Notes from Aug 13th Geiode Student Meetings
Morning: talk with Claus Rinner re Argoomap
Initially Patrick gave a quick demo of the Okanagan Collaborative Conservation Program map (http://occp.bc.ca/) that UBC-O is developing. Based on Google Maps. Will allow embedded images, video in the discussions. There will also be icons for discussions about the map overall. It will be preloaded with polygons for the regional boundaries, that people can converse about.
Re: Argoomap:
Steven showed us a demo of the current software. Interesting point: a comment can reference other markers, and one message can have multiple points associated with it.
Some questions:
Nick: Is there a reporting feature? for flagging spam for review, etc...
Patrick: Is there a printing out feature?
Nick: Does the organization have an archiving mandate? That's an issue for the city of Toronto, it is obligated to archive comments for open houses, etc. ...and that’s hard to satisfy if they're trying to do engagement over facebook, for example.
Steven: In the interviews, people asked about mass import... that shouldn't be that hard to do geocoding, kml or shp import...
They didn’t want discussions in the sidebar initially but it was cumbersome to follow threads in text bubbles. Patrick: I found this important because I personally find it a lot easier to follow the discussions and find discussions in the sidebar – versus searching through all the bubbles and opening them etc.. – representation issue
Claus presented some of the origins of Argoomap (and argumentation maps in general)... from a white paper he’s working on. Began in urban planning in Germany, a digital version of the public comment process. Based on "argumentation" theory from computer science, but didn't get as far as automatic reasoning in the case of maps.
One case study: just published this month in EnvPlanB... Queen St. W community organization "active 18" (..no longer active.) A student distributed flyers: 16 ppl registered, 12 participated. They looked at how many threads were created, replied to. Looked at web log to monitor when people usually login, (weekends) and where from (home), summer least use. Lots of messages talked about places, but didn't link to the map. The questionaire responses all said they were experienced computer users, but didn't know how to zoom (because it used a GIS-style interface), so that motivated Claus et al. to go to Google Maps. [Alan: Is this the name change from Argumap to Argoomap?]
Another case study: a student looked for geographic mentions in a non-geo Ryerson discussion forum (about the Ryerson master plan?) then copied them into Argoomap. So you can look at structure not only in space, but within thread, but also across threads. Even sometimes links between places on the Ryerson campus and the U of T campus (people saying UofT's nicer, older)
Do messages need different “types”? "pro" or "con" messages... "new idea" message type?
Kawarthas case study... underlying map is sparse, so will discussion be sparse too? [Alan: in BC, spread linearly along highways & mountain valleys?]
So who is going to be using Argoomap? Live Green Toronto, for one. Claus: had a talk with Randy & Yolanda about using it for observations... so perhaps Argoomap is overkill for that? But they want people to be able to comment on observations, so maybe it will be appropriate.
Patrick: Would we be able to load existing base layers in a clickable way? The people we work with have all their own mapping programs... they'd love to have discussion on top. Our Okanagan Ecosystem map for example is using Map Guide
Claus: could our tool load these maps as a layer? perhaps WMS only...
Patrick: in our Map Guide we have already made a request to get the bike commuter routes from city of Vernon, for example... we’d have to make a new request to get that data, so if we could use it straight off the Map Guide site we wouldn’t have to make a new request.
Claus: We found that discussions based on vector data is challenging because we needed two different colors for highlighting... for whether there are messages attached, objects referred to... etc. So rendering and clicking lines and polygons becomes problematic... easier to keep discussions linked to point markers.
Now there are attempts to rewrite Argoomap from scratch in more simple, structured way. Want to and will develop further, want to keep simple as possible. Aaron (Sani?) is working on this version. Will be more object-oriented, purely Java Script based (instead of current version with uses PHP). The current version is also stuck on an older version of Google’s API and is too hard to change.
The Argoomap version on sourceforge is being worked on in Germany by the original developer. Not the same version that Steven is working on. Claus hopes to replace this with the new totally rewritten version in aug/sept... this would be the official default version.
Also, someone's forked a version at Western, too... allows flipping between a more analytical view.
Currently people are building off Argoomap from different versions – not well organized at this point. Claus suggested this is not necessarily a bad thing... as academics we should be focusing on case studies and usage issues, not developing a perfect software package that works for everybody.
Alan agrees: duplication of effort and reinventing the wheel is bad, but we also shouldn’t get bogged down in over-engineering a new solution if we don’t have to, either.
Suggestion: perhaps focus on separating content from software? If we develop several different software projects, at least aim for some standard database fields supported by all of them. This way no organization is tied down to a specific program (or version of a program)... they could dump the discussions and load them into a different program? This also could help us test how much of the success or failure of particular Geoweb projects depends on the software or on the community itself... So, just an XML dump of the discussion?
Claus: Some minimum set of database fields that each project can choose to support... references can be complicated. Can message ID occur with multiple marker IDs, vice versa... lookup table. So you can have one message with multiple markers, for example. Object IDs would have to link into kml, shapefile... On the social/qualitative side... it’s beneficial just to be able to preserve the content of the messages beyond the life of the tool.
Alan: who needs to be involved in this creation of a data model? UBC, MUN, UBC-O, Ryerson...
Alan guesses he’s going to be taking the lead on this, unless anyone else wants to.
Claus: doesn't think it's a bad thing if we have several different projects going... and keeping projects alive: on the one hand as academics we'll always be here to support... and there’s also the benefit of making things open source. At least communities don’t have to pay software licensing to keep the software running (although they still need to worry about having some person to administer things)
Patrick: communities that get $40K one time funding to set up and atlas, then in 2 years its obsolete..
Alan – referred to Mapstraction - http://www.mapstraction.com/. It’s like a layer between your code and the Google API. The end user won’t notice the difference, but you can easily switch from Google to Yahoo or Microsoft or Open Street Map if you need to, without rewriting all your code.
Some discussion of how this type of program may help use discussion tool over different map programs (ie cmn atlases or local gov map programs) versus importing data layers into the discussion tool. What software is necessary to include other base layers in the discussion? Do the actual vector-based objects need to be imported, or just a static map in the background?
Back to Student Meeting Ryerson 2009