 |
 |
At MIchigan, we direct users to an interface within our library site. Over the last week, a bit more than 28% of searches against the Summon service were from on-campus (hardwired) computers. An additional 25% of searches were conducted from authenticated campus wireless networks. So slightly more than half of our searchescome from on campus. Of the rest, 39% in the last week were in the U.S. (and I'm boldly assuming that most of those are in and around Ann Arbor, but the GeoIP database we use is at the country level, not city). On Thu, Oct 25, 2012 at 9:09 AM, Brad Baxter <[log in to unmask]> wrote: > We at GALILEO have thought about this. We authenticate everyone. If > they're on campus it's by IP, if not, by password. So everyone sees > everything. This also lets us define which EDS profile to direct the > user to. > > -- > Brad > > On 10/24/12, Jonathan Rochkind <[log in to unmask]> wrote: > > On 10/24/2012 2:52 PM, Joe Hourcle wrote: > >> Do they even get a message that they've been restricted? > >> > >> I would think that having a message such as : > >> > >> 74 records not shown because you weren't authenticated > > > > As far as I know, none of the products have the ability to tell them > > _how many_ records in the total result set have been suppressed. > > > > All of them have the capability, of course, to just tell users "Some > > records may have been suppressed, try logging in." I was trying to get a > > sense of how many implementations use such a message, and if anyone > > knows if it results in logins. > > > > But I think my conclusion is that few implementers have thought about > > this, and most off-campus users probably don't get restricted content. > > :) Which may be just fine -- the amount of restricted content in a given > > product is also unclear (hard to compare between products, hard to even > > know for a single product if there's importnat 'restricted content' > > that's being missed). > > > > EDS is unique in having a weird way of suppressing content. If user is > > not authenticated, you get 'holes' in your result set. If you asked for > > a page of 10 results, you might only get 5 results, with 5 blank spots. > > On the one hand, this is a weird UI, on the other at least it makes it > > obvious you're missing content, and how many records from the _current > > page_ (if not the entire result set) have been suppressed. EDS does > > seem to have a fairly significant amount of restricted content, it's > > unclear if the other tools have around the same amount or less, like I > > said very hard to evaluate. > > > >> > >> would be enough to spur most folks to log in. > >> > >> What I hate is when you do a search for something that you *know* should > >> be there, and it's not ... then you find out that they're using IP range > >> or DNS matching, and not telling the user that they've intentionally hid > >> stuff. I think I've gotten most of the stuff straightened out with our > >> local library, but I have no way of knowing for sure. > >> > >> (my desktop machine's doesn't resolve in the 'gsfc.nasa.gov' domain, > and > >> not on the most common network here ... so most systems' test for 'is > this > >> a local person' fail, and I get treated as an outsider ... I actually > get > >> better service using my personal laptop on the wireless network for > >> visitors) > >> > >> -Joe > >> > >> > > > -- Ken Varnum | Web Systems Manager | MLibrary - University of Michigan - Ann Arbor [log in to unmask] | @varnum | http://www.lib.umich.edu/users/varnum | 734-615-3287