8. My notes prior- & post-Urban Tapestries trial
11/12/03

Venue

Victoria House, Vernon Place, Bloomsbury, London
12-2pm

Questions

think of the devices that could visualize the network/ontology
- how are the objects & spaces related?
- is this made explicit to the human-user?

walking between objects
- how is this handled by the system?
- are human-users static or mobile or both?
- how is movement recorded? [think of the degree of precision in masaki fujihata's work. how does it relate to this?]
- how could walking reveal the system's ontology
consider the density of information in specfic place of trial
- how would this work with different densities and areas?

-----------

feedback

conversations with rachel baker, alice angus and pete gomes.
proboscis have a very thorough design methodology - they have developed particular techniques such as 'bodystorming' to visualize the system to participants at early stages of the project - and have based their practice in some form of participatory design schema.
evidence of this process is pushed to the fore in this project - a scale model of the 'bodystorming' map is installed on a table in the ut workspace/reception area.
following a 45min induction into the use of the system, we were let loose into bloomsbury.

the trial

rb and i conducted the ut trial together - and soon ran into the limitations of the client device and infrastructure.
the client proved very slow - prohibitively so - the text input in particular being in lag of our need to make immediate notation of our location.
our expectation was also of a device that could work-out where it was in the world. we often had to guess where our location was - the map in the GUI being without street names. we projected ourselves into the system - our avatar being a circular 'target' icon - only to find that it was always in lag of our actual movement. we had to simulate the labour of our walking in dragging the icon around, as if behind us.

the core concepts of the system are the 'pocket' and the 'thread'. pockets appear at locations on the map and threads as lines between pockets.
the action of following 'threads' is called 'drifting' - which intimates situationist practices central to the psychogeographic project - that of sensing the borders of areas of similar character and of points of flow from one to another. 'drifting' in this case is more deterministic, as our overview map has already sensed the presence of 'Pockets' in our locality and draws lines directly between those in a specific thread. these threads are in effect desire-lines and do not find an analogue in our bodily experience of our journey between locations. this disjuncture is really marked - the system seems to want me to ignore, or find less-significant, my movement between locations. it prefers me to be stationary - a static user. in a sense, the knowledge that i can upload to the system is sedentary there is something of a paradox - which may in turn be all the more interesting - as we are using mobile devices to access what is actually sedentary knowledge. although an accretion , it is non-the-less fixed to a location. knowledge arranged in this way should be familiar to us - libraries and universities embody this: and as metaphor, so does the story of the 'memory palace', an architectural metaphor used to aid memory recall.
nowhere is there 'ambulant' knowledge - knowledge neither here nor there. between locations - more or less smearing one place onto another, in greater or lesser degrees. this knowledge is caught on the hop - it stumbles up a curb, it falls into a rut. what it doesn't do is assert itself in one location.
neither is it aimless drift - it does not sniff-out a localities ambience as a pass-time (sic).
driven by a desire to make sense, to give it time, to come to terms, it is always somewhere in negotiation.
extending this idea of the 'becoming in love' as the process of one ontology opening to another - rachel reckons that this is an important tactic of resistance at a personal level and shouldn't be revealed to, let alone expected from, the RDF project. this is entirely correct of course - and my intension is to point-out the inadequacy of this technology in this respect.
rb also talked about ideas for a conference, or collaborative mapping/sEmantic web exercise. hard outcomes are needed it seems. ie. some use-value will have to be found in the exercise, preferably applicable to more than the closed group of participants

with pete gomes

p had just also finished the ut trial and we headed out for a coffee nearby. pete was a participant in the 'locative media lab' in riga earlier this year - feb 2003 i think - and has an inside-track on things.
i asked if the ut project could be considered within the widEr lm field - it seems that many of the other proposed projects draw on RDF as a standard and hence have their emphasis on ontology and subsequent potential to link to the wider semantic web project.
in my view, the spacenamespace project in particular hits the mark in this respect, conscious as it is of the process of classification of spaces that has to be performed. i've yet to meet jo walsh, but i need to pursue the why, how of this study in classification.
the semantic element would appear to operate as salve to the obvious problem of trying to define an unequivocal world-view just from the construction of an object-hierarchy. even if synonyms can be found across ontologies, the challenge will be to keep-up with the full-range of dodgy practices that will change the meanings of those spaces and open them to constant re-interpretation.
[i can foresee a moment when someone will think that it's less bother to crack-down on possible re-interpretation of space than develop an inter-ontological, semantic web.]

p and i walked over to the Architecture Association to check pete's parkbenchTV site -- he's about to get his new plank , etched with wi-fi details, inserted into the seat of the bench.

I bought a copy of 'The AA Files' #44 which includes p's piece on location specific media and free networks.
- posted by spope @ 11:51 am

 
Feedback notes from Urban Tapestries blog
i came to the UT trial with a partiularly loaded set of questions to ask of it: i'm interested in how location-based information systems such as this accomodate the mobile user - that is, me, walking and talking with rachel baker - with information tagged to specific, static locations.

i use an ipaq as part of my everyday note-taking activities and am also very aware of their potential resale value. for this reason i rarely use one *openly* in public, just in case someone takes it for a next-generation phone.
so using an unfamiliar (and slow - ok you know this...) software application on this device in the street was unnerving to some extent, but we persisted nonetheless.

we took jean-paul's advice and headed north over bloomsbury square to look for less densely pocketed locations. Somewhere along the way my machine 'fell-over' - i tried to log-in but failed. this was the first time that i was aware of the network infrastructure on which the system depends and i really *wished* that the ipaq could tell me whether there was any 802.11 in the ether. (i guess it would have, but it was difficult to drop from one app to another through the the 'start' menu as i would usually have expected to do.)

being made so aware of the network infrastructure made me think of the continuing issues of ownership of wireless networks - this will continue to be an issue for projects such as UT. from what i can gather, it's 'access' that is the angle being pushed - this could as easily mean access to corporate/telco provision as it could local council/public service provision. either way, there will be implications for the conditions-of-use of any public authoring system using this technology.

back to my main, and loaded question of the mobile user: ok it seems obvious, we are carrying a mobile *device*, but we have a system (possibly any/every location-based system?) that privileges *static* locations above the movement between them. The user is asked to annotate single locations (and these will be classified into a strict object-hierarchy with RDF, or so i understand) - the demand is made for sedentary knowledge to be recorded, rather than ambulant knolwedge that finds itself between, on the way towards - neither here nor there..

following form this, i also felt some mis-match between the way that 'threads' are represented in the GUI and the actual bodily experience of moving through the physical environment when using the system. the labour of walking is almost entirely erased from the system. So our being diverted around fenced-off areas of park, stumbled down curbs, dodged traffic or got progressively more hungry didn't get noted until now. All that's left is a node and an arc on map.

A few comments on the project overall : i think there's some really good work being done through the design methodology that's been adopted/developed.
the participatory design methods - in particular the 'bodystorming' device, look as though they are really valuable in communicating both the concept of the project and in generating actual 'pockets' of information.

[http://diablo.proboscis.org.uk/MT/mt-comments.cgi?entry_id=23http://diablo.proboscis.org.uk/MT/mt-comments.cgi?entry_id=23]
- posted by spope @ 11:13 am