: Wed, 12 Sep 2012 16:58:05 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
On 9/11/2012 9:06 PM, Karen Coyle wrote:

> Not in the way that they are requiring it -- again, read their
> documentation. Attribution is on the data set, not individual data, and
> you only need to say that "WorldCat data was used in this project."

Huh, if this is indeed true that would be AWESOME.

So you're thinking they only require attribution on a single 'about this 
service' page or something like that? That's obviously a lot more tractable.

And looking at the page you linked to -- yeah, it does seem to say that, 
nice!

Weird, I swear I've spent a LOT of time looking at the various OCLC 
documentation on this (sometimes hard to find and confusing and 
contradictory)... but I don't think I could had come accross that page 
before:

http://www.oclc.org/us/en/data/attribution.html

Don't know if it's new or improved since I last looked (perhaps they 
responded to my bitching by clarifying!), or if I just missed it before 
amongst all the various documentation and clearifications and guidelines 
and norms documents.

But yeah, that's more reasonable, much better than I thought. Still not 
completely in the clear, there are still issues.

For one, despite OCLC's statement on Europeana.... OCLC's statement 
doesn't make a lot of sense. It's not clear if they are saying that this 
is a special exemption/license for Europeana, or if they're saying that 
their ODC-BY license already let you contribute data to Europeana. Note 
that Europeana is CC0, but just has a non-enforceable request for 
attribution.  So if they're saying the ODC-BY already let you contribute 
to Europeana.... that seems to mean that in their opinion anyone who 
wants to can take WorldCat data and put it in a CC0-licensed database 
with a simple non-enforceable request for attribution to OCLC. And then 
someone else can take the data from that CC0 database, and choose not to 
comply with that non-enforceable attribution. So why even license as 
ODC-BY with an attempt to legally enforce attribution, if you think 
anyone can 'launder' the data through a CC0 database and not be legally 
compelled to attribute?  Very confusing.

Also, there are still issues of "attribution overload" in an envisioned 
world of lots of remixing and deriviation and re-use.  You could easily 
have to include dozens of attribution notices on your 'about' page, 
including some for sources that in fact no longer have much contribution 
to your data.  Let's say I include data from A, B, and C, all of which 
are ODC-BY.  And B included data data from X, Y, and Z, all of which are 
ODC-BY.  And Z included data from OCLC.  Now my 'about' page needs to 
credit A, B, C, X, Y, Z, and OCLC.  It may be that over time all of the 
OCLC data was overwritten and merged with other data and no actual 
WorldCat contributions remain... but unless I (and A,B,C,X,Y, and Z!) 
were (expensively!) tracking all the provenance, there's no way to be 
sure of that, so the OCLC attribution requirement will effectively 
remain forever.  You'll keep picking up more and more attribution 
requirements the more data sources you use, all of which are 'viral' and 
effectively perpetual for lifetime of your database, you'll end up with 
a lot of credits on 'about'.  Still, it could be worse.
Received on Wed Sep 12 2012 - 17:00:17 EDT