[OSX-Users] Re: iPhone 3.0 battery life
Hugh Glaser
hg at ecs.soton.ac.uk
Wed Jul 22 14:53:56 BST 2009
Excellent.
Pleased to hear that.
If anyone is still wondering why...
Here is my (inexpert) take.
It keeps a connection open to the push server. And it needs to make sure it announces it's presence (I think every 9 minutes), to keep the TCP connection from timing out. So, yes, it is pushing, but the push server (at Apple) needs to know how to contact it all the time, which can only be by the phone contacting it.
Or think Dynamic DNS.
It is the same as my machine telling a DDNS server the IP address to map the domain to at regular intervals.
Or SMS (or even calls) - it might look different, but isn't.
The phone is constantly telling the network where it is, by contacting the base station.
So when the service provider has a message, it can push it straight to where it knows the phone is (unless you are at Glasto, in which case it may take a day or so :-( ).
The reason that it can do this is because the phone battery is being drained by the base station pinging.
Of course, this is the normal mode of a phone that is able to receive calls, so you don't notice a battery hit at all.
The difference is because push is at the top of the protocol stack, and so is an expensive connection to make.
If you built push lower down the stack into the data/phone network you could achieve the same thing as a side effect of the normal interaction, but the service providers would have to provide an API; mind you, maybe they do, along with location data? I have a feeling that Blackberry may have got some such deals.
Cheers
Hugh
On 22/07/2009 13:47, "Tim Chown" <tjc at ecs.soton.ac.uk> wrote:
For info, turning off push mail has made a huge difference.
--
Tim
More information about the Osx-users
mailing list