MacOS is still #3

Despite the halo effect from the iPod and the widespread use of Mac notebooks among speakers, MacOS still hasn’t managed to make much of a dent in the visitor logs:

1. Windows: 80.70%
2. Linux: 9.57%
3. MacOS: 9.44%

If MacOS can’t beat Linux on the desktop, I don’t know if it has a bright future.

Internet Explorer below 50%

Firefox is still #2 behind MSIE, but for this crowd, the gap is small:

1. MSIE: 49.61%
2. Firefox: 41.14%
3. Safari: 3.50%
4. Mozilla: 3.22%
5. Opera: 1.76%

If you’re designing or maintaining a web site with a tech audience, you’d better be testing on Firefox as well as MSIE.

Screen resolution and colour depth

I know that web designers like big layouts, but the sad fact remains that 1024×768 is still the most common resolution (and remember that the browser window may be much smaller than the screen):

1. 1024×768: 28.32%
2. 1280×1024: 25.84%
3. 1280×800: 10.61%

A long tail of resolutions follows, but it’s worth noting that the classic 800×600 has only 1.96%. Better news comes from colour depth, where almost everyone has 16bpp or better:

1. 32bpp: 80.29%
2. 24bpp: 11.89%
3. 16bpp: 7.37%

Traffic

Search engines, referrers, and direct access were all important traffic sources:

1. Search engines: 36.77%
2. Referring sites: 34.97%
3. Direct traffic: 28.22%

Blogs did show up among the referring sites, but the biggest traffic producers were traditional links from partner organizations (other conferences, IDEAlliance itself, etc.) — these were also the stickiest, since most people coming from these links went on to read more than one page.

As far as search engines go, I was surprised to find that nothing really matters but Google (assuming that Google Analytics isn’t biasing the numbers):

1. Google: 94.16%
2. Yahoo!: 3.46%
3. Live: 1.51%
4. MSN: 0.45%

I knew that Yahoo! and MSN were behind in search, but I had no idea just how bad it was (at least in the tech crowd). More than half of the people who found the site via a search engine went on to read more than one page.

The top search phrases were rather dull and predictable:

1. “xml 2007”: 28.50%
2. “xml conference”: 8.22%
3. “xml conference 2007”: 3.20%
4. “xml conferences” 3.04%

And so on through a very long tail. Individual speakers’ names start appearing soon, but none with more than 10 searches. I trolled through the low-frequency search phrases for something funny (and maybe risque), but all I came up with was the number “736”, which resulted in three visits. I gave up trying to find the site in the Google results for that number. Does anyone really search for a single three-digit integer, and if so, how many pages of results will that person scroll through?

Cold

It was cold this time, colder than XML 2006 last December in Boston. I brought suitable clothing with me, so I was still able to get outside for some long walks around the city, but with wind chills dropping well below -10degC, a lot of visitors probably weren’t equipped and had to spend most of the conference indoors. That’s not as bad as it sounds, since the hotel is connected to, literally, kilometers of indoor shopping, with pedestrian bridges across major streets. Unfortunately, they probably didn’t make it the 2.5 blocks outside to have banana or cranberry pancakes for breakfast at Charlie’s Sandwich Shoppe.

We also had snow — only a dusting of about three inches in Boston the night before the conference started, but since Boston’s not a snow city like Buffalo, Syracuse, Ottawa, Montreal, etc., I wasn’t sure if they’d have the roads clear and people would make it to the opening plenary. Fortunately, while many schools closed outside of Boston, the downtown kept moving as usual. We had over 200 people in a packed ballroom, and there was good audience participation.

Evenings

We had excellent 45- and 90-minute presentations during the day, of course, but some of the conference’s greatest energy was generated in the informal evening sessions. John Boyer‘s XForms evening on Monday, and Ken Holman‘s Standards and Specs Lightning Rounds on Tuesday both played to packed rooms with noisy audiences.

Both featured rapid-fire sequences of short presentations (15 minutes each for Monday, six minutes and twenty seconds each on Tuesday) giving the people there a chance to hear lots of different speakers and ideas in a short period of time. If you weren’t there on Tuesday night, you’ll find it hard to believe that an evening of standards talks could be exciting, but it was.

Consideration

XML is no longer mainly about a circle of people who know each-other and meet at conferences and committee meetings a few times a year. With so many new people, the change sometimes comes out as a lack of consideration — for example, speakers who wouldn’t bother showing up to hear the other presentation in their sessions (also leaving the session chair wondering if there would be a second speaker) — but overall, I think that people were back to being a bit more considerate than I’ve seen in the recent past. I saw fewer session chairs and speakers sitting in front of the room checking their e-mail while someone else was giving a presentation, had less trouble with people talking loudly in the hall outside while presentations were in session, and audiences tended to be friendly and supportive. Almost every presentation had a decent-sized audience ready to ask questions, and with all the work the speakers put in, they deserved at least that.

I experienced an example of consideration above and beyond any call of duty, when one person took time out from a serious family crisis to track me down, call me in my hotel room, and let me know about a change affecting the conference. You know who you are — thanks, and our thoughts are with you.

Boston

I love Boston. It’s a lot like Toronto, where I lived for six happy years and first learned how to love big cities properly (from near but not right in downtown, moving around on foot or public transit, and dealing with small stores and local merchants who get to know your name), but with its own special treats, like the antiquated trains (?) on the Green Line, a decent subway system that actually goes to the airport, truly fanatical baseball fans, and the Charles River. I’ve made a good number of visits to Boston over the past two years getting these conferences together, and I’ve come to feel very comfortable in the city.

Next year, I leave the conference to new people, and the conference leaves Boston for Arlington, VA. I hope that my life brings me back in contact with both the conference and Boston in the future, and I wish the best of luck (and stamina) to next year’s organizers.

Track chairs

• Documents and Publishing: Lisa Bos, Really Strategies
• XML on the Web: Simon St. Laurent, O’Reilly Media
• Enterprise XML: Richard Salz, IBM
• XML Training: Priscilla Walmsley, Datypic

Special events coordinators

• XForms evening: John Boyer, IBM
• Standards and specs Lightning rounds: Ken Holman, Crane Softwrights

Session chairs

• Simon St. Laurent, O’Reilly Media
• Debbie Lapeyre, Mulberry Technologies
• Tommie Usdin, Mulberry Technologies
• Norman Walsh, Sun Microsystems
• David Orchard, BEA Systems
• Eliot Kimber, Really Strategies
• Priscilla Walmsley, Datypic
• Rich Salz, IBM
• Michael Day, Yes Logic
• Anthony Coates, Miley Watts LLP
• Lisa Bos, Really Strategies
• John Boyer, IBM
• Robin Lafontaine, DeltaXML

Reviewers

• Mark Baker, Coactus Consulting
• Ronald Bourret, rpbourret.com
• Anthony Coates, Miley Watts LLP
• Gary Cornelius, CSW Group Ltd.
• John Cowan, Google
• Bob DuCharme, Innodata Isogen
• Edd Dumbill, Useful Information Company
• John Evdemon, Microsoft
• Peter Flynn, Silmaril Consultants
• Lars Marius Garshol, Bouvet
• Betty Harvey, ECC
• Ken Holman, Crane Softwrights
• Marcel Jemio, US Treasury/FMS
• Debbie Lapeyre, Mulberry Technologies
• Ken Laskey, MITRE
• Michael Leventhal, Tarari
• Eve Maler, Sun Microsystems
• James Mason, Y-12 National Security Complex
• Charles Myers, Adobe Systems
• Paul Prescod, Kinzin
• Liam Quin, W3C
• Zarella Rendon, PTC
• Michael Rys, Microsoft
• Eric Severson, Flatirons Solutions
• Ed Stevenson, Really Strategies
• Jeni Tennison, Jeni Tennison Consulting Ltd
• Henry S Thompson, University of Edinburgh
• Tommie Usdin, Mulberry Technologies, Inc.
• Eric van der Vlist, Dyomedea
• Claus von Riegen, SAP
• Norman Walsh, Sun Microsystems
• Philip Ward, Ford
• Jabin White, Silverchair
• Lauren Wood, Sun Microsystems

XML and British television reruns

An excellent illustration of that point is Matthew Browning’s and Robin Doran’s presentation, BBC iPlayer Content production: The Evolution of an XML Tool-Chain. Matthew and Robin talk about how, once the team stopped thinking about their information in relational database terms and started thinking in XML terms (RelaxNG, to be specific), it became a lot easier to manage changes and improvements to the project, which lets viewers watch shows they’ve missed from the last seven days of BBC television. It worked so well, in fact, that two other projects became redundant.

XML in practice

That’s not to say that it was necessarily an easy transition, but in the BBC’s case, at least, the rewards far outstripped the cost. That’s precisely the point of our 2007 conference them, “XML in Practice” — we’ve selected many papers that emphasize what did and didn’t work in real, large-scale projects, rather than focusing on new specs and prototypes, to help people make the same kinds of decisions in their own projects. XML won’t always make sense — in some cases, it will never make sense — but learning from others’ experience is a much better guide than speculation.

Cracking the bank with a white hat

Mark had the enviable opportunity to be part of a team of white-hat crackers, trying to break the security of web services at an unnamed bank — fortunately, he’s been allowed to share his findings with us. After listing vulnerabilities ranging from data smuggling in CDATA sections to SQL injection, he mentions that “the bank’s attempt to apply preventative security measures, such as SSL and XML Schema validation, actually proved to provide a false sense of security, and in fact introduced a number of security vulnerabilities of their own.”

This means you too, REST!

And lest the Restafarians get too smug, he found lots of vulnerabilities in REST as well as WS-*. Beyond Mark’s talk, it’s also worth noting that apps using OpenSocial, which is REST-based, have already been cracked twice, so no one in charge of protecting a net-based API (REST or WS-*) should breathe too easy.

See you in Boston.

XForms evening (Monday 3 December 2007)

7:30: Seeing is Believing: Intuitive Visual XForms Design

John Boyer, IBM Canada

XForms offers an order of magnitude simplification to the design and development of business applications.

7:45: The Pure Declarative Approach: XForms in Real Estate Forms Case Study

Dan McCreary, Dan McCreary & Associates

The declarative power of XForms empowers business units to maintain their own applications without IT involvement, using graphical specification capture.

8:00: Creating a Custom Editor for Everything

Doug Tidwell, IBM

Use XForms to create a custom editor for an XML vocabulary. The key to this magic is a set of XML configuration files.

8:15: XForms and the eXist XML database: a perfect couple

Erik Bruchez, Orbeon

XForms speaks XML natively, and so does the open source eXist XML
database. In this talk, we show how they form a particularly
attractive combination.

8:30: XForms, XHTML, and RDFa for Internet-Facing Applications

Mark Birbeck, x-port.net Ltd., W3C Invited Expert

Combine XForms, XHTML, and RDFa to build and test widgets, gadgets, and applications.

8:45: Composition and Choreography of Web Components in XForms

Charles Wiecha, IBM Research

Leveraging the MVC design of XForms, Web 2.0 applications can be designed as reusable components loosely coupled using XAC and SCXML.

9:00: Keynote: How XForms Can Win

Elliotte Rusty Harold, Dept. of Computer Science, Polytechnic University

XForms: will it be a dream, or a dud? In this keynote address to the XForms community, Elliotte Rusty Harold offers his vision and advice on the future of XForms. (30 minutes)

Background

While XML is used a lot on the web — for syndication, knowledge representation, open APIs, etc. — it certainly hasn’t developed the way we had planned back in 1997–98. XML Namespaces are often misused or simply ignored, other general web-related XML specs like XLink are barely noticed, and the payload in AJAX is often JSON rather than XML. Will there ever be a generic machine-readable web of information to parallel the human-readable web of documents? If so, will that web use plain old XML, RDF, or a non-XML format?

Think up some of your own questions, and we’ll see you in Boston from 3–5 December.

Call for participation

I’m happy to announce that the Call for proposals for the XML 2007 lightning rounds is open until November 16 (we fill slots on a first-come/first-served basis, though, so it’s better to submit early).

Our theme for this year is Public standards and (widely-implemented) specifications related to structured markup. We prefer presentations from groups or individuals involved in their development, but others can propose talks as well.

The lightning rounds are outside the regular conference program and open to the public.

What’s different?

Here’s what’s different about lightning rounds:

• They’re very fast: each presentation consists of exactly 20 slides, shown for 20 seconds each (at which point we cut off the speaker). You can learn about a lot of standards and specs in a very short time. There’s no time for audience questions.

• They’re free and open to the public: anyone in Boston on 4 December is welcome to drop in, even if you’re not attending the conference, and you don’t have to register for the conference to give one.

• They’re in the evening, and some of the audience might have been out wining and dining before hand.

• We encourage the audience to cheer, heckle, and otherwise act in ways unbecoming a regular tech conference presentation.

Act now

Michael Smith from Opera has kindly agreed to run the evening, and it will be a lot of fun. So if you have a standard or spec to promote, please get your proposal in ASAP to make sure that you get on the programme!

What’s hot in markup in late 2007?

I haven’t done any formal analysis yet, but from a first read through the proposals, the hottest topic this year is REST — even when it’s not the main point of a talk, people seem to go out of their way to work it into the title or the description, and it’s starting to spread out of the Web track into Publishing and Enterprise. WS-* is getting less and less common, but we still have some strong proposals in the area, and will almost certainly include a few on the conference programme.

Other frequent topics include microformats, DITA, ODF and OOXML (which is even in the non-tech press right now), metadata, XSLT, and XQuery.

Of course, there’s also a long tail, so we will be able to schedule interesting presentations on a wide variety of topics, including up-and-comers like Atom and non-XML markup like JSON.

Changes for 2007

There are a few changes for the conference this year:

• There will be no separate tutorial day; instead, the regular program (three tracks) ends at noon on Wednesday, then Wednesday afternoon will be devoted to a special training track for all different skill levels (no separate registration required).

• This is the only call for participation; there will be no late-breaking call this year.

• On one of the evenings, we plan to offer lightning sessions for standards groups — each group will have 20 slides and 6 minutes and 40 seconds to let us know what they’re working on. This will be a great way to learn a lot about a lot of specs and standards in a short time.

• We continue to encourage presentations on open data and document technologies other than XML, such as JSON.

We look forward to seeing you in Boston this December.

Recipe for pickling a web site

The original site was a hand-rolled LAMP implementation, but it was designed from the start to be amenable to a static copy. To pickle it, I started by doing a recursive slurp of the live site using wget (with the -m option) — that generated permanent, static HTML copies of the dynamic, database-driven pages on the site. At that point, I had an almost, but not quite perfect static copy of the site, because there were two things that wget missed:

1. Images referred to only in CSS stylesheets (such as the banner).
2. CSS stylesheets referred to by other CSS stylesheets.

It took only a few minutes to add all of that by hand, and the site was ready to go.

Why it worked

This will be old news to a lot of people reading, but a few simple advance steps (during site design) made later static preservation easy. Here’s what I did:

• Every page has its own URL, period, end of discussion. No AJAX, no POST.
• Every page (or at least, every page that we want to archive) is reachable, directly or indirectly, from the home page.
• Script names are not shown to the public, so there are no URLs ending in “php” (hint: exposed script extensions like “php”, “asp”, or “jsp” are signs of gross incompetence in web design).
• No web pages rely on exposed GET request parameters: for example, the URLs looked like /programme/presentations/123.html, not /programme/presentation?code=123, or even worse, /show-presentation.php?code=123.

And that’s it. Of course, if the site had included live forms, I would have had to remove those as well (and any links to them), but that wouldn’t have been much extra work.

On a final note, while the live site was hosted on an Apache server (the “A” in “LAMP”), the pickled site is hosted on a Microsoft IIS server. It made no difference at all — that’s the way Web standards are supposed to work.