tag:blogger.com,1999:blog-8470995059384390571.post6540592282478410804..comments2023-12-14T05:22:45.323-07:00Comments on geothought: So actually, Apple isn't recording your (accurate) iPhone locationPeter Battyhttp://www.blogger.com/profile/10803273537153805225noreply@blogger.comBlogger14125tag:blogger.com,1999:blog-8470995059384390571.post-77909401934160612032011-04-30T09:18:22.829-06:002011-04-30T09:18:22.829-06:00@geoist obviously you didn't read to the end a...@geoist obviously you didn't read to the end and click through to my analysis of the raw data :). This is no more indicative of where you live or work. But if you managed to hack into my computer to obtain this file, you could find that info much more easily by looking at my contacts, calendar or email. There's also more detail about what the cache is doing in my other posts.Peter Battyhttps://www.blogger.com/profile/10803273537153805225noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-89143019341453108062011-04-27T16:26:11.028-06:002011-04-27T16:26:11.028-06:00The Apple press release vindicates your post. As a...The Apple press release vindicates your post. As a Windows 7 user, I cracked my own data myself, and ran it through a few GIS programs, as well as Google Fusion Tables.<br /><br />It was quickly obvious that the data was not tracking data. For starters, it only had 10 timestamps for 7 months! Also, the locations I most frequent did not show up, even after temporal filtering.<br /><br />If you getTimbonoreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-77714357002779313342011-04-26T07:16:49.988-06:002011-04-26T07:16:49.988-06:00@tml: ‘a primary key … of course makes it even cle...@tml: ‘a primary key … of course makes it even clearer that it isn't "tracking" you’<br /><br />I disagree. The purpose of a primary key is to reference that data in another table. So there *could* be another table or file which records (hashed?) keys and timestamps, tracking you all the time.Laurence Penneynoreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-20601662782047065792011-04-25T14:05:21.850-06:002011-04-25T14:05:21.850-06:00@mapper yes, that's essentially what's hap...@mapper yes, that's essentially what's happening with the map display tool you see here. Look at the blog post I did after this to see the raw data without this aggregation.Peter Battyhttps://www.blogger.com/profile/10803273537153805225noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-78160676074857168032011-04-25T13:47:17.860-06:002011-04-25T13:47:17.860-06:00You can get uniform grids like that when decimal d...You can get uniform grids like that when decimal degrees are cut off at a specific decimal point. So if your actual locations are (-75.12349999, 50.12349999) and (-75.12341111, 50.12341111), but your long/lat is truncated to 4 decimal places, both locations will be mapped at (-75.1234, 50.1234).Mappernoreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-84216014449807030822011-04-25T08:21:00.581-06:002011-04-25T08:21:00.581-06:00@tml good point - as you say this reinforces the f...@tml good point - as you say this reinforces the fact that the data is a cache of cell tower locations, as I had concluded and as Apple talked about in the referenced document.Peter Battyhttps://www.blogger.com/profile/10803273537153805225noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-69409846609649514982011-04-25T02:51:50.101-06:002011-04-25T02:51:50.101-06:00One thing you seem to not have noticed is that the...One thing you seem to not have noticed is that the columns that identify the cell in the CellLocation table (MCC, MNC, LAC, CI) form a primary key for the table. I.e. each cell (for a network of a carrier) is stored only once. This of course makes it even clearer that it isn't "tracking" you.Tor Lillqvisthttps://www.blogger.com/profile/18405156881630719326noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-58937308926695606682011-04-23T16:08:15.776-06:002011-04-23T16:08:15.776-06:00@Pete, I agree they wouldn't want to use GPS u...@Pete, I agree they wouldn't want to use GPS unnecessarily, but I use GPS quite a bit for other applications (including Google Latitude) so if they were intentionally trying to record my location history, I would have thought they would log those location events when they happen. But like you say, all speculation for the moment :). As I said above, am playing with the raw data now, which is Peter Battyhttps://www.blogger.com/profile/10803273537153805225noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-34281323573678638152011-04-23T16:06:22.278-06:002011-04-23T16:06:22.278-06:00My guess is that they are "generalizing"...My guess is that they are "generalizing" the data and using it for iAds. And I don't really care how accurate the data is. As a geonerd, you should know that sometimes, zip code accuracy is enough. Sometimes City-level accuracy is enough. Either way, storing this data like this is a breach of customer trust, and THAT is why people are outraged.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-75998813162290167812011-04-23T15:42:00.485-06:002011-04-23T15:42:00.485-06:00Perhaps a count of the number of 'pings' r...Perhaps a count of the number of 'pings' received by towers from your phone. The count would be affected both by the duration of time spent in an area and the number of towers that are in it. This may explain the large bubbles in the densest areas.<br /><br />Is it possible that atmospheric conditions can allow sporadic pings at great distances? That may explain the outliers.Russhttps://www.blogger.com/profile/14670158274634939097noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-16526948148625842522011-04-23T15:39:28.572-06:002011-04-23T15:39:28.572-06:00Great post Peter. As far as the question of why th...Great post Peter. As far as the question of why they wouldn't use the GPS location, one hypothesis is that continuous use would drain more power, whereas some variant of tower triangulation could be done using cell signals without any extra hardware being turned on. They are extremely serious about battery-life optimization.<br /><br />Just an interesting speculation until we hear back from Pete Wardenhttps://www.blogger.com/profile/16752128834177298274noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-88795005694423842522011-04-23T15:32:15.824-06:002011-04-23T15:32:15.824-06:00@Russ sorry if i wasn't clear in my post, but ...@Russ sorry if i wasn't clear in my post, but the rectangular grid is due to the way that the visualization tool I was using (Pete Warden's app) aggregates the data - this doesn't represent the actual detailed data. I have managed to extract the underlying data and am looking at it now, so stay tuned.Peter Battyhttps://www.blogger.com/profile/10803273537153805225noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-27013480065018635402011-04-23T15:13:37.755-06:002011-04-23T15:13:37.755-06:00Very interesting. Cell phone towers are not arrang...Very interesting. Cell phone towers are not arranged in a neat symmetrical grid all over the world. So it appears to me that the data is an interpolation of something onto a regular grid. The question is, what is that something?Russhttps://www.blogger.com/profile/14670158274634939097noreply@blogger.comtag:blogger.com,1999:blog-8470995059384390571.post-6134793134283568662011-04-23T14:34:17.086-06:002011-04-23T14:34:17.086-06:00Good analysis Peter - data does not lie.Good analysis Peter - data does not lie.Andreihttp://geocomrade.comnoreply@blogger.com