Panlibus Blog

Should interoperability mandate partnership?

Alejandro Garza over on the Stupendous Amazing Library blog, extrapolates the fact that there is very little partnership between library system vendors to conclude that they are not interested in interoperation between their systems.  He is picking up on extracts from the JISC/SCONUL Library Management Systems Study as commented upon by the Disruptive Library Technology Jester.

Coming from a history of integration protocols, in the library world, where they were more a framework for agreement than a standard, it is easy to assume that the only way to get two systems to talk is for their suppliers to establish a partnership to get it to work.  My least favourite standard NCIP is a classic in this regard. 

As I commented on the Jester’s post, the questions for the study were:

… in the present tense. Answering with ‘our products will integrate, etc., etc.’, would have no doubt drawn equal scepticism, but for different reasons.

The answers you picked out are symptomatic of an industry in transition. Transition from products without exception based on architectures that never envisioned light-weight loosely-coupled integration. Transition to a REST based service oriented architecture where integration between library and non-library applications should be simple and based on simple and open standards.

The “Do you have partnerships with other LMS/ERM vendors?” question in the survey demonstrates an attachment to traditional thinking towards integration. So far, with the traditional heavy-weight protocols we are used to in the library world, the only reliable way to get integration that works has been through a partnership between suppliers. Web 2.0 has demonstrated that with simple light-weight protocols, integration is possible without the need for commercial partnerships. There are many benefits that arise from partnerships, but they shouldn’t be a prerequisite for successful integration.

It is not all doom and gloom though. Initiatives such as the DLF’s ILS API defining simple REST base protocols that all vendors should be able to support, have started to gather momentum in the last few months. A momentum that appears to be supported both by vendors and open source groups.

Since I made that comment I attended a JISC and SCONUL Library Management Systems Study Consultation Event in London.  This event was a get together of stakeholders in the UK academic library community, which were joined by representatives from system vendors for the afternoon session.  For those with a sadistic streak in must have made an entertaining spectacle, watching six vendor representatives (Ex Libris, Infor, Innovative, OCLC, SirsiDynix & Talis) trying to squeeze their views in to 5 minute slots.  From most of those presentations and the discussion that followed, it is clear that the vendors are just as much stakeholders in this as the rest of the community.

I feel there is a refreshing openness in opinion and approach that is starting to spread through the conversations in the world of library systems.   This openness has been in high evidence in the recent Library 2.0 Gang conversations on ILS APIs and Bolt-on OPACs

It was a good meeting in London, I only hope that the organisers can keep the momentum going and build a community around the concerns of all the stakeholders, vendors included.  If the initiative started by the study falls back in to the traditional model of projects and reports that we are used to, it will be a massive waste of an opportunity.

Back to my original question – do we need partnerships to enable interoperability?  No we don’t.  With loosely-coupled integration, facilitated by web native light-weight open APIs, interoperability should ‘just happen’.  Vendors should, and are starting to be in the position to, say my systems are open for you to interoperate with – who ever you are, partnership in place or not.  This won’t happen over night, but we are already on a new path, with a healthy does of credit for the DLF’s leadership in giving us some direction.

Photo from Flickr by Just.Luc.

Technorati Tags: , , , , , , , , , , ,

2 Responses

  1. Alejandro Garza Says:

    I think you are right that a simple API opens a system to external mining. So I guess, yes, interoperating does not require the stuff of “parterships” (IMO meeting, signing documents, holding press conferences, etc.) So I agree there with you, I guess I will post a followup =)

    I still think that, however, it would be better to work together (partner) towards open standards (even if it means having 3 or 4–better than 10 or more).

    My thinking is that the JISC study was more about the willingness to cooperate, not the mere fact that the system can talk X language.

    Heck, I have HTTPd and screenscraped my way to “interoperability” between Innovative and dozens of things, including Drupal. See http://drupal.org/project/millennium

Leave a Reply