So, thinking about some recent posts ....
In the Bosworth presentation [ppt] mentioned below, the following desirable attributes of a 'web of data' are proposed: simple, standard, sloppy, and scalable. This is based on his view of what has made the web successful. Constrain choices and keep it simple.
As the library community re-engineers itself in a network world, we face an interesting challenge in finding the right balance between simple, sloppy approaches and the heavyweight, highly structured legacy approaches we are accustomed to. There has been a lot of discussion about the need for greater simplicity recently.
Additionally, the specialist, heavyweight library approaches typically have not travelled beyond a library niche. This reduces library impact. And this becomes extremely important now because we need to provide services at the point of need, in the workflow. Simple approaches will help diffuse library services into user environments.
So, things should be as simple as possible, but, as Einstein reminds us, no simpler.
Related to this is the comment that really leapt out at me in the Chudnov et al paper on OpenURLs and autodiscovery.
The concrete examples presented above, and the more speculative discussion following the examples, perhaps represent a simplistic view of how to implement information services. Even so, at this point in the development of digital library services, we have too many standards and service models finding too few users. In the particular case of extending autodiscovery to OpenURLs and metadata, the rapid uptake of RSS autodiscovery indicates a strong market demand for simplistic implementations of extremely useful service integration techniques. [Main Articles: 'Opening up OpenURLs with Autodiscovery', Ariadne Issue 43]If the suggestion here is that we need to focus on a few simple building blocks, then I agree. Interestingly, in my colleagues' work in OCLC Research, we are seeing a focus on OAI-PMH, SRU, and OpenURL as we look at service platforms. Work on schema tranformation and terminology services, for example, can leverage this substrate. And this approach lives nicely with RSS and URL-based services. Of course, this is to put to one side metadata and content standards for a moment.
Looking at this short list, and if these really are emerging core 'plumbing' approaches, then something becomes quite apparent. This is the real achievement of Herbert Van de Sompel in helping move us as a community towards simpler, more webby approaches.