It has A-GPS, which is inferior in some ways to true GPS, but is much more accurate than triangulating from towers (which it will resort to under certain conditions.)
The iPhone (3G, 3GS, and 4) and the 3G iPads have fully functional GPS hardware, and also use A-GPS to acquire a position fix.
If you've ever used a normal standalone GPS in one city and gotten on a plane and flown across the country and tried to use the GPS again, you'll know that it can take 5+ minutes to reacquire a fix that first time off the plane.
Some AGPS devices cannot determine their own position from a satellite signal, whereas others can as long as they have received an almanac from a server in the recent past. It depends what you mean by "actual GPS". Some implementations are completely incapable of acting in a standalone capacity.
Here is what the Wikipedia article you linked to says: "Standalone" or "Autonomous" GPS operation use radio signals from satellites alone. A-GPS additionally uses …
I don't see how this contradicts what I said, which is that some AGPS implementations require assistance in order to accurately determine your location. Not every AGPS device can map radio signals to a lat-long; some can only do that with assistance.
It also lists some ways that various implementations require assistance:
Assistance falls into two categories:
Information used to more quickly acquire satellites
It can supply orbital data or almanac for the GPS satellites to the GPS receiver, enabling the GPS receiver to lock to the satellites more rapidly in some cases.
The network can provide precise time.
The device captures a snapshot of the GPS signal, with approximate time, for the server to later process into a position.
Accurate, surveyed coordinates for the cell site towers allow better knowledge of local ionospheric conditions and other conditions affecting the GPS signal than the GPS receiver alone, enabling more precise calculation of position. (See also Wide Area Augmentation System and CellHunter and openBmap.)
Calculation of position by the server using information from the GPS receiver
The assistance server has a good satellite signal, and plentiful computation power, so it can compare fragmentary signals relayed to it
That’s not what the article says. It says that AGPS devices are just like GPS devices. The only difference is that they use additional information to improve startup times.
Assistance is described as something that’s entirely optional. Why do you think that’s not the case? I googled around for a bit and it seems as though every source I can find tells me that AGPS is just like GPS if you have no data connection.
I'm sorry but I don't know how to make it more clear than I already said before: some AGPS implementations are not capable of operating in standalone mode. The Wikipedia A-GPS article states:
A typical A-GPS-enabled receiver will use a data connection (Internet or other) to contact the assistance server for aGPS information. If it also has functioning autonomous GPS, it may use standalone GPS, which is sometimes slower on time to first fix, but does not depend on the network, and therefore can work beyond network range, and without incurring data usage fees.[3] Some aGPS devices do not have the option of falling back to standalone or autonomous GPS.
I've added emphasis. The last point is all I was saying.
So first, someone asked whether the iPhone triangulates position from cell towers or uses GPS. I responded that it does AGPS, and that in some ways AGPS is inferior to standalone AGPS.
You objected to this, saying that AGPS is "actual GPS" which, in fact, it is not. If by "actual GPS" you mean standalone GPS, then you are wrong. I correctly pointed out to you that AGPS is not a standard definition, but a name for one of a wide range of techniques which involve assistance from a third party in determining position. You, in fact, were incorrect.
Failing to comprehend the article from Wikipedia which, in fact, enumerates the methods by which an AGPS device may receive assistance, you asserted that AGPS only refers to the technique of optionally downloading an almanac from a network resource instead of an orbitally transmitted signal. Again, you were incorrect.
Now that I've shown that you were incorrect, you want to object to raising irrelevant points? My only point was that the device uses AGPS, and that AGPS is not standalone GPS -- that in some ways it is inferior. I showed those ways because you asked, not because the Apple implementation is encumbered by them.
Ew, mate, you are using the wrong words there. Fair enough, AGPS may be inferior in a few devices but usually it is superior (i.e. has better startup times and is exactly like GPS in every other way).
Apologies if my tone of voice turned you off. I am getting fatigued with this thread. I feel like the distinction being seized here is minor, bordering on pedantic.
It uses towers to work out how to find the GPS satellites quicker than normal, and additionally can provide some location information when no GPS is available.
Really? My Garmin device can locate itself without entailing the possibility of communicating my position to a third-party. There is no possibility that my checking my position can enable anyone else to know it as well. That's not true with AGPS.
That is one pretty significant way that standalone is superior.
If that’s important to you, sure. I’m, however, a bit foggy on the actual implementation of AGPS and don’t really know whether it actually sends your location info to a server. Would be nice to get some implementation details about that.
I think it was pretty clear that we were talking about performance – time until and precision of the first lock (which AGPS does improve), overall precision (which AGPS doesn’t improve) and so on.
http://en.wikipedia.org/wiki/List_of_devices_with_Assisted_G...