back when the App Store was announced. Push notifications, which let any software firm’s servers publish data updates to your device, were initially due in Sept. 2008, and they’re still not here. Nobody knows quite what the problem is, or how big a fix it requires
and linked to Gizmodo:
As you can imagine, this makes push notification a Holy Grail for users and developers alike. The only people who may not be happy about these are the carriers. After all, the idea of an instant messaging application with push notification services taking over their lucrative SMS business doesn’t seem like a very good one.
Or maybe I should take off my tin foil hat and just assume that Apple has hit a roadblock that nobody at engineering ever expected. But a two month delay? Why? It just sounds too weird.
I think the issue has many parts.
There’s the initial technical issue of creating and scaling a secure service that will work with nearly 25 million iPhones and iPod touch devices. The latter are a challenge because they’re not connected all the time. We can point at RIM and say that “They managed” but RIM’s service is slightly more managed and we’ll get onto that in a moment.
The second issue is one of economics. It’s absolutely true that carrires are not happy about instant messaging on phones using all-you-can-eat data plans. Don’t believe me? Look at the furor created when Nokia bundled Skype with their latest phones – Orange and O2 indicated they might refuse to stock the N97. Those lucrative 600 texts I get a month used to be insufficient for my needs when I only had text and telephone to communicate. With all-you-can-eat data, I’m using less than half my allowance every month and my talk minutes usage has decreased even more.
The third issue is one of signal and noise. I think that Apple themselves are reeling from the fact there’s 25 000 apps in the App Store right now. I think they’re seeing the iPhone and iPod touch being used for much more than they planned and they’re working to accommodate that. And I think they’re being cautious about the process by which apps will be permitted to communicate by Push. If you have one instant messenger application and one email client, then you’re not going to see too many updates. But some people have 9 pages of apps. I currently have 33 non-built in apps installed. If half of those started Pushing notifications to me, I’d be forced to silence some of them. And this ability to push notifications also puts the developer directly in touch with the customer – something Apple has avoided for the most part with the current setup. They’re having to work out interface, policy and secure scalability.
It’s not just about messaging notifications – it’s about views. Being able to send a discrete message which will launch a view to fetch data. This allows you to keep your heavy business logic on your server – for your “Ghost” to be in the “cloud” and for it to send the notifications of the things that are important.
For example: We already have new details to our schedule being added by Push. What about a server based app that asks you every 15 minutes where you are? In order for it to tell you where you need to be to catch your next meeting. Maybe you can walk if you leave now? Maybe you’ll need a taxi if you leave in 15 minutes. Our mobiles know where we are, they know where we’re planning to be and they have access to maps, public transport, traffic reports and more. And that’s just using the stuff we already have! The possibilities are much more exciting when we think about the apps that haven’t been built yet!
I understand that this Push Server will take time to be built. I just wish it was here.
Wondering how battery life will be affected in the iPhone with all this pushing – it’s already a big issue.
I’ve worked closely with SMS servers in the mobile networks and can vouch first hand it’s surprising the sheer amount of time & effort that goes into making the SMS infrastructure work as quickly and reliably as it does across a mobile network – especially around peak times.
SMS as a service has been evolving and improving for 10-15 years now so I’m not overly surprised it’s taking Apple some time to figure this out.
In the early days of the Mobile Internet, push was also a late comer. Whilst the mobile industry finally settled on a generic push mechanism that could work across multiple bearers (SMS, GPRS/IP, 3G etc) SMS ended up being the predominant bearer for binary encoded push messages. Using SMS meant push messages could piggy back off the default features of SMS (retry, store and forward etc) for when the phone is turned off/unreachable.
It’ll be really interesting to see what Apple come up with. Along your lines Matt I would say an IP based request/response would be most likely. On Wifi it would be free and when mobile the 3G data costs could be buried in the monthly bundle or they’ll asking for another £x/month to have push added to your bundle.