Things I went to at Xtech…
XAML and Avalon – (microsoft)
Building with XUL – Ben Goodger (google)
Just fucking do it Principle to public data on the web – Tom Loosemore (BBC)
All XML Databases are equal – John Snelson
Connection social content service with RDF, FOAF and REST – Leigh Doods
RSS syndication for a worldwide audience – Ian Forrester (myself)
Structure and Chaos, wikis, xml and structured authoring – Paul Prescott (Blast Radius)
Achieving Interoperability between RDF and Topic Maps – Lars Marius Garshol (W3C RDFTM WG)
Simple Worldwide Aggregation Using XSLT – Ken Holman
XHTML2: Accessible, Usable, Device Independent, and Semantic – Steven Pemberton (W3C)
Bridging XHTML, XML and RDF with GRDDL – Dominique Hazaël-Massieux
Beagle: Free and Open Desktop Search – Jon Trowbridge
BBC News and RSS, Or: How We Learned To Stop Worrying And Love Open Data Services – Joel Chippindale and Kevin Hinde
Mozilla's Birds of a feather QandA – Mike Shaver and some Mozilla developers
Dynamic Topic Maps using Web Service Interfaces – Eric Freese (IBM)
The Oracle XSLT Virtual Machine (XVM) – Anguel Novoselsky (Oracle)
Managing Complex Document Generation through Pipelining – Jeni Tennison
Streaming XML with Jabber/XMPP – Peter Saint-Andre
So first up, generally it was an excellent conference (maybe the best I have been to). I have never known so many friendly people in the industry. The mood through-out was pretty cool and encoraged a quite flat structure where even the Gold speakers and sponsers were around chatting with people. I didnt even know there was speakers lounge till the last day, and I found no one was ever in there. Speaker became listeners and everyone made up the audience. Just what you need in a 3 day conference.
The BBC was out at the conference and made up 4 of the many talks at the conference. But the BBC was 2nd behind the huge array of Mozilla guys, who I believe used Xtech as a way of first time meeting each other. There was a birds of a feather session on the Thursday which I attended and then they disappeared leaving only Mozilla gear (I need to take a picture of my Firefox T-shirt really). According this blog they all went on a boat trip while most of the BBC guys spent time in the Japenese resturant with Edd Dumbill and a couple others. Great food, good company lots of flames.
Anyhow from the start… Xtech really did cover so many parts of the XML world and covered them very well. Some of the highlights of the first day for myself included the battle of the xml user interfaces, mozilla's xul vs microsoft's xaml vs Laszlo. I only stayed for the xul and xaml talks.
The Avalon/Xaml talk was interesting but i had already seen pretty much the same presentation on channel9 a month earlier. Anyhow it was useful to see xaml upclose and to ask questions. I also never knew the difference between Avalon express which runs in the IE browser and normal desktop Avalon (like a application/widget on the desktop). Theres also BAML which is the binary version of XAML. Someone asked the question of cross platform compatiblity, and Rob Relyea turned it back on the xtech crowd suggesting microsoft are very open to communties porting avalon to other platforms besides windows. Someone also asked about the seperations of concerns in regards to CSS and XAML, Rob made it clear CSS would need to be upgraded to be taken advantage of. But it would be possible to convert CSS to something XAML could use.
The xul questions were actually equally harsh but Ben Goodger from google had no problem putting his hands up and taking the criticism. What was interesting for myself was firefox 1.5 which once installed will have everything someone would need to run and develop a xul application. A trojan for xul adoption maybe? To add to this xul push, xul runner is in development. This will be the xul, gecko and chrome engines in a simple excutable program, which once run only needs the xul application parts to run. Best way of thinking about is like flash/shockwave player standalone. This reaslisticly means a mozilla browser is not needed to take advantage of xul. Goodger was also keen to point out the cross platform nature of xul. Linux, Windows will have xul runner first with mac support coming a little later (there was a little unrest on this last point). The general view is to make XUL applications as easy to write and deploy as extensions in Firefox right now.
Lunch was surved in te main hall around the booths of companies displaying there wares. This was a good idea but it all became a little too crowded sometimes. But finding somewhere to sit was not a problem at all. On the same point, wireless was excellent throughout the whole event. Plus there was enough electrical points for people to juice there batteries. I took a english 3 way adapter which was useful when others came along. I swear the mix of people on laptops and people not must have been almost half and half in some rooms. Trying to quote Matt Biddulph talking with Edd Dumbill one night,
sometimes it works sometimes it does not, in Xtech it did. Honestly most of the talks were so interesting I bearly had time to write notes.
After lunch, Tom Loosemore's presentation titled Applying the JFDI (just fucking do it) Principle to Public Data on the Web. Was great as expected and well received throughout the open data room. Tom is well established figure in the bbc and public data fields. Some of the best known projects include theyworkforyou.com, upmystreet and faxyourmp. This was followed by All XML Databases are equal by John Snelson which outlined all the different factors which need to be considered when thinking about XML storage. Quite useful as the SQL/XML Databases are now starting to get a lot of the coverage. It was also good to see Exist DB on the comparision list, not just the Oracle's and SQL Servers. Its a shame I missed Gavin Bell's presentation about bringing a open european constitution in xml and the ROME presentation, which I may start using for aggregation instead of the almost dead Flock. Leigh Dodds talk titled Connecting Social Content Services with FOAF, RDF and REST was great to hear. I had been following hints on his blog for some time now and I was pretty pleased to finally hear some of the very interesting results. I wont try and summarise Dodds but it was very good and raised tons of questions and thoughts about consistency and ease of use in the case of REST webservices.
The last talk of the day, was my own which I wont talk about here, because I was meant to write about it here instead.
Other sessions worth noting I went to.
Structure and Chaos, wikis, xml and structured authoring by Paul Prescott of Blast Radius, was useful as it compared the very loose structure of wikis against the rigid structure in most xml based content management systems. Which leads me to the Xwiki presentation which I missed but had a personal re-presentation later on the last day. Basicly the guys behind the open source project have built a scripting engine inline and allows for structured user input and webservice integration. Its all pretty much sumed up here and here. Anyway, it all leads me to consider using Xwiki for personal and maybe professional projects.
XHTML2: Accessible, Usable, Device Independent, and Semantic – Steven Pemberton from W3C. The room was packed full while Pemberton discussed XHTML 2.0 and the thoughts behind it. Besides the usual, more useability, better accessability, better independence, less scripting, better internationalisation, better semantics, less presentation and more structure while making the world a better place. There was talk about some of the new elements and attributes included. < h > and < section > allows for unlimited levels of structure simular to OPML. < hr > has changed to < seperator >. Paragraphs finally can include lists and other inline elements inside of its self. Images still exist but the src=”” attribute can be applied to almost element, aka allowing for another level of structure which was not possible with < image >. RDF/A was talked about and is RDF with attributes, allowing everyone to get involved in the semantic web without having to learn RDF. property and rel attributes can be applyed anywhere. Think of it like the class attribute but without the style thinking behind it. This in turns makes a complete
RDF triple without the HTML community learning RDF. Interestingly enough, Pemberton mentioned RSS and joked that there is no need for RSS when XHTML 2.0 comes around. He showed an example but didnt quite make it clear except to say there were already 7 versions, why not add enough?
Bridging XHTML, XML and RDF with GRDDL Glearning Resources Descriptions from Dialects of languages) by Dominique Hazaël-Massieux was quite mind blowing and had links to XHTML 2.0. I do not know where to start really. It seems to be a standard way to extract RDF semantics from XHTML and XML documents using XSLT. Which is very interesting when you consider RDF/A in XHTML 2.0.
Beagle: Free and Open Desktop Search by Jon Trowbridge was fantastic to see and talk about. It was a shame his laptop died the day before because the only demos he could show us were the flash movies. But Jon really went into detail about the relationship between him and Novell, the push for free and open desktop search and its rivials.
Managing Complex Document Generation through Pipelining by Jeni Tennison, was very good and gave me lots of extra ammo for pipelining as much more than just a concept. I actually caught up with Jeni at lunch time and asked her more about her presentation. She said she has used Cocoon quite a bit in the past but said I should check out Daisy too – which happens to be one of the presentations I missed earlier on Wednesday. And last of all talks worth mentiioning that I attended has to be Streaming XML with Jabber/XMPP for the same reasons as Jeni's Talk. Lots of ideas and thoughts and a general feeling that these technologies are not as far off as I'd been tricked into thinking recently. See for example in the case of pipelining and jabber, I have known about these and the advantages for a long time, but have not had the drive to push them into my work life, maybe that will change.
Some of the presentations which I also heard were good or interesting include.
The Future of XML at W3C – Community Participation by Liam Quin, which I didnt want to miss but had to for the Jabber presentation. Are Server-Side Implementations the Future of XForms?, Rich Web: SVG And Canvas In Mozilla which I heard had a display of a fully useable google running at 30 degrees using SVG foriegn object. Printing XML: Why CSS is better than XSL caused a massive disagreement I heard, Can OpenOffice be the new XML schema IDE? which I really wished was on another day and the same applies to Comparing XSLT and XQuery.
On a whole Xtech was well worth the time and I'm so what glad I picked it over WWW 2005. I'm sure to be there next year which might involve a change of venue but it will still be in Amsterdam hints Edd. Thanks to everyone who made Xtech 2005 such a good conference and that also includes speakers and the people who paid to go.