[OPEN-ILS-DEV] Question about Evergreen database

Sat Nov 24 00:17:19 EST 2007

--- Karen Collier <kcollier at kent.lib.md.us> wrote:

> Whether knowingly or not, you guys have hit on a pretty controversial
> topic in librarianship - whether or not to store data on what books
> patrons have checked out in the past... controversial at least in the
> US since the USA Patriot Act.
> 
> On one hand you've got the argument that you want to be able to give
> a patron a list of books they've checked out in the past when they
> ask for it, and they do ask.
> 
> On the other hand, some librarians oppose the USA Patriot Act on
> constitutional grounds related to patron privacy and for this reason,
> prefer to not retain this information, in case they are ordered to
> provide it.

That is precisely the reason why I asked the question.  I didn't want
to tip my hand initially, because I wanted to get an answer uncolored
by anyone's political predilections.

As a library patron, I do NOT want the government to be able to 
monitor my reading habits.  If I know, or even suspect, that the
government can see my borrowing history, it absolutely will have a
chilling effect on my reading.  There are certain books that I simply
wouldn't check out.  If I want badly enough to read them, I'll buy
them -- and I'll pay cash, so that the bookstore can't tie the 
purchase to the purchaser.

(Think I'm paranoid?  A year or two ago the FBI was monitoring the
sales records of grocery stores, trying to identify people who 
favored Middle Eastern cuisine as terrorism suspects.  I am not 
making this up.  I see no reason why the FBI would hesitate to 
monitor bookstores.)

Suppose I were a librarian, and suppose the FBI presented me with a
National Security Letter (NSL) demanding to know who had checked out
a copy of the Koran, or anything by Molly Ivins, in the last year.
I would like to be able to say that the information was not available,
because Evergreen didn't retain it.  Otherwise I would have to choose
between betraying my patrons and risking my own freedom.

Over the last few years the use of NSLs has increased more than a
hundredfold.  They are issued solely by the Executive branch, with
no effective oversight by Congress.  They don't require a warrant,
nor do they receive any form of judicial review, except when someone
is prosecuted for not complying.

Most NSLs are probably issued to telephone companies and Internet
service providers.  However at least one NSL is known to have been
issued to a library in Windsor, Connecticut.  We know about that
case only because the librarian was prosecuted for refusing.  Most
librarians are probably already aware of that case.  It's not 
hard to find it in Google.

We don't know about cases where the librarians complied, because 
everything is secret.

I don't trust this government not to abuse the powers it claims for
itself.  If you're a Republican, and you're willing to trust George,
would you trust Hillary?  I wouldn't.  Even if I would, I don't
trust this country always to elect governments that I can trust.

> With this in mind, it might make sense to have a configurable option
> where a branch/system/consortium could elect to save this type of
> information, erase it when a book is returned or when it has been
> checked out to another patron (as my library's current software
> does), or to let patrons opt in or out.
> 
> That much flexibility might overly complicate matters in development,
> but I doubt it would go unappreciated.

Personally I would rather Evergreen didn't store patron-level history
at all, but I think it's an excellent idea to make it configurable,
especially if each patron can opt in or opt out.  I have no idea
what it would take to implement such a thing, because I don't work
on those parts of the system.

> I am delighted to learn that information on when books were checked
> out is retained.  Our current ILS tosses that information out along
> with the patron specific information.  It would be very useful to be
> able to create a graph of circulations over time for an item or group
> of items, to see changes in popularity when you're deciding whether
> to weed an item from the collection or thinking about what types of
> books to add to the collection.

If I understand you correctly, it should be possible to support that
kind of analysis without retaining the identity of the patrons.

> Just some thoughts,
> Karen

Thanks to both of you for your replies.

I apologize for the long rant, but it's a subject that I feel
strongly about.  If fact freedom is more important to me even than
const-correctness.

Scott McKellar
http://home.swbell.net/mck9/ct/

> ----- Original Message -----
> From: "Brandon W. Uhlman" <brandon.uhlman at bclibrary.ca>
> To: open-ils-dev at list.georgialibraries.org, "Scott McKellar"
> <mck9 at swbell.net>
> Sent: Thursday, November 22, 2007 4:33:28 PM (GMT-0500)
> America/New_York
> Subject: Re: [OPEN-ILS-DEV] Question about Evergreen database
> 
> Hi, Scott.
> 
> I am not a EG developer yet, but as a humble sysadmin, I can take a  
> stab at your questions.
> 
> The short answer to all of them is 'yes'. When a book is circulated,
> a  
> record (or tuple, if you prefer) is created in the action.circulation
>  
> table of the Postgres data. The tuple includes:
> 
> - the borrower (actor.usr)
> - the item (asset.copy)
> - the time the item circulated
> - the staff person who circulated it
> - the location where it was circulated
> 
> The item type is determined from the circ_modifier in asset.copy, the
>  
> duration of the loan is determined in circ_item_config.js (in  
> /Open-ILS/src/javascript/backend/circ in the source, and in  
> $EG_HOME/var/circ in the compiled install) using with data in  
> config.rule_circ_duration, and the result is also stored as the due  
> date.
> 
> When the item is returned, the return location (actor.org_unit),  
> return staff person, and return time are all stored in the tuple as  
> well.
> 
> The database does retain this information for historical use. The  
> staff client does have some already-included functionality for
> viewing  
> this data, particularly for the scenario you mention in [2]. The  
> reporting interface would let (staff) answer the questions in both
> [1]  
> and [2].
> 
> It's worth noting, though, that some jurisdictions prohibit storing  
> this information under privacy legislation without the borrower's  
> consent, or, in some cases, at all. British Columbia, where our  
> installation is, is one of the former, so over time, we'll be  
> developing processes to vacuum this information from our database to 
> 
> comply.
> 
> Happy Turkey Day.
> 
> ~B
> 
> Quoting Scott McKellar <mck9 at swbell.net>:
> 
> > Not being a librarian, and knowing little about how libraries work,
> > I haven't paid a lot of attention to the OILS-specific parts of the
> > code base.  However I do have a question.
> >
> > Suppose I check out a book, and return it a week later.  Does
> > Evergreen remember that I checked it out at one time?  Or does it
> > just note that the book is back in circulation, available to be
> lent
> > to another patron?
> >
> > To address the issue a different way: is it possible (either
> through
> > existing tools or through customized ad hoc queries) to answer
> > questions such as the following:
> >
> > 1. Who checked out Moby Dick over the last year?
> >
> > 2. What books, or other items, did Suzie Grimshaw check out in
> March
> > of 2007?
> >
> > I am not asking so much about the existing tools, but about whether
> > the database retains this kind of patron-level usage history in the
> > first place.
> >
> > Scott McKellar
> > http://home.swbell.net/mck9/ct/
> >
> >
> 
> 
> 
> ======================================
> Brandon W. Uhlman, Systems Consultant
> Public Library Services Branch
> Ministry of Education
> Government of British Columbia
> 605 Robson Street, 5th Floor
> Vancouver, BC  V6B 5J3
> 
> Phone: (604) 660-2972
> E-mail: brandon.uhlman at gov.bc.ca
>          brandon.uhlman at bclibrary.ca
> 
> 
> 