Monday, April 06, 2009

GIS Meeting

I'm not marketing myself as a GIS expert, not primarily, and so know enough to recruit a current PSU student to fill me in on some of what's cooking.

We met at The Bagdad and went over some stuff, me explaining about the broken pipeline using the e | m | h | c motif, which looks cryptic, but just means elementary then middle then high school then college, or basically K-16. We see a break pre-college, when a lot of students feel heartbroken that they'll never "get math".

Our proposed fix is to use FOSS. GIS / GPS uses FOSS too (per FreeGIS etc.), so that's the bridge.

I learned more about SEEDS and Corpus Linguistics. There's both a physical and a mental geography, in the sense that zip code areas might be filled in with more DemocracyLab data.

I explained about the idea of deliberately spicing your rants with icons, changing the markup behind the scenes, because we want the search algorithms to take these rants into account i.e. we work with the grain of the design, to make the summary representations more cogent.

We're not summarizing in some distant board room, sharing exclusive secrets. The mirroring is right back to the ranting public, providing feedback for steering.

Anyway, the DemocracyLab stuff was more a sidelight. More to the point was the urban gardens, their interplay with the schools, institutional buyers. We want to improve the cafeteria food with Farmer's Market type produce, something Portland is already doing to a limited extent.

Even more to the point, the GIS systems need to show kids where the animals might be, including domesticated farm animals like sheep (a theme for today), and semi-wild animals like muskrats, herons, possums, coyotes, deer and raccoons, all native to Greater Portland.

Make that a Python list, then study some GIS maps, help with the data collection in some case. I saw both a heron and muskrat today, near Dignity Village.

I talked a lot about my visit to Immersive Media with my University of Rochester connection. If you get to practice your skills in some immersive world, like Second Life, then graduate to the real deal in some field trip up the Columbia Gorge, then maybe you'll be less likely to peel off from our math track. You'll think it's exciting and relevant and the GIS objects we give you (an API) will impress you with their power and ease of use.

Maybe you'll get to make mashups? Having mastery over a Controller (often a computer language) in the MVC design, is a key to future happiness in this regard. MIT's Media Lab's work on the Landman Report Card is a case in point.

Close to the surface in my thinking was Glenn's global matrix approach, which uses simple and elegant mnemonics to add more of a time dimension to these global data displays, i.e. he pushes global awareness back to those ancient civilizations we read about in Critical Path as well.

In my "road show" model, I'm able to segue to or from Glenn's segments pretty easily, the idea being to field a rotating cast with swappable members.

I look at Cubespace as a base for some of these trainings, Wanderers for others, e.g. I represented ISEPP at Pycon when doing my three hours, with segues to contributions from both Steve (PSF) and Ian (Tizard / Stanford).

In terms of fantasy worlds, I don't mind asking "where is Python Nation?" as like a question about Middle Earth. Now that Pycon has this Rivendale reputation, that's easy to do, plus the geography is actually quite well developed, thought out.

I alluded to Middle Earth in last year's Chicago talk as well, referring to Moria, a next stage on the journey after an abortive attempt to tackle an over-mountain pass -- metaphorically, Calculus Mountain, a chief obstacle on the older maths track, pre discrete math getting more of a footprint, thanks to FOSS and Portland's innovative math teaching subcultures (MLC a good example).