From: "Hosburgh, Nathan" <[log in to unmask]>
Date: Fri, 27 Sep 2013 16:33:06 -0600

Stevan,

If I define Green OA as simply "OA delivered by repositories" (as
defined by Peter Suber and others) then it becomes clear that the
discrepancies I mentioned are possible.  It sounds like you are using
a more narrow definition of Green OA as the final, peer-reviewed
draft.

Nathan,

I don't think so at all. 

First of all, Green OA means OA provided by author self-archiving,

on any website, not just IRs (but that point's minor and irrelevant).

What's relevant is the definition of OA, which is free online access

to the peer-reviewed article: That means any draft from the final,

refereed, revised, accepted one onward (i.e., refereed post prints,

not unrefereed preprints).

What is an Eprint?

Eprints are the digital texts of peer-reviewed research articles, 

before and after refereeing. Before refereeing and publication, the

draft is called a "preprint." The refereed, accepted final draft

is called a "postprint." (Note that this need not be the publisher's

proprietary PDF version!) Eprints include both preprints and

postprints (as well as any significant drafts in between, and any

postpublication updates). Researchers are encouraged to self-archive 

them all. The OAI tags keep track of all versions. All versions

should contain links to the publisher's official version of record.

I don't think there would be as much controversy surrounding this
issue if we were all talking about the final, peer-reviewed draft.

We are. That's what mandates are about. And that's what publisher

Green OA embargoes are about.

Repositories are not uniformly populated with final, peer-reviewed
drafts.  But, even if they were, this would still leave the issue of
further copy editing post peer review.

That's what I said was utterly trivial -- to someone who otherwise has

no access at all. (And that's what OA is about and for: those who have

no subscription access to the refereed postprint.)

Even if we are using "green" as designated by SHERPA/RoMEO, this still
leaves open the possibility that the repository version is a pre-print
(version of the paper before peer review).
http://www.sherpa.ac.uk/romeo/definitions.php?la=en&fIDnum=|&mode=advanced&version=#colours

No it doesn't. It says:

Now apart from the silly and superfluous distinction between publishers who

allow only (refereed) postprint self-archiving and publishers who allow both

postprint self-archiving and (unrefereed) preprint self-archiving:

The target of OA is the refereed postprint, hence both SHERPA "green"

and SHERPA "blue" are OA Green.

Moreover, IRs indicate clearly whether an item is refereed or unrefereed.

Since we're not living in a homogenous Green OA world, I would not use
the availability of Green OA as a deselection criteria as Rick
Anderson suggests.  

To repeat: Rick was proposing a Green (no-embargo) journal policy

on OA as a deselection criterion, not percentage of Green OA articles.

Having worked in ILL for a number of years, I
agree with Chuck Hamaker that "the goal is to provide the version of
record of an article as expeditiously as possible, and at the lowest
cost possible in the most convenient form".  

An ILL for every click by an institutional user is a pretty pricey 

proposition -- but never mind. We were not talking about toll

access but about open access. That means what the institution

can't or won't pay for, whether via subscriptions, licenses, or

pay-to-view (ILL).

That's what the Green OA postprint is for.

I can say from experience
that faculty/researchers/scholars are concerned with getting their
hands on the version of record.  If they have to pay for it out of
their own pocket, they will often do so.  I saw this firsthand even
when an OA version was available from a repository.

What you need to count is not how often authors pay to get

the priced version even when he has access to the Green OA

version, but how often they don't.

This is not about anecdotal incidents, but quotidial (and 

rational) user  practice.

To answer the following:

SH: For users deprived of access to any version at all,
all of these points are utter trivia.

This is true for some users, not all users.  Some users do not find
these points trivial.

How many? (Not how many librarians: how many users!)

Not content with Green compared to what? Nothing?

Faculty/researchers/scholars are sometimes not content with the Green
OA version compared to the version of record/publisher's PDF for the
reasons I've already mentioned.

But this is about when they cannot afford the publisher's priced version.

Stevan Harnad