For me, the data shows a lot of where I've been, but with only the sort of accuracy (and errors) you'd expect from tower triangulation.
It also misses some places that I frequent entirely (eg: it has almost no hits near my weekend house or some of the local places there my wife and I visit). Similarly shows no hits in the Pacific Northwest for me, but did seem to track my Vegas and Santa Barbara trips pretty well (though again with highly scattered accuracy).
Maybe it's just me, but I'm having trouble feeling very "violated" by this. If you were to look at my map, I think you could tell the area of the country where I live/work, but probably couldn't pick out my house or office with much accuracy (unless there are underlying timestamps the map doesn't show).
Yes all the points have timestamps and an accuracy estimate, and there is wifi fix data too that this script doesn't show.
There is often erronous data in there too - that's not this script, that's just that location fixes are inexact.
The real violation is that you just drag-dropped your whole iphone backup to a script written by people you don't know and running on servers you don't control... but don't worry, we don't care to profile you or sell you data... ;)
The real violation is that you just drag-dropped your whole iphone backup to a script
Yeah, I guess. I just did a quick strings on the backup files, and you could extract a fair bit of info, but it didn't seem like there was much in there that was truly juicy.
I'm growing increasingly convinced that the first iphone tracker desktop app reduced precision not so much to "protect privacy", but to cover up exactly how off the data was. I think selling it as though it were your location rather than tower location backs this up. Looking at it in Google Earth it seems to only be accurate when I was on I-95 in New Jersey.
I figured the data would be kind of rough, but for a month of my data in NYC it is just off. Apparently I spent most of my time in Bed Stuy and no time in my apartment in Park Slope. It also got Charlottesville, VA confused with Las Vegas. If that actually happened, there must be some great party stories that I haven't been told about yet.
Interestingly, that was my first thought too. I'm the guy who hacked together the javascript which passably reads the database file in-browser for this, and two things stood out to me as I was working: How much data there was, and how not-quite-that-accurate it was.
I think the initial app's way of presenting the data has merit because it does hide the swarm of identical points, but it also does make it 'feel' more accurate when you look at it because 'Yes! I live in that circle.' There's sure to be a happy middle ground somewhere.
What amuses me is that people complain that this information is being recorded and then proceed to upload the database to anonymous applications to visualise it, thus compounding the problem.
Sorry, I wasn't insinuating malicious intent. It's a cool app and I used it, it's much better than the openstreetmap version that was released!
My comment was more of a reflection of people's amusingly conflicting views. One thing this has made me realise though is that if Apple were to just ask their customers rather than operate with subterfuge then they could have avoided this whole thing being a problem.
What do these points actually represent? Mine shows a tremendous number of points in the regions where I've lived and been, but no points within about 1/4 mile of where I actually live and I have definitely used the GPS there. Also it shows some points near Las Vegas, which I'm pretty sure I haven't been to with iOS 4.
Some points even show up in the middle of the San Francisco Bay, which doesn't make sense if this is showing cell towers.
It doesn't seem to show everywhere I've been. I put the whole folder in, about 2k files, and it only showed points near San Francisco even thought I've travelled with my phone to the East Coast many times this year.
Is there a limit on your plotting points? If so, I'd do a geographic sampling to get a good representation of my phone's life.
For me, the data shows a lot of where I've been, but with only the sort of accuracy (and errors) you'd expect from tower triangulation.
It also misses some places that I frequent entirely (eg: it has almost no hits near my weekend house or some of the local places there my wife and I visit). Similarly shows no hits in the Pacific Northwest for me, but did seem to track my Vegas and Santa Barbara trips pretty well (though again with highly scattered accuracy).
Maybe it's just me, but I'm having trouble feeling very "violated" by this. If you were to look at my map, I think you could tell the area of the country where I live/work, but probably couldn't pick out my house or office with much accuracy (unless there are underlying timestamps the map doesn't show).