IPhone push notification

Thank you @ lawrencelry for inviting me to participate in the cocoaheads activity. I am lucky to have met ichm and buddyfeed author @ robinlu. The three of us discussed Apple's push, which is summarized below.

------------

Technology: a firm expression of pain

First, push is essentially a service. Therefore, if you do not consider the interests of mobile phone manufacturers and operators, the carrier will achieve the most ideal push. The best PUSH Service in the world, provided by BlackBerry in cooperation with carriers, is not accidental.

Apple, Nokia, and other vendors are determined to take another path: bypassing operators.

The word "Wrap" shows the embarrassment of this solution. However, the development of technology and the popularization of mobile networks make this goal more and more realistic. The article behind push notification analyzes Apple's push solution. But that explanation is incomplete. He only describedFrom apple to usersThis section. @ Robinlu, as a developer, completed the other part of the push puzzle for us:From developers to Apple.

 

Originally, except for Apple's push server, developers mustMaintain another Web server by yourselfTo collect the push generated by your program and send it to Apple's push server.

A brief description is as follows. Click to zoom in

 

Suppose that buddyfeed must support push...

A buddyfeed user sends a comment and submits an update at friendfeed.com. After,Webserver maintained by developersThe updated notification will be received from the iPhone's buddyfeed client (or from friendfeed.com.

The server on which the developer receives the notification is called app push web server. To handle this notification, it becomes an acceptable standard form for Apple push server to send to Apple. The push server of apple then pushes the message to the user in the following post on push notification.

------------

Business: Waist Soft and easy to overturn

Technically complete? However, considering the business considerations, it is not that good.

This solution requires developers to maintain a web server. This isContinuousExpenditure. The price of the push program on the App Store is mostlyOne-timePayment.

Im +: $4.99, boxcar: $2.99, gpush: $0.99!

Consider the relationship between the sales of the software and the traffic generated by the software. Sales increase, decrease, and decrease are normal, and users served by web server will always grow !!! More users are equal to more traffic, more bandwidth, and more expensive Server rental. Over time, when users are already very large, software sales have become saturated. What options will developers make?

You are charged again for the software you have purchased, or,Simply turn off the Web server he maintains !!!

Apple's push implementation is scrawled. The new notification covers the old ones. You don't know where to find the red numbers on several programs. However, this design problem is even more fatal. When sales fall to the inability to maintain the monthly rent for Web servers, what are the benefits of cheap push software?

------------

Way out?

There are at least three ways for Apple to push notification:

First, Apple provides app push server for developers.

The second is in app purchases, which is charged on a monthly basis.

Third, push ads.

Currently, the push software in app store has selected method 2, such as tweet push. Although he is more likely to provide reliable and long-term services, it is far from a paid push software, regardless of the rating or popularity.

------------

Even opportunities? Personal SaaS?

ERP and other enterprise-level applications develop the concept of SAAS. Software as a service (software as a service ). Instead of selling software, we sell a set of web and software-based Elastic Solutions and provide support. For this reason, monthly/yearly rent is collected. This billing method for SaaS applications has been widely accepted by enterprises.

As mentioned at the beginning of this article, push is a service. The implementation of push on the iPhone is actually the result of the decentralization of such enterprise-level concepts to individuals. In fact, a large number of individual users are also paying for services. On a traditional network, you have an account with a Flickr pro account. On mobile networks, a large number of users in Japan pay 350 yen/month to get mms while enjoying Push Mail. There are also many BlackBerry bis users.

Providing high-quality push services for iPhone users and charging by month may become the norm of the future market.

However, Apple's push technology is not detailed, making it difficult for common users to accept monthly subscriptions. The implementation of push is sloppy, so that the value of push should have been discounted. The low price and price war in app store make it difficult for developers who insist on monthly rent.

All of the above are damaging this opportunity.

Smartphones and Their Application markets are brand new and rapidly growing fields. From iPhone to App Store, Apple's rare design and innovation have also maintained a very high level of implementation. But I have to say that the design and implementation of push do not deserve this comment.

However, I believe that both Apple and developers are looking for better solutions. Apple will not stop improving push. In the end, the app store model should allow well-developed and responsible developers to stand out.

Time will test everything.