Pocket tips: Don't be afraid of mass models

Recently, "Mobile new discovery" Report, Web i reading application Pocket (formerly known as "read It Later") published an official blog article by its Android application development director Maxweinach (Max Weiner The history of Android pocket apps. Weiner revealed that the initial development and testing of the application used only two handsets, pointing out that Android app developers are not necessarily intimidated by massive models and suggest that they make the most of the user community and new resources.

If you're interested in developing Android apps, some recent reports may be hitting you. As Android smartphones and tablet models become more and more (3,997 of Android devices are currently available according to Opensignalmaps), the development of Android apps seems to be getting a lot of headaches-if you buy so many devices, That's too expensive.

But from our successful launch of the Android Pocket app, and the previous experience with the Android version of Read It later apps, what I have to say is: "Don't be afraid!" ”

One can be compatible with most mainstream models of excellent applications, in fact, within reach. The following is a brief introduction to the Android Pocket application development, as well as some of our experiences.

(Note: Because the performance requirements of the game tend to be more subtle, I cannot guarantee that our approach is also applicable to game development.) ）

The first step: optional Android device

In the fall of 2010, I began to develop the Android apps for Pocket (then called "Read It Later").

I first bought a Samsung fascinate phone with a two-year contract on Amazon, which cost only 1 cents. There were four versions of the Android system: Cupcake (1.5) and donut (1.6), Éclair (2.1) and Froyo (2.2), with Éclair's market share approaching 50%.

Android also had simulators at the time, but it was too slow to do so. So in the next few months, my development work was done entirely on this phone.

After the application was gradually formed, the pocket founder, my brother Nate, also wanted to play. But I live in Oregon State and he lives in San Francisco, so he buys a second-hand Nexus One mobile phone (froyo system) on Craigslist, a classified ad site.

At this time, our internal QC process has two mobile phones.

Step two: Use the Android user community

One months before the application was released, we opened the internal tests to 50 users who were not using the models we had.

In March 2011, our app became the first paid news application in two days, and kept it until April this year, when we started to be free. In this time of more than a year, our application got 4.7 points (full score of 5 points).

Although we used only two handsets in our development and testing, 90% of Android phones are available on the market, and few users complain about device incompatibility.

(Note: When we started to develop Android apps, Android tablets were not listed yet.) If you switch to today, an Android tablet is also essential. ）

Step three: Make full use of new resources

Android has come a long way now, and VM simulators can be useful. With the simulator, you can test different versions of Android and screen sizes without having to buy one by one of those devices. In addition, Android's official version of the Compatible library (compatibility libraries) makes it easy for developers to add new features to older devices.

The design of Android apps was once haphazard, and the later Holo themes and design guidelines made Android find its own style and make the design process more enjoyable-and now it's much easier to design a nice Android app than it used to be.

Fourth step: Looking at Android from the user's point of view

For developers, the Android platform has many great things, such as push messages (C2DM), Background services, window tools, and so on. The world is diverse and users are diverse, and Android meets a wide variety of personalized requirements.

But Android developers should also learn to look at Android from a user's perspective. We should dive into the Android user community and understand the needs and perceptions of users so that they can build a heavyweight product that users like.

Pocket Status: Increased equipment

Pocket has developed a lot in the past year, so we have gradually added some equipment. We now have a total of 13 devices, some of which are bought, some are our own, and some are donated by others. These devices cover various versions of Android from 2.1 to 4.0, as well as different screen sizes and resolutions.

While there are occasional equipment problems, this does not prevent us from succeeding on the Android platform.

As the pocket continues to grow, our quality testing strength will continue to increase, and someday we will have "a 7-day, 24-hour quality testing team led by a former US Air Force officer" or "a room full of 400 units", but we started out with a 1-cent phone price. Also do a lot of things.

So what's so scary about it? If you want to develop Android apps, come on, "Start price" can be as low as 0 yuan!

Mobile New Discovery Reviews: Pocket is one of the first applications to test the Android platform and succeed in many Web  applications. Its rival, Instapaper, has been belatedly landing Android, but it has been hard to match the pocket of a year on the Android platform. We also have the question that game apps like "Temple Run" tend to have very serious compatibility issues with different Android devices-so is the pocket-type application itself better compatible? What's the secret of the Pocket team? In any case, the pocket's positive attitude toward Android is well worth the benefit of Android developers, after all, tens of millions of Android users are a huge market to be reckoned with. (Yau LA)