Tuesday, April 21, 2009

GIS 2009

GIS in Action 2009

:: gis in action 2009 @ hilton, vancouver, wa ::

I'm at a Hilton in Vancouver, WA. Unlike at Pycon, almost no one is using a laptop. My 10:30 - 12:00 talk went very well, everything worked, room full, audience laughing at my jokes, asking intelligent questions, no complaints.

I still need to hover more in the booth area, which splits exhibits between town and gown i.e. academic posters and private company booths safely cohabit, plus some of the private booths are university departments. This is a blend we might be shooting for at Pycons, thanks to Vern Ceder's new position. Currently we're pretty much all business.

I spoke at this seventeen year old conference a few years ago doing an uphill slog with lots of source code. Today's talk was far more autobiographical, winding through dad's Libyan sojourn, visiting our 8th grade class in Italy that time (About Zoning), the high school math teaching interlude, some mention of McGraw-Hill.

Given the 2005 lunch talk was about Civil War era submarines, I had a good excuse to segue to my uncle Bill Lightfoot's work, on submarines pre-WW1 (there's a slide with the book cover).

I was trying to explain my relationship with PPS, per local politics. I said nothing about Alaska, or Bhutan for that matter (given the lunch topic, the Mongolian Altai by Jim Meacham more stories from the Himalayas would have been appropriate, with links to Shirley MacLaine).

Fresh from Pycon 2009 (Chicago), I had lots of news. Then I connected the usual dots between New Math (e.g. Venn Diagrams) and Gnu Math (e.g. SQL), the latter being our campaign to run more FOSS on Intel and AMD chips, not just have calculators for brains in math labs. That's a well honed rap by this time, lots of heads nodding.

I went into Google Streets with Immersive dodecacams, place based education, PY3K. Ian's work with the ATM in the UK got some focus, as did some South African pilots. I told my Y2K story about watching the BBC in Lesotho, USAers freaking out about their superpowers (there'd been some apocalyptic literature, a crisis mindset).

Talking so much about my dad and urban / regional planning, was my way of reaching out to this other culture. Later, dad became an education planner, which is sort of where I pick up, given all my curriculum writing experiments, mom likewise an influence (she wrote quite a few textbooks during their stint in Thimphu).

It's OK to see prehistoric dino Pythons as a basket of hissy snakes, happy to work for ya for many years to come (no pressure to migrate 2to3 if you have no economic reason to do so). Unlike cars and trucks, bit-perfect Pythons don't wear out or break down.

I ended with a brief summary of CSN planning, given that's getting my focus these days.

When Boost goes 3K, that'll be a big milestone (a clue from one of my astute listeners -- these folks are pretty Python aware, thanks to ESRI). Will there be a Google Summer of Code focused on that project?

GPS-Photo Link with TerraSync looks cool. It's not a live demo yet, just screen shots (wait, here we go...). Demos are risky, but worth doing.

Although I rotated my cube between exhibits a lot, I didn't do much in live Python beyond disassemble a function (dis module), explaining what byte codes are all about, in the context of narrating about the VM summit, which I followed from a distance.

I steered a couple of folks to http://www.flickr.com/photos/17157315@N00/3321022118/ as where to get some real Python overview, access to trainings.

I have Alan Potkin's Water in Mainland Southeast Asia PDF for Bob Pool, who expressed interest in learning how to do more with less. Perhaps ReportLab would help customers of GIS data get it all in one highly usable CD / DVD.

Adobe's PDF format has many under-exploited features. Patrick's idea of just distributing websites (per recent traffic with Synovate) is also state of the art i.e. browsers are the "universal client" these days. Microsoft was prescient in seeing them as replacing / augmenting the GUI.

People loved the spinning Compiz cube. I'm clearly an accomplished showman, had a whole table of artifacts, actually poured beans between polyhedra (Saturday Academy hat).

This wasn't a carbon copy of the IEEE talk at Portland Center Stage, but had some commonalities. Having Trevor touch base this morning, regarding some manuscript he's writing, got us off to a good start. He's my source for the "other tomorrow" meme, a feature in today's slides (3.x meg).

Hybrid SQL ~SQL applications appear likely in this domain as well (not just in the medical field). Come to think of it, Anatomy & Physiology is really just another kind of geography (interior mappings), as Fantastic Voyage reminds us.

Some data is amenable to row and column treatment, other documents might come in through mapreduce. The HttpResponse object is their confluence (a stream of suitcases). The browser doesn't care where these came from, so long as they appear secure (what that means is up to your firewall).

I was glad to finally meet Tim Welch who works with GeoDjango for Ecotrust. We chatted at the OSGIS booth, yakked about FoodHub some. I'll hope to make a future user group meeting. The PSU booth heard my rant on the Red Cross web site, something I've been sharing with Cubespace engineers.