The Memory Leak

The launch of Google Maps with My Location yesterday has stirred up discussion of spatial privacy in the NY Times.

Privacy issues have long been discussed in other sectors, but only relatively recently in GeoData arena. Bob Blakley has been studying this for quite a while, and has promoted the concept of an “Identity Oracle”. While it has nothing to do with Oracle Corporation, private industry does play a key role.

The idea is that I should own and control my private information. Presumably this would include my location. I want to decide who does and does not have access to my location (and my children’s location).

Bob explains it this way:

The Identity Oracle is not a technology. It’s a business. Its business plan says “We allow people to enjoy the benefits of their identities while protecting them against the risks of misuse of their identities”. It charges money for its services.

While privacy issues have been discussed in context of Location Based Services (LBS), I haven’t seen any discussion of how Identity Oracle concepts might fit into the mix.

An Identity Oracle would not allow others to know where I am, but only enough information to provide the service I want from them.

Tell us, Identity Oracle, is Kirk near a gas station?

For example, maybe I want to allow Google to sell my information to advertisers, but I don’t want them to reveal my location. I want gas stations to only know that I’m near them - but not my precise coordinates.

This is not a simple concept, but one worth exploring. Google will soon likely be under greater scrutiny with respect to spatial privacy, perhaps we in the geospatial community should consider how an Identity Oracle might fit into LBS business plans.

All Points has pointed out an excellent in-depth writeup by CorpWatch describing the GEOINT conference held here in San Antonio.

Speaking of San Antonio and spatial privacy, take a look at what AT&T (also headquartered here) has been up to.

Bob Blakley points out a highly relevant article describing how AT&T gave the feds full access to massive amounts of internet traffic without any warrants.

Now that AT&T has rolled out remote monitoring, this potentially gives the government access to a lot more imagery. In effect the DHS has outsourced spying not just to corporations, but to private citizens as well.

Couple this with AT&T’s RFID program, and the spatial privacy issues become even more apparent. No doubt webcams tied to RFID readers can be programmed to follow certain RFID tags through space.

How would you like it if, for instance, one day you realized your underwear was reporting on your whereabouts?

— California State Senator Debra Bowen, at a 2003 hearing as quoted by wikipedia.

Even with all the spying tools though, AT&T decided to cut back on telecommuting anyway. Maybe they just don’t want to risk having the feds spy on their telecommuters communications. Better keep’em safe behind the firewall.

This copy of a Wall Street Journal article makes me wonder how much we in the civilian geospatial community are being co-opted:

James Devine, a senior adviser to the director of the Geological Survey, who is chairman of the committee now overseeing satellite-access requests, said traditional users of the spy-satellite data in the scientific community are concerned that their needs will be marginalized in favor of security concerns. Mr. Devine said DHS has promised him that won’t be the case, and also has promised to include a geological official on a new interagency executive oversight committee that will monitor the activities of the National Applications Office.

(Replace Uniface with ArcGIS.)

I’m excited to see that ESRI has hired Dirk Gorter as Director of Product Management.

I googled around a bit and see he was involved with Uniface at Compuware. I’ve never heard of it before, but it looks like a cool product. Note he does not have a GIS background. I think this is a good thing. ArcGIS has plenty of spatial capability, what it needs are easier ways to integrate it into the enterprise.

Looks like Walmart.com got more customers than it could handle.

Today, CNET quoted Walmart :

“Walmart.com said some of its most popular selling items on Thanksgiving Day were the Garmin Nuvi 650 portable global positioning system …”

I wonder how many of these people will use the POI loader.

OK, I’d like to think I’m not an ecofascist, but when I found litter in our front yard I began to think how RFID tags might allow me to track down the perpetrators.

Some day we will all be using credit cards for everything we purchase. Many of these purchases will have RFID tags. This means we could be held responsible for how we dispose of these products. Police could scan litter, find out who purchased it and send them a ticket. Even the blind judge from Alice’s Restaurant would have to accept it as evidence.

De-Evolution of GeoData market is needed to give us more freedom of choice.

Ed Parsons’ post comparing music and geodata has generated some interesting discussion.

I don’t have precise numbers, but still, I think it’s useful to compare the two markets in terms of Pareto distribution.

I would guess the top two vendors (artists?), Navteq and Teleatlas, account for 99% of the geodata sold to the consumer market for use with the GPS devices. Five years ago this percentage must have been much smaller. Clearly the Gini Index for geodata is rising. With so many GPS devices being sold and a growing awareness of geodata by the general public, can Web2.0 somehow help the market de-evolve so its tail becomes more proportional to its head?

Compare that to the music industry where the top 10 artists account for (maybe) 10% of all revenue. That’s just a guess. You could probably get a better estimate with data from RIAA. At any rate, I’d bet Gini index for the music market is falling, thanks to companies like Amazon.

The term “Long Tail” was coined to describe Amazon. As an industry we need to be thinking about how the market for GeoData can grow a long tail. As Ed points out:

Remember this is not always about making information free, it is about making it accessible.

I think it’s worth watching what Amazon does. Of the GAMY four horsemen (Google,Amazon,Microsoft,Yahoo) they are the only ones not leveraging their geodata assets. Amazon must be sitting on a goldmine of consumer spending geodata. Imagine the kinds of maps that could created by summarizing shipment locations for particular products. Google, Microsoft and Yahoo all seem to view geodata as a means to sell future advertising. Amazon doesn’t have an advertising based business model, so if/when they do get into geodata I think we can expect something different.

Perhaps their selection of WeoGeo as a finalist in their web services challenge is an indication. Perhaps they are planning a spatial equivalent to their popular MP3 store? Even though their MP3 store competes with iTunes, it’s not to say they find fault in the iTunes approach - isn’t Kindle basically an iPod for books?

Perhaps Amazon Web Services could offer a WMS service using data mined from their sales activity, then open that up so that third parties could publish their own geodata. To grow an even longer tail, Amazon needs to look at ways to reduce shipping costs. It seems like they could open up traditional bricks and mortar stores (or franchises) so that I could order something and pick it up at a location near my house, reducing shipping costs.

Then again, maybe the prospect of Christmas shopping has made me delusional.

WeoGeo, discussed in an earlier blog, has been named as a finalist in Amazon’s Startup Challenge.

I think they really deserve it. Wish them luck!

This guy did a really good job, it seems like something similar could be done to illustrate map projection concepts.

The geospatial community has blogged about Kindle (like High Earth Orbit, Henri Bergius and Peter Batty and AnyGeo)

I don’t see anyone asking about spatial awareness though. I think it would be great if I could download some sort of spatial documents to a device then query those documents using location. Doesn’t EV-DO support some sort of geolocation capability that amazon could tap into and let me browse documents relevant to my location?

Can’t make the blades any sharper? Then add more blades (Gillette “Fusion”).

Dave Bouwman has a good writeup on choosing new hardware. It’s interesting that he opted for the dual core instead of quad core.

I’d really be curious to see quad vs. dual core comparisons for ArcGIS Explorer (AGX). Even though AGX has “ArcGIS” in its name, it carries no legacy COM baggage it does not require extensions to be registered with the COM Interop like the rest of the ArcGIS family. Parallel processing seems a lot more doable in the pure .NET custom Task Framework compared to dealing with the COM interop. I’d be interested if anyone has written an asynchronous custom task that leverages multicore. Maybe a spherical version of Life running on a different core?

Then again, as more processing is off-loaded to the GPU, maybe more CPU cores become less relevant. Maybe how tightly integrated the GPU and CPU will become more crucial. AMD’s project (also called “Fusion”) seems to be betting on that with their plans to integrate graphics technology purchased with ATI:

AMD thinks that integrating the GPU will be essential around the end of the decade because so many applications–games and videos, for starters–will want to latch onto the GPU architecture and because the relative performance of a GPU is way beyond the CPU right now.

Remember how a revision of MapObjects morphed into ArcObjects when the value of COM became apparent?

Maybe the same thing could happen with AGX as the value of parallelization becomes apparent.

Google Earth seems to be winning the beauty contest. Maybe I’m missing something, but the Google Earth COM API would be just as difficult to parallelize as ArcObjects. COM is the bottleneck. AGX has no COM requires no COM Interop for extensibility, so wouldn’t it be better?