Cornelia's Weblog

Just read Dan Farber‘s post on structured blogging that in turn refers to some other posts on the subject.He and some of the other posts he references are addressing the concern that the overhead of adding the structure isn’t worth it for the blogger. His conclusion, BTW, is that giving up (even before it’s really even started) would be defeatist – I agree completely.But that specific discussion is not really the point of my post here. Rather, it is the structure part I am thinking about…I live in a world where information almost always has some structure applied to it – content management. Yeah, we manage unstructured content, however, we layer on structure in the management process. Absolutely no doubt that without that structure there would be very little management. To add to the standard capabilities of the web, render, download, rich text searching, … you need more structured information – no way around it (today’s web content has some structure, enough to support these operations). Blogging itself is an example of adding more structure to serve a particular need – when I finish this post you’ll be able to see the values of specialized fields both in the actual web rendering and even more so if you look at the rss. Clearly, I’m all for more structure to support more capabilities – as I said, services over content is the world I live in.Anyway, in iECM we are working on a formal model for interoperability of content management, we are addressing the information sharing problem. And we’ve had some very interesting conversations about the web being an information sharing infrastructure. How does the iECM information sharing infrastructure distinguish itself from the information sharing infrastructure of the web? It’s all about adding the right structure, and the right set of services to support the set of content management capabilities we need.

In the work that we are doing in iECM we keep coming back to a discussion around what the right term is for what we are creating. I’ve gotten a bit distracted for a few hours trying to find a consistent definition of the various “xyz architecture” terms (reference architecture, technical architecture, abstract architecture, etc.). Guess what. All of these terms are a bit overloaded (I know, surprising ).Take just one of these – Reference Architecture – a bit of investigation finds little consistency with how this term is used. First off, there are many different reasons for the creation of a reference architecture including providing a baseline against which developers can build a solution, providing deployment guidelines and even describing very specific solution. What is in them also varies wildly from being extremely high level (providing what my iECM colleague Stuart Williams of HP would call a conceptual model), to very detailed; some assign vendor products or standards, others remain entirely distinct from specific technologies.The conclusion I have come to is that any “xyz architecture” document should first define the term in the manner it is used addressing the purpose of the architecture as well as a characterization of contents or style thereof.For what we are doing in iECM, I particularly like the term “reference model”; what we are proposing is consistent with description given in the OASIS Reference Model for Service Oriented Architectures.

This Reference Model for Service Oriented Architectures is an abstract framework for understanding the significant entities and relationships between them within service oriented systems, and for the development of consistent standards or specifications supporting that environment. It is based on core unifying concepts of SOA and may be used by architects developing specific service oriented architectures or by those needing to explain SOA principles. A reference model is not directly tied to any standards, technologies or other concrete implementation details. It does seek to provide a common semantics that can be used unambiguously across and between different implementations.

I know it’s pretty old but I just saw it for the first time several months ago and found it delightful.  Tim Bray on WS-Opposition.

Okay, I’m stumped. I had a document that I received from a colleague and when I opened it the thing looked like this:The wierdness is the “1.” on the left of each line number. Other documents with line numbers didn’t show the “1.” so after banging my head against the wall for 3 hours (okay, I’m a little stubborn) I asked some colleagues to have a look. The “1.”s didn’t show up on their machines. Ugh. So I finally decide to ignore it for a little bit. Later (after a reboot) I reopen the doc and the mystery “1.”s are gone. Don’t get me wrong, I’m glad they’re gone but… ARGH!!