I think that's a great point, and I agree that it's not the ideal metric for illuminating the issue. However, I'd argue that, to a limited degree, even the "harmless equipment" acquired by Law Enforcement agencies from the Department of Defense serve to "Militarize" the police forces in a way that reinforces conflict rather than cooperation.
If you let the cops dress up like soldiers, train like soldiers, and equip them like soldiers to the extent that we currently do in the US, eventually they start to believe they are in a war.
But yes, I do agree that the ratio of seemingly mundane items to the big ticket MRAP's and APC's is skewed toward stuff you can get at REI or OfficeMax, which sort of muffles the message. There's interestingness and insight buried in this 1033 program data, I believe. This probably isn't quite it, yet.
Sure, if they're wearing BDUs, army-style helmets, etc., it can be reinforcing a culture that isn't appropriate to civic policing. Maybe some of that stuff is part of this program, but even that's still buried among the flashlights and first-aid kits.
Perhaps you could add to the dataset tags that identify whether an item is e.g. "weaponry/armor", "military regalia", "general equipment", etc. and make these available as map filters. Or, even more basically, a simple indication of whether the item is also available for civilian purchase.
Some stuff might be a bit ambiguous too. For example, the entire acquisition for the county I live in consisted of four "utility trucks". Are those MRAPs, or are they just cheap vans that might as easily be used by the animal control department as by the police?
Thanks! Considering I probably couldn't even spell choropleth a month or 2 ago, I'm pretty grateful for Mike Bostock and his proclivity for documentation (for D3 and mapping concepts in general) by creating working examples. http://bl.ocks.org/mbostock is my favorite thing on the internets.
The data was sourced by using ogr2ogr to grab the data from an arcGIS server that the city runs and convert it to geojson. There was also an associated Sales db that I was able to query all results for, so I grabbed those and stuck them in MongoDB for the purpose of linking the parcel data with the property number in the db. When a user clicks on a property, I search mongo for the associated records and update the html via socket.io.
I recently was showing an acquaintance of mine a demo of something I was playing around with and a potential business model. He responded with "I wouldn't be that interested but if you could use those same capabilities to deliver X, now that is something I would pay for."
When I went and looked at your demo I thought here could be someone who is already walking down the road toward being able to deliver a critical component of X.
If you want more details lets switch over to email. You can find one of my contact emails in my HN profile.
I showed this last week, but the app crashed and burned after about an hour. I'm pretty sure I figured out the issue(s) and fixed them, but I'll be interested to find out if that's really the case.
http://www.mlive.com/news/index.ssf/2017/04/mackinac_bridge_...