« Visualising the blogosphere | Main | Future of Web Apps retrospective »
22 February 2007
mashup* event, BT, London
Posted by Paul Miller at February 22, 2007 11:28 PM
As mentioned earlier this week, I headed back down to London this evening to attend mashup*.
With the title of “What's next, Web3.0? - The coming semantic web”, the event was highly relevant to our activities around the Talis Platform.
The event was chaired by Vecosys' Sam Sethi, and opened with presentations from Mark Birbeck, Talis podcast subject Paul Walsh, and Tony Fish, followed by an open discussion. I had to leave before we got to the post-discussion wine, but maybe next time!
Sam got the ball rolling by defining the Semantic Web (which Wikipedia describes thus, and W3C thus, demonstrating just why SWEO is so important!) as allowing people to “query the web like a database”. Not bad, for those who get what a database is.
He then embarked upon a useful illustrated tour of some microformat-consuming tools, describing microformats themselves as “making data re-usable” and suggesting that microformats are a practical instantiation of core Semantic Web principles. Highlighting microformats.org, and specific microformats such as hcard and hcalendar, he then showed browser extensions that read microformats, illustrating his points with Tails and WebCards specifically.
Sites such as upcoming.org and worldcupkickoff.com embed microformatted data within their markup; invisible to humans who just see a 'normal' web page, but very useful to microformat-aware browsers. Although we need browser extensions like Tails just now, there is a suggestion that Firefox 3 will include native microformat support this year.
Next up was Mark Birbeck, CEO of X-Port. Mark showed two products from his company that make use of some Semantic Web technologies from the W3C. formsPlayer leverages the XForms recommendation and is a;
“full-featured XForms processor that provides a complete implementation of the W3C's XForms specification, made available in today's browsers.”
The demonstration we saw suggested that it's similar in concept to Sam's microformat parsers, extracting value from RDFa markup in web pages.
Sidewinder was one I really didn't get; it
“is a new kind of desktop application framework. Like many frameworks it makes the task of producing desktop applications easier, but whilst other frameworks will use C#, Visual Basic or Java, Sidewinder uses high-level, standard languages, such as XHTML and XForms.”
Mark made much of its power to take applications in the cloud, and run them like desktop apps. Erm... why would I want to?
Next up was Paul Walsh of Segala, a fellow member of W3C's Semantic Web Education & Outreach group (SWEO). With only a few minutes on the stage, it's probably better to point you to his comprehensive podcast on the same subject... :-)
Finally, Tony Fish took the stage, to offer a (possibly too?) contrarian view.
He suggested (reasonably) that most users don't care about the detail, and simply want
“everything we want, where, when, and how we want it”
No argument there.
He went on to suggest that the big question is one of how we facilitate delivery of these requirements, and broke delivery down into three broad strands of;
- supply & demand
- create & consume
- input & output
The first, he suggests, is simple economics, and well understood.
The second is a (re-?) discovery that people are instinctively creators rather than passive consumers.
The third is where things went off the rails for me, ultimately arriving at an assertion that people want a sensory web, not a semantic web. I'm not sure I understood the argument, as I assume he meant something more than creating a web of content that we can smell, taste, and touch as well as (currently) see and hear?
In Q&A, Mark suggested;
“the data is what people are after [not the website offering access to it], and that causes problems for today's business models”
Too right, but it's fun making the incumbents wobble...
Andrew Hardie from DemSoc suggested that the Semantic Web has difficulty gaining traction because it
“shifts the effort to return ratio”
It's hard to add all the structure and metadata that many of today's Semantic Web applications require, the datasets are pretty small, and the use cases tend to be niche.
After some protracted pounding of Paul Walsh (they should have listened to the podcast first!), the conversation degenerated somewhat to the level of many Semantic Web discussions; I could swear there were people with microscopes out, counting the angels dancing on the head of their badge pins... A conversation that started well, with recognition that the end-user story was more important than any particular technical solution became remarkably caught up in itself, and in notions of what is/are (?) 'semantic', which was a shame.
Andrew Hardie made a good point toward the end, though, emphasising the role of the Semantic Web (whatever 'semantic' might be!) in moving from machine readable to machine understandable data, and in enabling the user to ask 'high quality questions'.
A good event, in a good venue (although you would have thought BT could provide free wifi in their own conference centre, even if they couldn't cope with selling a decent wifi connection to another conference venue), and I hope to make it next time, too. The more events like this that I attend, the more convinced I am that we're on the right track with the Platform; and a track where no one else has a better solution.
Posted via a GNER train's wifi network, too late at night, diverted off the main line and somewhere in Lincolnshire, heading home to my own bed after a week spent in a different hotel room every night...
Some concluding thoughts from FoWA may have to wait until the morning.
Update: David Lenehan also covers the event for Read/WriteWeb...
Technorati Tags: microformats, Podcasting, Semantic Web, Talis, Talking with Talis, TDN, Vecosys
Trackback Pings
TrackBack URL for this entry:
http://blogs.talis.com/mt/mt-tb.r280.cgi/719
Listed below are links to weblogs that reference mashup* event, BT, London:
» London mashup* Semantic Web or Web 3.0? from Segala
Thanks to an invitation from Vecosys, I gave a keynote presentation at last night’s mashup * event.
It was held at BT’s astounding conference centre, which possibly has the best conference facilities I’ve seen. It was fully equipped w... [Read More]
Tracked on February 23, 2007 01:15 PM
Comments
Thanks Paul. I really hope it was worth the long journey. Great post will link it to Vecosys tomorrow.
Posted by: Sam Sethi at February 23, 2007 02:30 AM
Thanks Paul. I've linked to you and ReadWrite/Web from my post.
Posted by: Paul Walsh at February 23, 2007 02:41 PM
Hi Paul,
You say:
Mark made much of its power to take applications in the cloud, and run them like desktop apps. Erm... why would I want to?
I think I made the mistake of assuming that this was obvious. The web apps conference earlier in the week had a number of discussions about bridging the gap between the desktop and the web, and there are umpteen conferences on 'rich internet applications' coming up that will be looking at the same subject matter. And so I foolishly took for granted that the audience at mashup* would think what I was describing was a good development. However, since you're certainly not the only person who has asked that question, I obviously judged wrongly. :(
Apologies for that, but the result is that I'm going to try to write a better introduction to what Sidewinder does, and why.
In the meantime, over in the comments section of a photo of me, there is a short discussion on Sidewinder and Apollo that might help a little. (The thread of comments is not about the photo of me you'll be pleased to hear! It's just that the person who took the photo at the mashup* event, also made a reference to Sidewinder.)
Also, I hope you don't mind if I link to a post of mine that has a screenshot of the final demo I did, which may be of interest since there's an explanation of the components.
Thanks for the mention.
Mark
Posted by: Mark Birbeck at February 26, 2007 12:45 AM
Mark
thanks for the comments, and I'll follow the links you offered.
I do - before following your links - totally get and accept the 'rich internet applications' space, and the bridging of the gap between the desktop and the cloud. However, I'd always perceived the value as being the other way around; having our remaining desktop applications draw, transparently, upon the capabilities of remote data and services rather than taking perfectly good web apps and chaining them to the desktop.
However, maybe that was what you were actually getting at, and I should follow those links before engaging in any more thinking-out-loud...!
Paul
Posted by: Paul Miller at February 26, 2007 08:08 AM
Hi Paul,
No...I really, really am getting at the fact that you can take a "perfectly good web app", and give it an additional interface. :)
To illustrate: you've spent months creating a web-based accounts package that rivals Sage (for example); with Sidewinder you can give it the final twist by allowing it to run just as if it was a desktop application. It's light years from being an 'either/or' proposition...we're saying that the exact same mark-up and code that you used to create your online version, now works for your offline one. Take a look at the number of web-based applications that have an additional downloadable, installable component--Flickr, Picasa, Google Desktop, Yahoo! toolbars, and on, and on--and you'll see that this isn't a 'maybe one day' feature, it's something that is needed now. But why fire up your C# compiler to produce a Flickr download manager? Why not just use the same HTML, script, XForms, or whatever that you used for the web version, and wrap it in a 'host' that gives you features that are closer to the operating system: you could use system tray messages instead of alert boxes; you could queue up images to be sent to the server if the machine was off-line; you could use batch files to post the contents of a certain directory each night. The key point here is that you don't need to use a different language to write your application, provided that your HTML/XForms/script gets some beefier features than are normally available in a browser.
And not only does this make it easier and quicker to build applications, but it also lifts the constraint on applications that they need to be either web-based (and so run in a browser) or desktop based (and so written in Java or C#). Many applications can be written with web technologies, but that doesn't mean they fit the 'browser' paradigm. I'm constantly pressing 'File > Save' on the browser menus when using Google Docs, for example!
All the best,
Mark
Posted by: Mark Birbeck at February 26, 2007 11:38 AM
I think Mark's SideWinder has an interesting tool but comes up against many RIA appls out there (incl Apollo). Several of these already have traction and some with large companies behind them
His saving grace is his Xforms standards based toolset. However I think he'd have more of a chance if it was open sourced and pushed as a viable alternative. He also needs to add on/off-line sync capability - without that u won't get the 'in the cloud' appls.
Lal
Posted by: Lal at March 30, 2007 06:07 PM
