Annual leave and conflicting activities have meant that the STEEV blog has been on a forced diet. I do have a couple of morcels of news which may assuage the hunger of this social media channel however…
In the middle of July we had a Skype call with Herbert Van de Sompel (Los Alamos National Laboratory) about how the Memento framework could be implemented in order to facilitate the spatio-temporal visualisation of energy variables.
As mentioned in the blog post ‘evolution of an interface’ a decision was made to use Output Areas as the spatial unit of analysis. It was our collective opinion that visualising the energy efficiency variables at house-level was too fine grained to offer any meaningful representation (even although the data supplied by the Welsh School of Architecture was provided at this level). This has implications in terms of the use of MasterMap as the large scale context mapping product* which was initially earmarked as the potential web-content upon which Memento could act. As it stands Herbert didn’t see any purposeful use case involving STEEV and Memento which would add to the functionality of the visualisation tool. Memento can be implemented (by linking to unique or presistent URIs generated by the tool) and we intend to portray this in the interface. This will not however showcase Memento’s temporal sweep through web-content (as highlighted on BBC/CNN content) as intended by its originators.
I parapharase Herbert (from correspondence) when he says:
Memento deals with web-time i.e. the state of a web resource at a given point in time. It allows one to use the HTTP URI of a web resource (say an HTML page, an image, even a database) and request a representation of the state that resource was in at the point in time that is of interest. This means, Memento looks at the Web as a “system” and can refer to and and look into the state of that system (and its constituent resources) at some point in the past.
This is a very different notion to looking at the state of some other “system”, e.g. a simulation of the real world, and checking in what state the real world was or will be at some point in “real world time” according to that simulation. In the case of STEEV, we are referring to a temporal state within the simulation. Not a temporal state of the Web.
Thus the temporal semantics involved in Memento and in STEEV are fundamentally different and that, as a result, applying Memento concepts to the latter is inappropriate.
A useful output from our conversation with Herbert however was the proposed creation of a STEEV RESTful API which would define a URI and its arguments for the determining and retrieval of a particular state. The URI arguments would represent the value of the model, each variable, the year, the map extends and the map zoom level. A mechanism would be added to the interface for saving its current state as a URI – perhaps to the clipboard, a popup or envoking an email client. Fetching the URL would return the client to the state when saved.
example URI:
http://steevsrv.edina.ac.uk/?model=sap&grid=1&renew=2&enery=3&year=1990&bbox=248354.304,203198.976,270931.968,225776.64&zoom=2
A meeting has been set up with Herbert here at EDINA for 19 September where we will discuss our implementation of the framework and other possible enhancements that have synergies with Memento or perhaps even Annotate (another utility developed at the Los Alamos National Laboratory).
* The upside of using Open Layers and Open Stream as context and navigational mapping utilities is that there should be no licencing restrictions on the use of the STEEV tool.