2019 was the outbreak of open source projects I produce, and domestic manufacturers continue to have a heavy open-source project. And the world's largest open source community to embrace China Github quite fast trend in the 2019 annual meeting of China Open Source (COSCon'19), chairman of CSDN Jiang Tao, vice president of GitHub Dohmke to expand the "Fireside Conversation" GitHub currently 40 million registered users worldwide, developers from China and from the number of contribution were ranked second, more and more domestic enterprises play an important role in the open source project of international cooperation. Our active contributor to open source projects, more than 40% are added in the year 2019, most of them are 90 young people, completely out of interest in participating in open source projects. Can be said that open source is the future of the IT industry, now any company can not be divorced from the software, let alone out of open source. On a deeper level it is currently open is closed source even reverse swallow, gradually assimilated world.
Open dispute both standards battle: the current leader of the open source project is equivalent to 20 years of standard-setters. For example Google's deep learning framework Tensorflow has been able to become the industry standard for artificial intelligence, must rely on word of mouth open source user, you can say who mastered the most popular open source projects, grasps the right to speak, thus leading the industry Direction of development.
Open Source dispute is the entrance of the dispute: the current major IT vendors has launched its own operating system, AI framework open source project, in fact, the nature of the business logic or compete inlet flow of users, who can say who will be able to master the flow inlet competition in the future in seize the initiative.
Open source full-stack dispute is the dispute : Currently similar to Tencent, Baidu that most of the enterprises will own the entire front to back all of the open source technology stack, this strategy is to consolidate its leading edge in the whole technology stack for other industry reference by imitation, as its brand value and technical capabilities campaign trail.
Open source has become the highest form of competition in various IT companies, the number of STAR open source community marked the firm's technical strength to a large extent. Here we look back on the highlights of the domestic open source projects this year.
The sudden emergence of Tencent
Tencent this year before a change in strategy, opening the madness of the open source model. In sub-forum on open source technology Tencent Techo recent developers conference, Tencent will for the first time publicly expressed its open source program has now entered a "deep water" and change the past, "bottom-up" of the open source model, the "bottom up" and "top-down" evolution of collaborative development combination. Tencent open source will build on the basis of internal coordination, promote lower level, opening up more heavy technology, closely involved in the open source community building, and constantly improve the open source governance, creating an ecological developer build of.
Tencent has now joined Linux, Apache and other 9 major open source foundation, the depth of cooperation to become the highest levels of membership, and donated three major outstanding open source project Open Source Foundation. June 2018, Tencent will develop high-performance RPC framework TARS, lightweight and name service program TSeer donated to the Linux Foundation; August 2018, Tencent will donate an industry-leading deep learning framework to Linux Angel's focus on artificial intelligence LF AI Foundation; September 2019, Tencent one trillion distributed messaging middleware TubeMQ officially open source, have also been donated to the Apache Foundation, to become officially recognized Incubator project. Tencent technology accumulation in a strong flow and operating system message queues, output more open source projects.
Tencent high Star Open Source Project
TubeMQ : (Github Address: https: //github.com/Tencent/TubeMQ/) Tencent research since the distributed messaging middleware system (MQ), particularly good storage and transmission of large quantities of data, compared to other components of the mainstream MQ , TubeMQ experienced test can be described as the most, after all, the highest daily trading volume of 33 trillion data put in here, I believe in the future it will be quickly replaced by kafka TubeMQ become the new C-bit MQ software system. Its structure is shown below:
Tiny OS tencent: (Github Address: https: //github.com/Tencent/TencentOS-tiny) The IOT operating system provides streamlined RTOS kernel, but the official release time is relatively short, but has grown more mature and in terms of core control and technical support to do quite well, its structure is as follows:
Completely open Baidu
与其它厂商拼命将将自己的开源项目归并到自家旗下来为自己争夺star和贡献度相比，百度甚至将没有将其主导贡献的项目Echart（https://github.com/apache/incubator-echarts）和飞浆（https://github.com/PaddlePaddle）放在自身（https://github.com/baidu）旗下。而且也完全没有门派的偏见，开源了大量的GO语言项目，从这个角度上看，百度可以说是国内对于技术态度最为开放的大厂，做开源的想法也很纯粹。同时也受益于开源，使百度的很多项目得以低成本的快速发展。百度目前在前端技术和AI技术有一定的优势，这两方面输出项目比较多。
百度高星开源项目简介
Echart:（Github地址：https://github.com/apache/incubator-echarts）这个项目绝对是笔者这种不太会美工的程序员福音，各种Python和R语言的接口也都十分完备，随便几行代码就能达到一个特别棒的展示效果，如有兴趣可以参考笔者之前的博客（https://blog.csdn.net/BEYONDMA/article/details/85345831）。
PaddlePaddle: （Github地址：https://github.com/PaddlePaddle）目前是唯一一个能和Pytorch、TensorFlow分庭抗礼的国产深度学习平台。它集深度学习训练和预测框架、模型库、工具组件和服务平台为一体，拥有兼顾灵活性和高性能的开发机制、工业级的模型库、超大规模分布式训练技术、高速推理引擎以及系统化的社区服务等五大优势。而且百度提供免费的在线NOTEBOOK（https://aistudio.baidu.com/aistudio/index）环境供大家体验，下图是飞浆端到端的部署方案：
滴滴高星开源项目简介
Dokit：（Github地址：https://github.com/didi/DoraemonKit）也被移动开发者亲切称为哆啦A梦kit，这是一款功能齐全的客户端（ iOS 、Android、微信小程序 ）研发助手。能够让每一个 App 快速接入一些常用的辅助开发工具、测试效率工具、视觉辅助工具，而且可以支持完美在操作面板中接入用户的非通用辅助功能，功能强大，便于扩展。
cube-ui：（Github地址：https://github.com/didi/cube-ui ）cube-ui 是滴滴基于Vue重构并开发了移动端组件库。每个模块都经过严格的测试，并且对于响应速度也有严格要求，真正做到了质量可靠、体验极致、扩展性强。
小而美的RT-Thread和TDEngine
当然国内的开源项目也不光是大厂的天下，一些小厂商凭借在某一领域的长期经验也有很多惊艳的项目。比如Tdengine的star数就已过万，RT-Thread的star数也已经接近四千，可以是小型IT厂商中的佼佼者。
RT-Thread:（github地址：https://github.com/SummerGGift/rt-thread）提到咱们国产的Iot操作系统RTthread也是一个无论如何也绕不过去的标志，如果RTthread在06年初就发布了第一版，而且经过了这十几年的上亿台设备装机的技术沉淀，RTthread的社区也已经俨然发展成为国内最强的嵌入式社区了，笔者刚刚到他们的论坛上逛了一圈，真的是强手如云，大咖云集，而且各方面的文档和培训资料一应俱全。其架构图如下：
TDEngine:（Github地址：https://github.com/taosdata/TDengine）有关这个时序数据库我们已经介绍过多次了，读者可以参阅《这位创造GitHub冠军项目的“老男人”，堪称10倍程序员本尊》、《巨头垂涎却不能染指，loT 数据库风口已至》这里不加赘述了。
交友婚恋类项目热度不减
据笔者观察咱们程序员这个以理工男占主导地位的群体，哪怕是在开源网站上也常有一些交友项目成为爆款，比如我们下面要说的这个抖音机器人就已经积累了6千多star，看来边学技术边找对象也是属于IT人的硬核需求。
抖音机器人：（Github地址：https://github.com/wangshub/Douyin-Bot），此项目的思路是先用手机adb+爬虫技术获取抖音上的头像信息，再使用腾讯的人脸识别AI平台（可在https://ai.qq.com/case/facedoor.shtml申请），根据返回的年龄、魅力等数据进行来找到漂亮的小姐姐。
awesome-lover：（Github地址：https://github.com/xixinjiejie/awesome-lover）Github真是个包容的平台，除了开源代码，你也可以把他看成一个信息发布的渠道，这里也有不少单纯发布交友信息的项目，比如这个awesome-lover，这是一个聚焦在程序员相亲、恋爱的信息交换项目，所以直接把信息放在了Github上。
2019年是我国在开源项目上的爆发之年，收获之年，随着越来越多的年轻开发者的加入，开源已经逐渐成为IT的未来，我们看到目前国内的开源项目在技术社区中呈现出百花齐放的趋势，无论大厂还是初创企业均有展示的机会，愿我们国内IT产业能借开源之势继续蓬勃发展。
