I wonder if real-world testing brought up all kinds of 11th hour issues / use case problems. You think they went back to the drawing board on some things?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Did we forget push notifications on the iPhone?
- Thread starter dude963841
- Start date
- Sort by reaction score
I wonder if real-world testing brought up all kinds of 11th hour issues / use case problems. You think they went back to the drawing board on some things?
LOL...you and me both. It should prove interesting, especially considering I also have my own Exchange server I'm not giving up.
It should be very interesting, indeed. I bet you a lot of people never thought they would have to give up their fancy new firewire-based HD DVD recorders...until they were pleasantly surprised by the new MBPs.
Okay, so I have to tell a communications war story here. Seems appropriate, and some of you might appreciate it.
Turn the way-back machine to about 1991. We designed and built a futuristic gambling setup, where thousands of touchscreen casino VGA terminals around North America were tied together via central 68020 VMEBus computers over 9600 bps modems.
One game consisted of giving out electronic lottery cards to all the players, and then a central server "called" out lottery numbers very quickly in sequence. When you got a winning pattern, your terminal was supposed to tell the central server to stop calling, and declare itself the winner for all of the continent.
Well, you can figure the difficulties. First, terminals and entire casinos could drop offline at any time, meaning an actual winner might not be seen. Communications at 9600 bps was too slow for thousands of terminals to acknowledge getting each number, either. We faced a common casino problem of accidentally giving out a win to a device that came after the real winning device, if the numbers were delayed in transmission.
(Kind of reminds me of Apple's notification scheme, with all its inherent pitfalls.)
In the end, we came up with this clever method: BEFORE the game actually started, the central server computed all the random lottery numbers, and broadcast the entire set to all the linked up terminals. This meant that each terminal could precompute exactly when/if it would win ahead of time. (Since it's the same random numbers given to everyone, it was total legal and fair to do this.) So each terminal, when you joined the game, told the central server when it would win in the number sequence.
Sweet. Now the central server knew well ahead of time when to STOP "calling" numbers (really, just telling the terminals when to USE the next precomputed number) and declare a winner. It worked a treat, and to players it looked like an amazingly fast game that was impossible to do at the time.
Turn the way-back machine to about 1991. We designed and built a futuristic gambling setup, where thousands of touchscreen casino VGA terminals around North America were tied together via central 68020 VMEBus computers over 9600 bps modems.
One game consisted of giving out electronic lottery cards to all the players, and then a central server "called" out lottery numbers very quickly in sequence. When you got a winning pattern, your terminal was supposed to tell the central server to stop calling, and declare itself the winner for all of the continent.
Well, you can figure the difficulties. First, terminals and entire casinos could drop offline at any time, meaning an actual winner might not be seen. Communications at 9600 bps was too slow for thousands of terminals to acknowledge getting each number, either. We faced a common casino problem of accidentally giving out a win to a device that came after the real winning device, if the numbers were delayed in transmission.
(Kind of reminds me of Apple's notification scheme, with all its inherent pitfalls.)
In the end, we came up with this clever method: BEFORE the game actually started, the central server computed all the random lottery numbers, and broadcast the entire set to all the linked up terminals. This meant that each terminal could precompute exactly when/if it would win ahead of time. (Since it's the same random numbers given to everyone, it was total legal and fair to do this.) So each terminal, when you joined the game, told the central server when it would win in the number sequence.
Sweet. Now the central server knew well ahead of time when to STOP "calling" numbers (really, just telling the terminals when to USE the next precomputed number) and declare a winner. It worked a treat, and to players it looked like an amazingly fast game that was impossible to do at the time.
Ha....nice! My thing back then was doing distributed security systems, also via 9600 baud connections. It's not quite as much fun as that story, tho.
I think we might actually see some form of a RAM increase in the next iPhone which will probably allow some multitasking to take place and probably a task switcher of some sorts. Especially with the announcement of that new Palm which allows it, and considering every other smartphone does it.
I think they probably realised the limitations of the Push Notification system whilst trying to impliment it. It's been 7 months now since the announcement of the 3G so we're on the downhill slope to the 3.0 firmware/hardware now anyway.
The only thing which makes me question whether they will adopt a true multitasking approach is that they're going to look pretty silly after lambasting microsoft for having it. It'll be like saying "Yeah, they were right all along, we were wrong."
I think they probably realised the limitations of the Push Notification system whilst trying to impliment it. It's been 7 months now since the announcement of the 3G so we're on the downhill slope to the 3.0 firmware/hardware now anyway.
The only thing which makes me question whether they will adopt a true multitasking approach is that they're going to look pretty silly after lambasting microsoft for having it. It'll be like saying "Yeah, they were right all along, we were wrong."
Agreed. But what to do about the earlier models?
Might be why Jobs is letting Schiller take the stage... and the heat.
