What is the difference between App testing and traditional software testing? What aspects of test cases should be added?

With the increasing influence of mobile phones on people's lives, App testing has gradually become familiar to everyone. From crowdsourcing at the beginning to automation exploration now, the technical development of mobile phone testing is also changing with each passing day.

App testing is much more complicated and difficult than traditional soft armor testing.

According to my work experience, I will summarize how to do a good job in app testing as follows.

( 1) ? Non-functional test

An important aspect of app testing is the non-functional requirements of app. Before the mobile app is put on the market or further developed, testers have a certain responsibility to track such requirements.

The first test conducted in the early development stage should be the usability test. Usually done by alpha users or colleagues. Walk into a cafe or restaurant and ask the people inside about their application usage. Let them look at the first version developed at this stage and collect feedback to see if users can make good use of the new features in order to get a first impression.

(2) ? functional test

Every new function developed needs to be tested. Functional testing is an important aspect of application testing. Testers should perform manual testing and subsequent automated test maintenance. At the beginning of the test, the tester must manually test the application as a "black box" to see if the provided function is correct and whether it works as designed. In addition to classic software tests, such as clicking buttons and submitting orders, testers must also perform more functional application tests.

In addition to the entire manual testing process, test automation is also very important for mobile apps. Every code change or new function may affect existing functions and their status. Usually, there is not enough time for manual regression testing, so testers must find a tool for automatic regression testing. There are many automated testing tools in the market, both commercial and open source, which are oriented to different platforms, such as Android, iPhone, WindowsPhone7, BlackBerry and mobile Webapp. According to the development strategy and structure, quality management testing experts need to find the most suitable automation tools for their environment.

(3) ? Client performance test

Whether an App is good or not is not only reflected in its function. The performance of the tested app on low-end computers is also very important. For example, a fun game or application can only run smoothly on high-end computers, but not on low-end computers, and it will not have a good reputation.

Regarding the performance test of App, the parameters we pay attention to are: CPU, memory, power consumption, traffic, FPS. At the same time, we also need to pay attention to the time-consuming installation and startup of the App.

At present, everyone may be confused. How high are CPU, memory, power consumption, traffic and FPS? Here, I can tell you that you can refer to some values of boutique games and compare your own developed app with the app data of industry boutiques.

(4) ? Adaptive compatibility test

After the functional test, the App also needs to be tested for compatibility. The items to be checked mainly include the following points:

(a) Whether the installation, pull-up, click and uninstall on different tablet brands are normal;

(b) Whether the installation, pull-up, click and uninstall on different operating systems are normal;

In the actual test, we often encounter the following problems:

(a) app cannot be installed on a flat system;

(b) The app cannot be pulled up on the flat card system.

(c) On the flat card system, there is no response after the app is pulled up, or a black screen or a flower screen appears after it is pulled up;

(d) On a tablet brand system, the app cannot be uninstalled smoothly;

(WeTest Tencent Quality Open Platform) This product can realize the adaptation compatibility test of many popular models.

(5) ? Weak network test

In the process of using App, it is inevitable to encounter weak network environment, such as on buses and subways. In this case, network jitter, uplink or downlink timeout often occur, resulting in packet loss in the application.

As a tester, we should make a weak network environment model before the app goes online to see if there are some unknown problems in the weak network environment. The following are common situations in a weak network environment:

(1) 3G weak network signal scene simulation;

(b) Simulating low-speed moving scenes in urban areas;

(c) Simulating high-speed moving scenes in suburbs;

(d) request response timeout _ uplink timeout scenario simulation;

(e) request response timeout _ downlink timeout scenario simulation;

(f) network jitter scene simulation;

(6) ? Power consumption test

The performance of the App on the mobile phone, in addition to the function, whether the app consumes electricity is also the focus of the test process. How long can this App play when the mobile phone device is fully charged? What is the power consumption of the App per hour? What is the power consumption of 10 minutes when the App hangs up in a certain scene? These are the points that we usually pay attention to in power consumption testing.

(7) ? Protocol testing

Simulate the client sending the protocol package directly to the server to see if the server has a certain check and recognizes the data sent by the client. Protocol testing is mainly to deal with users sending malicious protocols to the server to fool the server's verification.

(8) Safety test

Before the App goes online, you need to do a detailed security test. Security testing is mainly to detect whether the application is easy to be cracked by the outside world; Whether there is a risk of being injected by malicious code; The risk of plug-in is high after going online.

(9) Server performance test

Server performance test, mainly including single-machine capacity test and 24-hour stability test. Stand-alone capacity test is to detect how many users a stand-alone server can carry under the premise of 90% response time and success rate. Use a specific game mode for 24 hours, the service does not restart, the memory does not leak, and the success rate of each transaction reaches the standard.

This can be booked at the wettest entrance.

(10) Server Disaster Tolerance Test

Server disaster tolerance test mainly refers to whether a service process has the ability to recover itself after it crashes. For example, after the game logic process disappears, will it automatically pull up; Will memcached restart when it crashes, and will it affect all players? These are all factors that need to be considered in the process of app testing.

(1 1) interrupt test

Aiming at the service level classification and real-time characteristics of intelligent terminal applications, this paper puts forward testing methods, such as interactive testing between App and key applications such as incoming calls, file downloading and music appreciation when running in the foreground and background. Test the app's response when the phone, SMS, MMS, Weibo or other notifications come in.

(12) public opinion tracking in the late online period

After the launch of the new app, users' comments on the app, bugs that were not noticed during the test, hot posts on the forum, and word-of-mouth of the app in the app store are all points that testers need to pay attention to after the launch of the app. If you need a Bug that was not found during the test, you need a new test suit to confirm and fix it according to the problem.