Trip Report WWW2005 3rd day (friday, may 13)

querying the past, present and future: where we are and where we will be

this was a very interesting panel about the past, present and future of the Web. each panelist gave his or her impression about what the Web was, where we are now and where we may be in the future. some statements were rather controversial. one panelist basically said, what we have today is quite satisfactory, another panelist had the vision, that the Web will fade into the background like databases did in the past. the Web will remain the underlying infrastructure, but Web applications will emerge and become the "thing" that we will deal with. this sounded very appealing to me.

the panel was chaired by ling liu. during her introduction, she ask: "how do we query the web today ?" her answer was: by searching, browsing and navigating. but we use the Web not only when we sit at the desk in our office or at home, we use the Web while we are on the move. therefore we want a device spanning experience.
further, she asked what query models do we use to access information and query the web ? there is a passive, ad-hoc approach and there is an active, event-driven approach. by active and event driven, she means that we are look up the next train or check the movies they play at the cinema next door. she calls these activities "present". but there are also future related activities, such as looking up the games our favorite team will be playing next season.
ling said, the past is the history of the present web, so there is a need to archive information in order to preserve the past. but how do we query the Web in the future ? and can the Web be used as a common infrastructure for querying the past, the present and the future ?

first panelist: do we need a time machine ? my answer is YES! to query the past, just crawl the Web repeatedly and buy lots of disks. since not all documents were created at the same time, when we follow hyperlinks, we constantly move forward and backward in time. one way to handle the future is notification. we can subscribe to events or matches, so we will know, when something happens. Google alerts does it today, so there is no need for new protocols. however, crawling and monitoring requires some further research.

second panelist: conventionally written media is usually static and fixed in time, whereas the Web content is continually changing. however, many websites are abandoned because maintenance requires (too) much effort. so there is a great number of "decayed" documents on the Web. humans can often rather easy detect decayed documents based on indications such as outdated data (e.g. reference to old dates or events way back in the past), machines have much more difficulties to detect old information. they did a formal decay measure and found that from 1000 pages 475 were already dead after 4 months. search engines remove inexistent pages from the index, but they hold dead pages basically infinitively.

third panelist: for life science scientist - as probably for all scientists - it is very important to be able to know how they got to a particular piece of information and that the information will be preserved. one way to do this is to record browsing activities so they can replay and reproduce what they have done. furthermore, they invented what they call "LifeScience ID" (LSID). an LSID records the history of an object and what is the relationship between objects. this information is described in RDF, they have basically created "a shadow web". in the future, they want to have event-busses, a means to transport what they call "e-Science" events.

fourth panelist: he showed a number of interactive web services which are already available today, but what he thinks will the common user experience on the Web in the future. there won't be no more queries, there will be action and interaction. the Web will fade into the background as the infrastructure - like databases did maybe a decade ago.

to the top

interaction and the Web: the future browser

even if the title suggested that they might show a future browser, this was not really the case. so from that point of view, i was a bit disappointed. nevertheless, the information about future enhancements to various parts of the Web were very interesting anyway.

(click on the images for an enlargement)

past, present and future of CSS layout by bert bos: CSS level 1 - which basically supported margin and float - became a W3C recommendation in 1996, but was not reliable before 2000. CSS level 2 - which added support for tables and positioning - is the current W3C standard but is right now being revised. it may become reliable by the end of 2005. CSS level 3 - which will treat a document as a grid similar to the layout of a newspaper - will support alignment as in tables and order-independence as in positioning. it will be possible to do the layout with a graphical tool and there will be support for several grids for different devices. CSS level 3 will not ready before 2007.
see slides for details ...

the sematic browser: improving the user experience by steven pemberton: XHTML2 is the next iteration in the HTML family. it is as generic XML as possible; with less presentation and more structure in it. this will provide more usability, more accessibility and better internationalization. in addition, it is more device independent, will provide better forms and will require less scripting. finally, there is much better support for semantics. the goal was to keep the old community happy and to keep the new community happy. the relationship to the semantic Web comes from the support to integrate RDF in XHTML2, which means you don't have to learn RDF to be able to benefit from it.
it seems to me XHTML2 is a great next step. it is fully backward compatible with HTML V4.0 while providing exciting new features to support the semantic Web by providing native support for RDF expressions.
see slides for details ...

Web applications in XML by tv raman: the anatomy of a Web application is to deploy user interaction to a universal client. XForms will significantly decrease the cost of Web applications. they will help to design user interfaces for cross-devices and multimodal access. XForms will deliver well-formed XML to the server and enable end-user programming of Web front-ends.
see slides for details ...

welcome back browser by dean jackson: with the end of the first browser war between Netscape and Internet Explorer, the innovation somewhat eased off. the advent of Firefox seems to bring back some enthusiasm in developing new browsers and extending existing ones. while some of the common web browsers still do not support SVG natively, Tiny SVG is very successful with mobile devices.
see slides for details ...

presentations of the W3C ...

to the top

closing ceremony:

the closing ceremony was split into two parts. first, they presented awards for best paper, best presentation, best poster and best student poster.

(click on the images for an enlargement)

then, dr. leslie carr, presented next year's conference (WWW2006), which will take place in edinburgh, scotland, UK (not england !) from may 22 until may 26, 2005. he invited us to bring our families with us and to plan next year's vacation in scotland. in order to support this, they will even provide child care during the conference ! he announced other changes such as there will be no tutorial/workshop day and no developers' day. instead, there will be a five days conference with tutorials, workshops and activities for developers during the whole conference. i consider this good news, since i think this conference series can really need some changes. the presentation of WWW2006 was very invitingly and i hope, i can be there ...
leslie remembered the audience, that the Web was born in 1989 at CERN in switzerland and that on a global scale, scotland is just around the corner ...

(click on the images for an enlargement)

to the top

to the trip report main page