Fake fake real-time blog from Open Publish 2007, Sydney

by Rick Jelliffe

This is a fake fake blog, because even though it is not real-time, I was actually at the Open Publish 2007, Sydney conference, unlike my real fake real-time blogs. They will be putting presentations online next week.

I chaired a full-day Symposium on Standard Open Document formats on Wednesday. Alistair Spiers has blogged on this. We had four sections.First I talked about the history of document technologies and standards, for context, looked at the modern ZIP-based formats, and at the similarities and differences between ODF and Open XML, especially looking at the differences in their goals from their standards.

Rather than being theoretical, we then spent most of the day with presentations from two developers talking about their actual experiences in integrating systems around Open XML and ODF. We certainly have arrived at the point where real implementation experience is available that trumps fact-free blather and FUD.

Jason Harrop is well-known from his SpeedLegal days, and showed a collabarative multi-version editing system built on custom XML (adding structures to the linear text) in Office 2007: he had a helper application in Word and the data was sent to a backed Java-based repository which handled shredding and storage issues. He didn't report any particular problems with formats, just that you need to be careful with namespaces. The system would be good for people collaborating on many kind of legal documents.

Dr Ian Barnes from Australian National University is working in a similar area to Dr Peter Sefton (they are mates, I gather) and his talk was concerned with his Digital Scholar's Workbench project, whch uses Open Office and ODF to pivot through DOCBOOK to various output formats: and various issues related to multi-publishing and making it appealing for data entry people to use stylesheets. Interestingly, the project grew out of student assignments, where changes in technology allowed a toy problem to inspire a practical result.

I finished off the afternoon with a walk through the standards process, and where Open XML was up to. At the end, we had a welcome lurking visit from Chandi Perera, who emphasized that at the commercial CIO level the issue of standards is entirely subservient to issues of ROI, execution etc.

I didn't attend any conference presentations, but I had reports that it was the best program yet. The Dr Raymond Wong paper on his new compressed, indexed, DOM/XML-friendly format caused quite a stir in particular. The paper will be up at the Open Publish site, but from what I understand from the drinks session (which I did attend) the format splits the document into a novel tree/index structure at its top, and the data content at the bottom: because the document is smaller than the raw XML it is suitable for transmission and because it can be used directly (it is not a compression per se) without decompression it can be loaded and used directly by DOMs. It sounded to good to be true, but Raymond answered the various questions I had, so it looks good enough to be true now!

I gave the closing keynote, on "The True Saga of Wikigate" and the audience laughed and cried at the right moments. There were many more questions than I anticipated, particularly about whether Wikipedia was reliable. I said that I thought it was excellent general, and that my experience of them was really constructive, but that if you see a page where only one side is represented (e.g. if there are many links to one side of the story only) you should beware.

In the final conference debriefing session, Nick Carr made an interesting point. He said that this was the first time in all the years this conference (and its predecessors, XML Open and SGML Open) has been running that there have been *no* presentations giving some trick or product or technology for trying to shoehorn Word into an XML or SGML production process. In previous years there have always been two or three, typically with some combination of VB and massage. It is quite remarkable how far we have come since 2004 where I reported of the same conference in
this blog If I were to pick a theme or meme, it was that the decision on whether and how to support Word was the by far the most critical decision for most large XML deployments.

The sister conference Open Standards 2007 will be held 15,17 November in Sydney, again in partnership with OASIS.