[Koha-devel] Social Engineering, was: How to gather better popularity data?

[Galen Charlton]

Hi,

On Thu, May 26, 2011 at 4:42 PM, Breeding, Marshall
<marshall.breeding at vanderbilt.edu> wrote:
> I believe that libraries have vital interests in having users find them on the Web.  I can
> think of nothing more damaging to libraries than insisting that they obscure their Web
> presence by restricting the pathways that lead users to their Web sites and catalogs.
> It is also in the interest of persons who work in libraries to know the automation
>  systems used by their peers so that they can make well-informed decisions
> regarding technology strategies.

And of course, part of that decision-making includes evaluating
providers of services for the technology.  This is particularly
important for open source ILSs.  For most proprietary ILSs, you can
get support only from its creator and possibly some country-specific
or library-type-specific resellers.  There is a much broader ecosystem
of support for open source ILSs, so if a library has decided to adopt
Koha, they then have a choice of deciding whether to use a provider's
services, and if so, which one.  Knowing who is using what provider
(and *not* just relying on the references supplied by the provider)
gives libraries more information with which to make their procurement
decisions.  In my opinion, the benefits of that information *alone*
outweigh any theoretical security risk.

If there's a major security glitch in Koha, it would affect lots of
Koha users, irrespective of whoever is supporting or hosting them.  If
there's a major security glitch that is peculiar to a particular
provider's implementation of Koha?  Well, frankly that is information
that should be known broadly within the library community, not
obscured.

As I'm sure is the case with most regular users of lib-web-cats, there
is some data that I wish was recorded that isn't, and some data that I
wish was recorded differently.  But I do consider it a very valuable
resource and I, for one, thank Marshall for all of the effort he's put
into it over the years.

> An anonymous Popularity Contest daemon such as used in Debian would not necessarily provide
> data that would help libraries considering or running Koha to find peer sites for comparison.  It
> would also not be a reliable indicator of number of libraries that actually use Koha.  If it's optional,
>  then the numbers would be under-reported.  It would also include the large number of Koha
>  instances that are used for development and evaluation in addition to those that actually
> run in production in libraries.  It would also not include the libraries that use Koha within a
> restricted intranet, or in local networks that have not access to the Internet, which is common
> in the developing world.  Such a technical approach is helpful with an OS where you want
> to measure overall deployment; it's different for automation software where you care
> more about what libraries use it in production.  (This is in response to the IRC comment
>  that I should have compared lib-web-cats to popcon and not the wiki.)

Well, it's not an either-or situation.  A popcon-like system would
provide data that would be of immense use to Koha developers.  Very
few libraries are likely to update their lib-web-cats entry every time
they do a minor release upgrade, for example, but a popcon would let
us know to a rough degree how frequently those upgrades do occur, what
platforms seem to be the most commonly used, and so forth.  I view a
popcon as a useful complement to directories like the wiki and
lib-web-cats.

Regards,

Galen
-- 
Galen Charlton
gmcharlt at gmail.com