Wednesday, April 10, 2013

Computers in Libraries 2013 - Day 2 - In the Cloud : Personalized Virtual Desktop

Anastasia Diamond-Ortiz, C. J. Lynz & Olivia Hoge of Cleveland Public Library presented this really interesting talk discussing their efforts at providing virtualized desktop environments for their patrons.  This wasn't something I was really considering doing before the presentation and after the presentation I'm probably considering it even less, but it was truly interesting nonetheless.

Cleveland has had a problem that probably all public libraries have had in that patrons come in with special computer needs, like unusual software requirements or the desire to have some kind of persistent  desktop that can be made available for them.  In most libraries I'm familiar with this problem is addressed by either "I'm sorry..." or a kind of half-measure (e.g. having a computer that is not connected to the Internet on which they can install software, but after the computer is rebooted it returns to its prior state forgetting all that had happened).  In rare cases I've been able to add software for patrons who needed it and it was free and something I knew I could trust, but most of the time it's just something we can't do due to the nature that we have machines that need to be used by a wide variety of people.

Cleveland, which being in an urban area and having a lot of patrons who don't have computers at home but may need to do special things with computers to get certifications and the like, has had this problem on a pretty large scale.  They decided to address it by setting up a system called MyCloud in which users could setup an account with the library and check out a thin client laptop for 3 hours that was usable only within the library.  When using that thin client laptop, they were connected via a Citrix server to their own Windows 7 virtual machine with 5 GB of storage space.  They have administrative rights on the virtual machine and can install whatever they like and save whatever they like as long as they don't use up the 5 GB of storage space.  They take complete responsibility for the machine and if they accidentally delete all of their files or get the virtual machine horribly infected with viruses they have to live with the consequences of that, but because of the nature of the network and the virtual machines, no other users are affected.

It's an interesting system and it introduces a host of possibilities and a host of questions.  The back-end that Cleveland had to implement to make this work was enormously expensive and they are already out of hard drive space with only 75 people using the system.  They can (and I'm sure will) upgrade the storage space and the rest of the server will probably last them for a long time.  Cleveland is also considering some architectural changes to hopefully make it possible to store more virtual machines in less space (currently each takes 30 GB -- 25 for Windows 7 and 5 for patron use).

When patrons sign up for the service they have to take an hour-long class and agree to a waiver of rights.  The hour-long class has probably been something that has made patrons less eager to sign up for the service.  The waiver is something unusual for libraries.  Libraries (at least where I'm from) pride themselves on storing no data on anyone beyond what is absolutely necessary to run the library.  However in this case the library is, by the nature of the service, storing all sorts of files that the patron creates or downloads, some of which might be incriminating to a patron doing something illegal.  Because they are storing it they have to be able to comply with state and federal law if the data was subpoenaed, and because of that they have to make the patrons aware that they will turn over this data if asked to under court order.

It's a really interesting program because it pushes the boundaries of what libraries can do for their users at the same time that it opens up a kind of can of privacy issues for libraries and patrons.  It will be interesting to see what happens and if this idea catches on with other libraries and what the long-term consequences might be.

No comments: