Yeah, the problem with the OCLC auto-IP-based gateway, as you say, is that
it only works for people in IP address ranges registered to their
university.

Even when just focusing on _our own_ users, we realized that _most_ of our
patron hours were spent "off campus". And decided that solutions that
relied on recognized IP addresses were unacceptable, our goal was always to
make the best solution we could for off-campus users too.

So I think the "best" (within existing environment) solution would  be to
maybe take the IP-address-identified link resolver as a _default_, but
always allow (with good UI) the user to select another institution from the
directory as a preference. I'm not sure if OCLC APIs support this (but they
_ought_ to, it's the right use case).

Note that I think this solution is what Google Scholar itself does --
connect you to IP-address-identified link resolver as a default, but allow
you to set a preference.

I also think that simply linking the user to a Google search on relevant
metadata (definitely including DOI if present) is a fine compromise
solution -- on it's own it's a lot of value for relatively little
development work. I think it will _often_ get the user to the right place.
  (In an ideal world, doing this link-resolver-preference-based solution
would be just as little work.... but I'm not sure how close we are to
that).

Jonathan

On Thu, Jan 26, 2017 at 9:06 AM, Haefele, Chad Mark <[log in to unmask]>
wrote:

> I realize I’m not directly answering the LibX vs Scholar question, but
> this all sounded very familiar to me:
>
> A department on campus here is doing a similar project of putting a
> massive bibliography online - we were involved in some of the early phases
> a couple years ago. They have a large bibliography of resources on a
> webpage, which is built on exports of data from Zotero. It’s behind a login
> while in draft mode, so I can’t link directly to it. But see attached
> screenshot for an example.
>
> Each article has a “Search locally” button associated with it. Clicking it
> goes to a link like this one:
>
> http://worldcatlibraries.org/registry/gateway?ctx_ver=Z39.
> 88-2004&atitle=Collaboration%2C%20Resistance%20and%
> 20Reform%3A%20Experiences%20and%20Historiographies%20of%
> 20the%20Napoleonic%20Wars%20in%20Central%20Europe&title=
> Central%20European%20History&stitle=Collaboration%2C%
> 20Resistance%20and%20Reform&date=2006&volume=39&issue=4&
> spage=547&epage=579&aulast=Aaslestad&aufirst=Katherine&
> au=Hagemann%2C%20Karen&sid=Biblio%3ABibliography
>
> This is where my memory of how we set it up gets fuzzy, but I think if
> you’re in the IP range of a resolver that Worldcat/OCLC knows about, you’ll
> get links to check for the article through it. The weakness is that if
> you’re not in one of those IP ranges, you’re at a dead end.
>
> -Chad
>
>
> ---
> Chad Haefele
> Interim Head of User Experience
> UNC Chapel Hill Libraries
> [log in to unmask]
> 919-962-3702
>
>
> On 1/26/17, 8:07 AM, "Code for Libraries on behalf of Terry Reese" <
> [log in to unmask] on behalf of [log in to unmask]> wrote:
>
>     I believe Jonathan is thinking about the OCLC Knowledge Base API (
> http://www.oclc.org/developer/develop/web-services/worldcat-knowledge-
> base-api.en.html) which does provide a way to do openurl for institutions
> in the resource.  Of course, the information in the resource would need to
> be valid, and that is dependent on member libraries keep it up to date,
> which they may or may not be doing depending on if they are depending on a
> service from OCLC that might require OCLC knowing that information.
>
>     --tr
>
>     -----Original Message-----
>     From: Code for Libraries [mailto:[log in to unmask]] On Behalf
> Of Jonathan Rochkind
>     Sent: Wednesday, January 25, 2017 11:31 PM
>     To: [log in to unmask]
>     Subject: Re: [CODE4LIB] LibX, Google Scholar with "library links", and
> other interfaces to licensed resources
>
>     I don't have numbers and this is really just my guess impression, but
> I'd be shocked if there weren't a lot more libraries in Google Scholar than
> in LibX.  It should be possible to get both numbers from publically
> available lists though.
>
>     I think the google scholar search is a good one.
>
>     The other thought would be letting users somehow direct themselves to
> their _own_ link resolvers, just like vendors do. You'd need to either
> build a directory of link resolvers for them to choose from, or I have some
> memory that there is _some_ relevant OCLC service. I don't recall if it's
> actually sufficient or has a decent API, or how extensive it's directory is.
>
>     On Wed, Jan 25, 2017 at 10:16 PM, Kevin Hawkins <
> [log in to unmask]> wrote:
>
>     > I've got two simple questions but first a long statement of
> background
>     > so you know why I'm asking.  If you're in a hurry, feel free to skip
>     > to the questions.
>     >
>     > == Background ==
>     >
>     > We've got a research group on campus that has built up an annotated
>     > bibliography of literature in their topic of study over the years,
>     > which graduate students have carefully organized according to a
>     > taxonomy of topics.  They are happy to continue to host this
>     > bibliography online as a searchable database.
>     >
>     > However, they've asked us in the library to help them think through
>     > what it would mean for them to help people reach the full text when
>     > available online (often through licensed library resources).
>     > Supporting users at our institution is easy -- we can prepend the
>     > proxy server string to the DOI, or query our link resolver, or query
>     > our library discovery product.  But I'd also like to offer some
>     > options for what they could do to help users not at our institution.
>     >
>     > Including a hyperlinked DOI or other publisher-provided permalink
> will
>     > allow users at other institutions who are on campus and using VPN to
>     > reach the resource directly, assuming their institution subscribes.
>     > So while that's an easy solution, it only helps certain users.
>     >
>     > In brainstorming with a colleague, we've come up with two other
>     > options that we might offer to the research group for helping users
> at
>     > other institutions reach the full text:
>     >
>     > a) They could recommend that users install LibX and then use its
>     > "magic button" feature to select the citation and search for it
>     > through their institution's library.
>     >
>     > b) They could build queries that go to Google Scholar and suggest
> that
>     > Google Scholar users configure "library links" in order to prepend
>     > their institution's proxy server link or send them to their link
> resolver.
>     >
>     > While I know that Google Scholar's database includes plenty of bare
>     > citations (without any links to full text), I'm not sure it's big
>     > enough to actually include the sorts of book chapters and possibly
>     > obscure journal articles found in this bibliography.  But I'm content
>     > to leave that to the research group to experiment with.
>     >
>     > == The questions ==
>     >
>     > 1. Does anyone have a sense of the number of institutions with
> "editions"
>     > for LibX versus the number of institutions with "library links" set
> up
>     > in Google Scholar?  I can't find full lists on the websites of
> either,
>     > but if we know that one is more comprehensive than the other, I'd
> just
>     > as soon recommend that.
>     >
>     > 2. Are there other options you'd suggest besides LibX and Google
>     > Scholar (with "library links") for helping people people find a
>     > library-licensed version of a resource when you don't know what
> institution the user is at?
>     >
>     > Thanks in advance for your thoughts,
>     >
>     > Kevin
>     >
>
>
>