There is nothing random about a GPS location update based on movement. It's quite systematic.
If it's done constantly, then of course you're right. But if it's only done upon cell tower switch, then I'm still not convinced that it's worth it:
a) If I'm on the move (like in a car or train) then it wouldn't help knowing the exact position at the very instant of the tower switch because seconds later I'll be away anyway.
b) If I'm, say, strolling though town, then the location information would be valid for a longer period, but still I'd be able to walk away quite a bit. And Loopt and other tools that tell you where your friends are would need to rely on a more exact position I guess.
Furthermore, because there's a cell tower switch involved, even without GPS the service should know that I'm on the cross-section of both involved cell towers.
By the way: I keep reading that the cell-tower-based location determination relies (i) on triangulation of the signals of all towers by which the phone is covered, and (ii) on signal travel time to the currently active tower. Is this correct? I don't think so because it appears to me that my position (on Google Maps / Latitude) just "jumps" whenever I've moved far enough.
If it's done constantly, then of course you're right. But if it's only done upon cell tower switch, then I'm still not convinced that it's worth it:
a) If I'm on the move (like in a car or train) then it wouldn't help knowing the exact position at the very instant of the tower switch because seconds later I'll be away anyway.
b) If I'm, say, strolling though town, then the location information would be valid for a longer period, but still I'd be able to walk away quite a bit. And Loopt and other tools that tell you where your friends are would need to rely on a more exact position I guess.
Furthermore, because there's a cell tower switch involved, even without GPS the service should know that I'm on the cross-section of both involved cell towers.
By the way: I keep reading that the cell-tower-based location determination relies (i) on triangulation of the signals of all towers by which the phone is covered, and (ii) on signal travel time to the currently active tower. Is this correct? I don't think so because it appears to me that my position (on Google Maps / Latitude) just "jumps" whenever I've moved far enough.
That is how the service works on the iPhone so it's not polling constantly. Plus, A-GPS doesn't take long since it's using cell towers to get data from sats. This is why the GPS in cars can take minutes to connect yet you can pop open your phone and have Google Maps know your location in a few seconds.
Hmm...
I have several iPhones (1G, 3G, 3GS) and an iPad. In every case the iPad A-GPS is more accurate than any of the iPhones (even though the 3G and 3GS have GPS chips).
The biggest problem I have with this article is the lacking mention of WebOS. Is a good reason for this?
If you're going to talk about Multitasking on a smartphone, the first thing to come out your mouth should be Palm Pre/ WebOS. The fact this article didn't seem to bring it into the spotlight is a bit...questionable.
I have several iPhones (1G, 3G, 3GS) and an iPad. In every case the iPad A-GPS is more accurate than any of the iPhones (even though the 3G and 3GS have GPS chips).
Any idea why?
.
The iPhones use A-GPS. The 3G iPad using A-GPS. The WiFi-only iPad and the original iPhone have no GPS of any kind.
Any mapping on the iPad would be done by your WiFi location. It doesn't use triangulation, but a DB has many WiFi hotspots in them so it can often know where you are.
The biggest problem I have with this article is the lacking mention of WebOS. Is a good reason for this?
If you're going to talk about Multitasking on a smartphone, the first thing to come out your mouth should be Palm Pre/ WebOS. The fact this article didn't seem to bring it into the spotlight is a bit...questionable.
Because WebOS is used on Palm phones and Palm isn't doing well. It's pretty much a non-starter at this point even though they one of the companies with a modern mobile OS.
I assumed the iPad would be at least using skyhook
Quote:
Originally Posted by solipsism
The iPhones use A-GPS. The 3G iPad using A-GPS. The WiFi-only iPad and the original iPhone have no GPS of any kind.
Any mapping on the iPad would be done by your WiFi location. It doesn't use triangulation, but a DB has many WiFi hotspots in them so it can often know where you are.
The key missing component for background processing in iPhone OS 4 is time line based applications, like IM and twitter. These can run in the background on an Android phone and can nicely stack up new messages until the user wants to read them. iPhone OS 4, bizarrely, can't do this (surely the popularity of Twitter can't have escaped Jobs and co), and requires the twitter/IM client to log in anew and refresh each time the user wants to see all the new messages. The push notification system is totally usless in this case (and in most cases to be frank).
Isn't this exactly the scenario that Push Notifications is made for? For example:
If you want to read each IM/Twitter as it happens, then the app can support push notifications.
If you want to "stack up new messages" from IM/Twitter "until the user wants to read them", then the app can just 'refresh' on startup.
I really don't understand what deficiency in the iPhone OS you're trying to point out... I just don't see your point.
I assumed the iPad would be at least using skyhook
That is the DB I was referring to.
Quote:
Originally Posted by illimiter
Isn't this exactly the scenario that Push Notifications is made for? For example:
If you want to read each IM/Twitter as it happens, then the app can support push notifications.
If you want to "stack up new messages" from IM/Twitter "until the user wants to read them", then the app can just 'refresh' on startup.
I really don't understand what deficiency in the iPhone OS you're trying to point out... I just don't see your point.
They want the app to load these IMs as they come in so that when you re-access the app it doesn't have to then poll the server to get these messages put into the app. Basically, they are trying to save 2 seconds even though there are APIs that would appear to work for them, such as the 600 seconds before the app is killed API (though I'm not sure how strict Apple will be with its usage) and the suspension API so that the app isn't restarted each time you go into it.
Comments
There is nothing random about a GPS location update based on movement. It's quite systematic.
If it's done constantly, then of course you're right. But if it's only done upon cell tower switch, then I'm still not convinced that it's worth it:
a) If I'm on the move (like in a car or train) then it wouldn't help knowing the exact position at the very instant of the tower switch because seconds later I'll be away anyway.
b) If I'm, say, strolling though town, then the location information would be valid for a longer period, but still I'd be able to walk away quite a bit. And Loopt and other tools that tell you where your friends are would need to rely on a more exact position I guess.
Furthermore, because there's a cell tower switch involved, even without GPS the service should know that I'm on the cross-section of both involved cell towers.
By the way: I keep reading that the cell-tower-based location determination relies (i) on triangulation of the signals of all towers by which the phone is covered, and (ii) on signal travel time to the currently active tower. Is this correct? I don't think so because it appears to me that my position (on Google Maps / Latitude) just "jumps" whenever I've moved far enough.
If it's done constantly, then of course you're right. But if it's only done upon cell tower switch, then I'm still not convinced that it's worth it:
a) If I'm on the move (like in a car or train) then it wouldn't help knowing the exact position at the very instant of the tower switch because seconds later I'll be away anyway.
b) If I'm, say, strolling though town, then the location information would be valid for a longer period, but still I'd be able to walk away quite a bit. And Loopt and other tools that tell you where your friends are would need to rely on a more exact position I guess.
Furthermore, because there's a cell tower switch involved, even without GPS the service should know that I'm on the cross-section of both involved cell towers.
By the way: I keep reading that the cell-tower-based location determination relies (i) on triangulation of the signals of all towers by which the phone is covered, and (ii) on signal travel time to the currently active tower. Is this correct? I don't think so because it appears to me that my position (on Google Maps / Latitude) just "jumps" whenever I've moved far enough.
Watch the event!
Watch the event!
Sorry, can't follow you.
That is how the service works on the iPhone so it's not polling constantly. Plus, A-GPS doesn't take long since it's using cell towers to get data from sats. This is why the GPS in cars can take minutes to connect yet you can pop open your phone and have Google Maps know your location in a few seconds.
Hmm...
I have several iPhones (1G, 3G, 3GS) and an iPad. In every case the iPad A-GPS is more accurate than any of the iPhones (even though the 3G and 3GS have GPS chips).
Any idea why?
.
If you're going to talk about Multitasking on a smartphone, the first thing to come out your mouth should be Palm Pre/ WebOS. The fact this article didn't seem to bring it into the spotlight is a bit...questionable.
Hmm...
I have several iPhones (1G, 3G, 3GS) and an iPad. In every case the iPad A-GPS is more accurate than any of the iPhones (even though the 3G and 3GS have GPS chips).
Any idea why?
.
The iPhones use A-GPS. The 3G iPad using A-GPS. The WiFi-only iPad and the original iPhone have no GPS of any kind.
Any mapping on the iPad would be done by your WiFi location. It doesn't use triangulation, but a DB has many WiFi hotspots in them so it can often know where you are.
The biggest problem I have with this article is the lacking mention of WebOS. Is a good reason for this?
If you're going to talk about Multitasking on a smartphone, the first thing to come out your mouth should be Palm Pre/ WebOS. The fact this article didn't seem to bring it into the spotlight is a bit...questionable.
Because WebOS is used on Palm phones and Palm isn't doing well. It's pretty much a non-starter at this point even though they one of the companies with a modern mobile OS.
The iPhones use A-GPS. The 3G iPad using A-GPS. The WiFi-only iPad and the original iPhone have no GPS of any kind.
Any mapping on the iPad would be done by your WiFi location. It doesn't use triangulation, but a DB has many WiFi hotspots in them so it can often know where you are.
The key missing component for background processing in iPhone OS 4 is time line based applications, like IM and twitter. These can run in the background on an Android phone and can nicely stack up new messages until the user wants to read them. iPhone OS 4, bizarrely, can't do this (surely the popularity of Twitter can't have escaped Jobs and co), and requires the twitter/IM client to log in anew and refresh each time the user wants to see all the new messages. The push notification system is totally usless in this case (and in most cases to be frank).
Isn't this exactly the scenario that Push Notifications is made for? For example:
If you want to read each IM/Twitter as it happens, then the app can support push notifications.
If you want to "stack up new messages" from IM/Twitter "until the user wants to read them", then the app can just 'refresh' on startup.
I really don't understand what deficiency in the iPhone OS you're trying to point out... I just don't see your point.
I assumed the iPad would be at least using skyhook
That is the DB I was referring to.
Isn't this exactly the scenario that Push Notifications is made for? For example:
If you want to read each IM/Twitter as it happens, then the app can support push notifications.
If you want to "stack up new messages" from IM/Twitter "until the user wants to read them", then the app can just 'refresh' on startup.
I really don't understand what deficiency in the iPhone OS you're trying to point out... I just don't see your point.
They want the app to load these IMs as they come in so that when you re-access the app it doesn't have to then poll the server to get these messages put into the app. Basically, they are trying to save 2 seconds even though there are APIs that would appear to work for them, such as the 600 seconds before the app is killed API (though I'm not sure how strict Apple will be with its usage) and the suspension API so that the app isn't restarted each time you go into it.