Comparison and Analysis of the Three Programming Models in Google Android
Comparison and Analysis of the Three Programming Models in Google Android
复制标题
Google Android中三种编程模型的比较与分析
DOI：
--
复制
发表时间：
2012
期刊：
影响因子：
--
通讯作者：
Xiangxing Qian
中科院分区：
文献类型：
--
作者：
Xiangxing Qian
智能手机和平板电脑正变得越来越像个人电脑。理解这类移动设备中不同编程模型的优缺点非常重要。为了充分理解它们对平台架构的影响及其技术相关性，我们在b谷歌Android提供的三套api（即Java SDK， c++ NDK和新的强大的Renderscript）中开发了相同的代表性Android应用程序的变体。基于相同的应用程序，我们进行了详细的苹果到苹果的分析，重点是动手编程的便利性、运行时行为和不同编程模型的技术相关性。我们发现目前Android提供的编程模型是可以改进的。我们提出了一个统一的解决方案，作为一个编程模型，我们希望它既能满足可编程性和性能的要求，又能保持应用程序在移动设备上部署时的安全性和可移植性。
Smartphone and tablet are becoming more like personal computer. It is important to understand the pros and cons of different programming models in these kinds of mobile devices. In order to fully understand their implications to the platform architecture and their technical correlation, we develop variants of the same representative Android application in all of the three sets of APIs provided in Google Android, i.e., the Java SDK, C++ NDK, and the new powerful Renderscript. Based on the same application, we conduct detailed apple-to-apple analysis, with focus on hands-on programming convenience, runtime behavior, and technical correlation of the different programming models. We find that the current programming models provided in Android can be improved. We propose a unified solution that we expect to satisfy the requirement of both programmability and performance as a programming model, and also maintain the applications’ security and portability when deployed in mobile devices.