Apple issues statement on iOS location controversy, says fix is coming - Page 5

You seriously dont consider a consumer products user manual as layman oriented?

Steve Jobs did say that iDevices are magical.

Why don't you provide us with something better right here, where we're all discussing it?

Could you provide a link to that?

It's not getting lost. It's an interesting discussion. And it's part of what the subject is all about.

We're mostly concerned with what we receive while on the ground. Obviously, the higher we go, the further away the horizon is.

Regardless why would they lie in the documentation? Why don't they just say four are required?

Again from the user manual:

"The GPS unit needs to acquire good signals from at least three satellites to determine a position on the the surface of the earth. A fourth signal is required to get three-dimentional positioning which determines elevation or altitude."

I'm not scolding you at all. It's a very common error to think three sats would triangulalate your position and this give you an accurate location. Common-sense really. But that's not how GPS works.

In the most simple of terms, it's just timing. Very, very precise timing. Milliseconds. You have most of the picture correct.

Three satellites have placed you in a fairly specific spot on a two dimensional plane. But GPS timing signals can't work in two dimensions. Why? Because your navigation device is using the time it takes to receive the GPS signals as a reference to determine where you are in relation to the known position of the satellites. Without that 3rd dimension how would your device determine how far you are from the signal source? You're dealing with a spherical object rather than a flat piece of paper. Without that 4th point, there's no way to compute the distance, and thus time difference, between the source and your device. Elevation is not just a nice thing to know. It's a required element in the mathematical equation.

MStone: Garmin isn't lying. They've told you that the forth satellite locked will provide elevation. I just expalined why that's needed. Garmin could have been more clear that they mention the 4th sat because it's a required piece of the puzzle. But the manual never says that only three sats are needed for positioning. It could have been stated more clearly.

Im not sure what part of that is hard to understand. I copy and pasted a good ovutline or what 1, 2, 3 and 4 satellites can do for your location. I even posted the damn pictures.

It took rocket science to get them into geosynchronous orbits, but it doesnt take rocket science to understand what is being discussed and how a technical websites about GPS trumps your consumer user manuals.

What seems to be happening, is that there are two things being measured. Horizontal position, on an x/y axis, which requires three satellites, and height position on the z axis, which requires four.

I'm reading from some people that four are required, but I'm also reading here, and in other places that only three are required for land use. I don't believe that the articles that state three are needed are wrong. And until someone provides a decimate link that can be trusted, that proves, conclusively, that four are needed, at a minimum, it isn't proving believable. In addition to the manual posted, I distantly remember, from the GPs units I've owned over the years, that three were needed, unless altitude was required. In the three articles I've read in the past day, all three also say that three are needed for terrestrial purposes, and that four are needed for flight purposes.

So again, unless someone posts information that definitely shows us this is untrue, I see no reason to believe it, no matter how insistent some people here are. If information is posted that is reliable that does prove that four are always needed, then I will believe it. The same thing is true for maximum numbers.

When I screw up either deliberatly or by mistake I apolgize, why can't Apple do that?

It doesn't matter whether they are layman or not. What matters is whether they are right. I believe that Garmin, and other GPS manufacturers know what they're talking about. I'm not so sure about others here who insist on giving different information.

I dont think they are wrong, either, I even posted how less than 3 can determine your position, but three is to determine your position on a plane which satellites 12,000 miles up measuring a tiny object on the surface of a sphere cant accurately accomplish without 4 satellites. Since they are geosynchronous and the Earths surface isnt changing with any great degree (even Mount Everest is only 5.5 miles above sea level) they will often do a decent job with 3 with some assumed data on elevation.

The manual is saying exactly what you're saying it isn't saying. How can you disagree with that? If you're saying the manual is wrong, that's different. But it very definitely is saying that three satellites are needed for horizontal positioning. It's quite clear.

Like the "Birthers", it's not about facts, it's about what you are predisposed to believe. This full disclosure will give confidence and back-up to those who saw this "spying" issue for the hysteria that it was. Similarly, it will be rejected out of hand by those who are natively paranoid, Apple haters, or both.

Obama released the original birth certificate today knowing full well that those who are either willfully ignorant or cynically manipulative about his American citizenship will find many ways to dismiss it. It's a lose/lose and he knew it, which is why he didn't do this sooner.

I can't "prove" that I wasn't born somewhere else so how can he. My birth certificate, and yours, could be fakes, right? You can't prove a negative.

Sorry, had to vent. I know I should have done it elsewhere, but I am here. What can I say?

It's not. You need three satellites.

Simple answer. There are four unknowns in the equations to solve for GPS posisitions. 3 of them are positional unknowns and the other is time because it's impossible to sync the clocks on the satalite together to the accurancy needed. Because there are 4 unknowns in the equations there needs to be 4 measurements from different satalites to determine a complete position.

However if they assume an elevation they can find Lat and Long on 3


Also I like Apple's way of working the GPS data better than my old garmin. The Garmin's filter would struggle to converge if your previous location was much different than your current position (think flying across country). The iphone and the Tom Tom app doesn't have that problem

The P code can be encrypted as a so-called P(Y) code that is only available to military equipment with a proper decryption key.

Mel, I already gave you a "trusted source" besides me. GPSInformation.net is an authoritative source for gps-related technical papers and both basic and intensive articles on the uses and functions of gps devices and features.

I listed my qualifications earlier. I'm considered a trusted source within the navigation community. I've now given you another trusted source in addition to a gps specific forum from earlier, GPSReview.net. That's three authoritative sources.

The articles you are depending on for your information are not what you would consider trusted sources are they? You really would look at them as experts in the GPS field simply because thy have an article on the web, yet doubt actual GPS sources who specialize in that field?

Why not test my statement?. Take your gps device outside, pull up the satellite screen and see how many satellites you show when your device shows it's ready to navigate. Simple enough.

As a quick clarification the GPS constellation is not at geosynchronous orbit

Youre literally thinking two dimensionally and ignoring what 1, 2 and 4 satellites are establishing.

But, it's the other way around. You can position accurately on the surface of the earth with three satellites, but you need four in a three dimensional space which is the case with flight, where height is a critical component.

Even if I'm traveling in the mountains, I rarely care about how high I am, but I care WHERE I am in relation to other locations. In that case, three will work. Four MAY be required for tracking purposes, but I'm not certain of that, as my older models didn't need more than three for routing and waypoints.

The way they do all of this is very clever, and looking at the math in wikipedia's article makes it pretty clear to me as to how clever it all is.

I think this disagreement is about the difference between theory and practice. Theoretically, it requires 4 satellites to determine an unambiguous position. Practically, a location on the "surface" can be determined with 3 by a GPS unit assuming you aren't out in "space" -- you are in one of two places, and one is discarded as not within expected parameters.

That's true, which is why emphemeris data needs to be constantly updated. Signal timing corrections requiring approx 30 seconds to download the entire packet. But that could take your device longer than 30 seconds to receive it if you "discovered" it midway thru the data send. It repeats every (I think) 30 seconds, 24 hours a day.

Just a technicality, but note that the GPS constellation is not in geosychronous orbit. Not only would that require them all to be over the equator, but would also place them so far out that signal-to-noise issues at the receivers would be difficult to handle.

The ephemeris data only require updating due to orbital variations - not specifically because the orbits are not geosynchronous.

Correct. But still a reason that it does need to be updated regularly. A non-geosynchronous orbit would be an example of an orbital variation.

There is no disagreement between the theory and in practice. In theory 4 satellites are needed to determine an accurate position estimate. In practice 4 satellites are needed to determine an accurate position estimate.

All the disagreement and doubting won't change that from fact. I already gave you a simple test. Did anyone bother to try it, or would you rather just continue guessing that I must be wrong.

This is the problem when reading an article from Wikipedia. They usually get it pretty correct, not not completely correct. This is one of those times. Their error is the usuall Wikipedia issue of omission. In order to understand this issue, you really need to read the number 59 footnote, which take you to the original article from which the information about P codes and the rest came from.

If you do, you will see a number of I treating statements. One is that "mostly millitary" equipment uses these codes. Other equipment dies as well. Then there is the date, which is 1994. These codes were open up in 2000, as has been mentioned earlier. The military may encrypt the information going to and from their own equipment, but it's still there, in unencrypted form for everyone else today.

http://www.kowoma.de/en/gps/signals.htm

Well, while we're providing links to our arguments, were still waiting for yours. I'm conceding that you may very well be correct about everything you've said, despite information showing differently. But until you provide a good link, we don't know. I would be very happy to acknowledge whatever you say, if you can show it to us in a way that invalidates whatever else we've got. It's up to you.

I wrote geosynchronous. Geostationary specifically refers to a specific kind of geosynchronous orbit at 0° latitude (equator). Geosynchronous orbits can be elliptical (as well as other types of orbits I can’t remember).

I specifically choose not to refer to them as "elliptical geosynchronous orbits” since we can’t even agree on how many satellites it takes to determine a point on a three dimensional Cartesian coordinate system.

Unless I’m missing something and you two are saying these satellites don’t continuously follow the same elliptical path.

How many satellites are required by my Garmin device to determine latitude and longitude? They say three, I have observed this in practice. In my 'layman' understanding of this is that the elevation is assumed to be the last known elevation or sea level if there is no fourth satellite.

I imagine that you have a sphere with a line intersecting it through the center of the sphere. There are two locations where the line intersects the surface of the sphere. One location is very near where you were last known to be and the other is on the opposite side to the earth. The one on the opposite side is not relevant. Assuming that the software in the Garmin is smart enough to work with three satellites and the assumption that you are on the earth's surface, only three satellites are necessary. I have never received a message on the device that there aren't enough satellites to calculate my position when I have at least three.

So, practically, a GPS, using appropriate heuristics, and assuming the earth is a sphere can determine your position on the earth's surface with 3 satellites. You may or may not have some knowledge of GPS theory, but, as usual, you don't know what you are talking about.

How many more links would you like? I've posted at least two professional and gps-industry specific sources as well as listed my own qualifications. I explained the source of your misunderstanding of the Garmin manual. And I've given you a test to prove me wrong.

I'll try one more time. This one has pictures
http://www.trimble.com/gps/howgps.shtml

NOTE: Be sure to read the whole thing, not just the first page. Step 3 and it's explanation of timing is important.

Ok that site is really bad. Please just find the words: "GPS requires four satellites" on an authoritative source and link to it like I did with the documents posted. I'm not going to read through a whole website trying find what you are claiming is in there somewhere. And by the way the pictures on the page you referenced show only 3 satellites.

Well that's not quite correct. A geosynchronous orbit is one that matches the rotation of the earth, but in all other respects it is just an orbit like any other, i.e. from the point of view of the satellite the fact that the earth has the same rotation period has no effect on the mechanics of the orbit. All orbits, geosynchronous or not, are described by ephemeris parameters, which only need to be updated due to orbital shifts - either intentional orbit corrections or natural variations due to deviation of the system from the ideal model used to construct the data.

If you're not willing to take 3 minutes to get as far as Step 3, I've wasted hours to try and explain. I 'm not going to bother wasting another hour on finding and posting links that you can't be bothered to read.

It's apparently only important to you as something to argue about rather than take the time to understand. And that's fine. Not everyone is curious about how GPS works, only that it does.

Now you’re just being obtuse.

Seriously, what part of post #101 is confusing about how 1, 2, 3 and 4 satellites work? What part of Cartesian coordinate system are you not getting? There is only one way to get a single point in a 3-dimensional space. As it’s been stated ad nauseam it assumes your terrestrial position but it actually doesn’t know it until 4th satellite is used. Don’t they teach geometry in middle school any more? I’m done with this conversation. If I wanted to teach elementary maths I’d get a job at a school.

Ah, finally someone besides Solipsism that has taken the time to try and understand the whys and whatnots of GPS. Welcome to the discussion Muppetry

(Posted in the other iOS location snafu thread since I was replying to specific posts there, but probably belongs here, as it's regarding info from this post today)

Overall, I was pretty pleased to see Apple's official response today. It took a little longer than I think many people would have liked, but it's generally plausible, and I still feel that Apple is generally one of the "good guys" in the realm of privacy. Minor issues pointed out here:

When they say that "Apple is not tracking the location of your phone", it may be true from a pedantic nature, but general location information about your phone was (and will continue to be if you opt to allow it) being passed along to Apple. It may not be highly accurate data, and it may have been purged or otherwise aggregated on their end to the point of not identifiable, but location data was passed on.

When they say "This data is sent to Apple in an anonymous and encrypted form. Apple cannot identify the source of this data.", I don't quite buy that. If they choose to take note of where the data came from, it could be identified, even if the identifying info is not explicitly in the data itself. Yes, this is a minor point, just throwing it out for sake of completeness.

Lastly, on this PR statement, I don't believe the two issues were really "bugs", but I could be convinced that they were simple oversights. Yes, there is a difference, although it's not really that important in the scope of explanation to the public.

But overall, these are small nits, and I applaud Apple for coming out and not only explaining what's going on, but taking steps to rectify the problems. To be honest though, they had little choice. The media and public at large were not going to let go of this easily.


As I alluded to above, any sizable amount of location-based data, even with no user-identifying bits, should not be assumed to be anonymous. It's possible that this particular data was not accurate enough to be identifiable, but I haven't seen anything (yet) that leads me to believe that's the case. Here is a great article, worth reading in its entirety:

http://33bits.org/2009/05/13/your-mo...ocation-pairs/

The PARC article referenced within is also good, but more technical than most people will want to read.


Key words: so far ! ;-)


As for reading fine print, you and I are in a very small minority!

As for being naive, that's a pretty harsh statement. But I do think everyone should consider that whenever vast amounts of personal information is in the hands of any 3rd party, it's dangerous. Apple does seem like one of the "good guys", but:

1) management changes, the data they own never goes away.
2) security breaches occur all the time, and the more data in any one place, the bigger a target it is.
3) companies are indeed gathering more and more profiling information all the time (not just location-based)
4) when the feds do step in and make use of that data, the public absolutely does not get to know about it (this is a fact).

Now, this brings up the real elephant trying to sneak around in the back of the room: Google!

Apple is in business to sell computers, iPhones, etc. Gathering personal information is helpful to them, but ancillary. It's easy to see why they'll want to enable various features that rely on personal information, but making the vast majority of that data opt-in does not run completely contrary to their business model.

Google, on the other hand, needs to know as much personal profile information about their users as possible because it's the crux of their business. I am dying to see how Google responds to this inquiry!

Ok you posted a link, I went to it, it did not mention anything about four satellites, I came back here to find that you had edited your comment saying read all the pages. My experience with the site was not good as the navigation sucks and in my opinion is a very low quality site which I would not have any more confidence in than your unsubstantiated claims that if you don't have four satellites locked you will lose your GPS fix.

Find the authoritative page like a university paper, manufacturer document etc, link to the exact paragraph, sentence that supports your claim and I will be happy to write to Garmin and ask them for an explanation on the apparent contradiction.