I’m sitting in a quiet coffee shop on the mist-shrouded Oregon coast, taking a much needed break from family in the wee morning hours to put down some thoughts on the recent O’Reilly Open Source Conference in Portland. I’ll be heading back to Victoria over the next couple of days, nursing my poor, ailing Saturn back to the island and no doubt stopping in Seattle to indulge my daughter’s mania for all things Japanese anime related. She tells me that she’s a dedicated Otaku, regaling me with the plot-lines from half a dozen Japanese comics, many of which she’s now reading (more or less) in the original Japanese (”They always get the translations wrong, Dad!” she says with the conviction that only a fourteen year old teenager can have).

The conference itself was immensely enjoyable, and very eye-opening. I did get a chance to meet with Simon St. Laurent (an old friend and acquaintance that I haven’t seen in nearly a decade) and hung out with M. David Peterson, Kevin Farnham and James Turner, all of O’Reilly, spent some time talking business with Jason Gilmore and Terry O’Donnell, Managing Editors of Apress and DevX respectively, and sat in on some very good presentations (and hopefully gave a good one, though its always hard to tell when you’re on the stage side of the presentation).

Sorting through the blitz of sensory impressions from a busy conference like that is always a challenge, but a few observations did become obvious to me. The first was the XQuery (and somewhat secondarily) wass an intriguing phenomenon to a lot of people, but in general there’s comparatively little innovation being done in the OSS community around XML anymore. Many of the web-centric talks tended to revolve around AJAX frameworks (Django and the Google Web Toolkit (GWT) tended to compete for mind share), with Ruby and Python each taking up a fair amount of that pie as well. Curiously enough, Perl was still one of the dominant topics at the OSCON, though I’m not sure whether that was due to historical factors (O’Reilly of course sprang to prominence on the basis of Larry Wall’s Perl book, the quite famous “Camel Book”) or if Perl is experiencing a resurgence of sorts. Well received as well was a talk by the IBM’s Joe Gregario on AtomPub, the Atom Publishing Protocol formerly known as APP, which, as Joe pointed out, tended to be notoriously difficult to search for on Google. I hope to do a more expansive piece soon on AtomPub for either this blog or XML.com, using Joe’s presentation as a starting point.

Speaking of Google - the search giant was very much in evidence at the conference, and technologies such as Google Gears amd the aforementioned GWT were attracting quite a bit of attention. I think people are beginning to see the Google platform as being a viable one for developing applications beyond simple mashups; indeed, I found that the term Mashup was increasingly being met with grimaces - Google-centric developers are beginning to gain a sense of community, and the former Web 2.0 uber-term is increasingly being seen as quaint in light of more complex applications.

Microsoft used the conference’s timing to announce that it was submitting the Microsoft Permissive License (or MSPL) to the Open Source Initiative, the organization that handles the certification of open source licenses. The MSPL is based upon the BSD license, and is likely to be about the farthest that the Redmond company is likely to go into the OSS territory. The announcement was greeted with some skepticism at the conference (not unexpectedly) but I think most people I talked to were at least willing to take the announcement as a necessary first step for Microsoft. While wearing a Microsoft t-shirt at an OSCON party still pretty much guarantees that you won’t be asked to dance, I noticed that there was at least a tentative truce in the Microsoft bashing.

Back to XML for a few moments, one thing that jumped out at me was that the next areas for XML development will be occurring in the realm of search and relational management - what’s being increasingly tagged as the Semantic Web. I don’t think that this means that RDF is likely to see a huge surge (though it is becoming increasingly well known) but rather that people are beginning to realize that the manipulation of metadata (and meta-metadata, etc.) in a distributed world will become one of the most significant aspects of the web, and Semantic Web technologies will likely be at the forefront of that. Watch RDFa closely. DevX has started a new Semantic Web section on their site (which I’ll be contributing to) and O’Reilly is investigating that area as well.

A final note - after the conference proper, many of us went on a field trip to the Portland headquarters of FreeGeek, an “open-source” cooperative that both recycles computer parts and creates new working computers from old ones for sale at bargain basement prices (and that are available for free to people willing to volunteer a set period of time with FreeGeek learning how to build those computers and handling the sorting). It’s an innovative idea, one that I’ll be covering in far more depth in a subsequent blog, but I wanted to thank the good people at FreeGeek for giving us a thought-provoking tour of their facilities.

I enjoy conferences like OSCon because they serve to spark ideas and make people see what is not only happening but what is possible. I had a chance to moderate a “birds of a feather” session involving journalists and open source, which included a rather impromptu gathering of journalists and IT managers from the New York Times, the Washington Post, USA Today, the McClatchy/Gannett group and others. While there were obvious rivalries (and a rather condescending attitude toward bloggers that likely masks a profound nervousness) nonetheless it was intriguing to hear representatives from these companies talk candidly about the changes that have been going there, about the increasing power (and problems) that IT faces with respect to the traditional management infrastructure, the growing realization that the real value of such papers exist as much in their archives as they do in their day to day operations, and the distinct concern about both the integrity of the business and its future.

I should be back in Victoria soon, and will get back to my regularly scheduled sniping at targets large and small and digging away at the goings on in the XML community, but OSCON served (for me anyway) as a rather pleasant interlude that’s opened up my eyes to so much else that’s going on in the open source world.

Kurt Cagle is an author and information architect living in Victoria, British Columbia, Canada. He is also the standards architect for CASRAI and is the webmaster for XForms.org.