URL : www.walk2web.com
Description:
Walk2web is a link visualizer that draws a rather rudimentary link diagram dynamically as the user navigates (clicks on) an original node that is a user submitted URL. The intent of the site is to provide a diagrammatic means of traversing the web.
The link diagram does provide the advantage of providing an instantaneous view of the variety of different paths that branch of a particular node. The interface also includes a miniature site viewing window that displays a fully functional preview of the node.site that the user points to while navigating the map. In this window we have options to mark the site as ‘like’ or ‘dislike’ and to bookmark the site privately or publicly. The map window also provides a means of ‘flagging’ a ‘location’ by placing a little green flag on your map.
The site is fairly impressive in terms of design with the links maps drawn beautifully in flash and ably representing the network as far as the usual 2 dimensional schema allows. The map gradually reveals the ‘next step’ as the user navigates from one node to the next providing for the impression of walking through the web ‘landscape’. Unfortunately this is as much positive spin I can give for an interface and visualization that offers little in the way of expanding on the browser experience except in its ability to trace paths and thus remember the paths not yet explored. I suppose this might be understood as given the user a means of more completely exploring particular link arrays although the ‘specificity’ of the mapping undermines this utility.
The like/dislike categories do not alter the weighting of links – they appear only to publicly flag the fact that I have ‘ranked’ the site. This appears to have neglected a golden opportunity to provide the link map with a certain temporal dynamic as recent posts would generate more recent traffic allowing a design that incorporates a vision of a web currently emerging and an old web receding into the background. Here however the old links are as fresh as the new which means the overall experience is less than satisfactory. The other problem confounding this site is that it appears only/mostly concerned with mapping the domain name level. This means that the site suffers terribly from the movement to at once a more centralized, proprietary and temporally dynamic web. Today’s web is characterized by large domains with much activity ocuring within the directory structure of these domains. This is as much at the level of the individual blog level as it is at the ‘myspace’ scale. At the blog level a single domain carries what we could call a dynamic series, or perhaps more evocatively a ‘stream’, of posts that the domain merely represents as a location. At the ‘myspace’ level the Domain loses all specificity and thus navigational value because of the mass of streams of data the occur under that domain. As i have argued elsewhere this temporal dimension of the web has fundamentally altered its topology making the old metaphor largely redundant – this is no longer a web but an oceanic system of data flows, currents, and eddies. Walk2Web’s method exemplifies this shift as its focus on the domain rarely produces a usefully indication of page level ‘link flows’ and this means that we tend to end up wandering (within one or two steps) into large domains with no specificity (myspace, livetype, blogger, wordpress). These large domains then open rather randomly onto sites under that domain. Systems like Walk2Web need new ways of mapping or rather signifying/presenting these flows and for feeding back into the modulation or agitation of these flows in the service of realizing new potential. As I said the like/dislike or maybe integrating and visualizing traffic flows is a way of achieving this. The last.fm model uses a(n aural) version of both like/dislike and traffic measurement to provide for, and encourage, such emergent dynamism.