We're, it seems, fairly unique, at least amongst the respondents on this
list. And I completely understand that folks will disagree with our
decision. But we do encourage (promote) an interface that forces
off-campus authentication to our Summon instance. Of course, if one knows
how the host pattern works, it's pretty easy to get to UNC's Summon page
without authenticating (http://unc.summon.serialssolutions.com/). And
we've done nothing to discourage generic use other than failing to promote
The decision was one that was taken only after a great deal of discussion.
And one we would need to revisit if we looked to Summon (or some other
product) as a catalog+periodical literature hybrid. Right now we have
separate discovery layers (yet another interesting, protracted, and
continuing conversation). Here's the email that went to staff, see below.
I wanted to inform everyone of a decision that has been taken, and more
importantly, communicate the reasons behind that decision.
Summon (Articles+) is a product from Serials Solutions that enables our
users to search across the vast majority of our licensed electronic
At UNC (and we seem to be unique at this time) we have chosen to send our
off-campus users directly from their search to an authentication page,
prior to being delivered to the page where they see the results of their
search. This potentially can feel like an unnecessary step and can be
perceived as a barrier, because *anyone* from off-campus may search a
Summon site and see "results."
There are several motivations for our approach:
1. Restricted results:
Serials Solutions engages in agreements with publishers for their content.
Most publishers permit Serials Solutions to share the citation level
information with any potential user (affiliates and non-affiliates), only
restricting download/use of the document/pdf to affiliates.
However, there are some major vendors who do not permit their content to
be seen by non-affiliates. One of these is ISI who license "Web of
Science" a massive source of searchable content.
That is to say, from off-campus Summon will only return a subset of all
results to non-affiliates.
Affiliates, in this case, means anyone on campus, including walk-ins, and
anyone who authenticates from off-campus.
If you are off-campus, and have not authenticated, you will not see the
full result set that you would see on-campus. I found this out personally
when I did a search from home that I had done earlier on-campus, and few
of the items I'd identified from my office search appeared on the screen.
I was momentarily flummoxed. (Upon logging in, the expected results did
2. User interface issues:
If you are off-campus, and do an unauthenticated Summon search, there is a
small addition to the webpage at the top, that says "Off Campus? Log in
to access full text and more content." We have found that users do not
3. Authentication happens anyhow:
To actually use resources, the off-campus user will be forced to
authenticate. Authentication will have to happen, the question is when in
the process it happens.
So. Given that to actually use a resource, users will have to
authenticate. That authentication returns the richest result set. And
that the mechanism to log in is challenging to find. We put the
authentication step at the front of the process.
If you know someone who would like to explore our Summon interface from
off-campus, for instance a colleague at another institution who would like
to see ours, they can go directly to it, rather than beginning a search
from our home page. The url for this is:
Please let Kim Vassiliadis or me know if you have questions or concerns.
On 10/24/12 9:19 PM, "Jonathan Rochkind" <[log in to unmask]> wrote:
>Good to have some numbers, thanks! Even taking your largest number, 25%
>+ 12% == 37% coming from on-campus is definitely less than half, and not
>'most' use being from on-campus -- which does not surprise me at all,
>it's what I would expect.
>This is an interesting discussion, I think. Thanks all. (Except for Ross
>and that other guy having a flamewar about things entirely unrelated to
>the topic! Just kidding, we love you Ross and that other guy. But yeah,
>unrelated to the topic.)
>From: Code for Libraries [[log in to unmask]] on behalf of David
>Friggens [[log in to unmask]]
>Sent: Wednesday, October 24, 2012 9:15 PM
>To: [log in to unmask]
>Subject: Re: [CODE4LIB] Q: "Discovery" products and authentication (esp
>>> a) most queries come from on-campus
>> Really? Are people just assuming this, or do they actually have data?
>> would surprise me for most contemporary american places of higher
>For the last two months, 25.4% of our Summon traffic has come from the
>IP addresses we've given as "on campus", according to the stats
>Serials Solutions provides. Note that another 11.8% came from the
>local ISP that provides wireless for our students, so most of that
>would be "on campus" at other institutions.
>> But it may very well be the extra "restricted" content is not important
>> nobody minds it's absence. (Which would make one wonder why the vendor
>> bothers to spend resources putting it in there!).
>That's been our view (though you're making me think we should perhaps
>try and understand better what the difference is).
>The A&I results are interesting. EDS seems to promote results from
>their own A&I databases more highly than I would expect, and they're
>certainly noticeable when blanked out with "cannot be displayed to
>When Summon started showing A&I results there was some interesting
>discussion on the mailing list - they're not immediately accessible,
>so they're arguably not "in the library's collection". And Summon (as
>does Primo) has an option to "add results beyond your library's
>collection". There was some argument on the other side, that A&I
>results are important to be included, so it seems that there is
>librarian pressure as well as commercial/licence pressure.