Upcoming Changes to the MongoDB C++ Driver
By Andrew Morrow, Kernel Engineer at MongoDB In coordination with the upcoming MongoDB 2.6 release, the Kernel team is changing how the MongoDB C++ driver is developed, licensed, released, and distributed. What are we doing? As of MongoDB
By Andrew Morrow, Kernel Engineer at MongoDB
In coordination with the upcoming MongoDB 2.6 release, the Kernel team is changing how the MongoDB C++ driver is developed, licensed, released, and distributed.
What are we doing?
As of MongoDB 2.6.0-rc1, it will no longer be possible to build the C++ driver from the MongoDB server sources. The files required to build the driver have been forked to a new repository on GitHub, and will be maintained independently of the server sources.
Can I start using it now?
Yes. We are still fixing up some post-fork rough edges though, so please consider the current state of the new repository to be unstable. We should be announcing release candidates soon. Release candidates are useable for testing, but should not be used in production.
Why are we doing this?
By separating the development streams for the server and the C++ driver we will be able to respond to C++ driver issues without being tied to the server release schedule, as well as make changes that were previously difficult to apply in the context of the server repository. We want to make developing and using the C++ driver a better experience for everyone involved.
How will the new driver be versioned?
The new repository will initially offer two release streams: the ‘26compat’ release stream, and the ‘legacy’ release stream.
The ‘26compat’ branch and release stream will contain a minimally altered snapshot of the C++ driver files from the server repository as of MongoDB 2.6.0-rc0. It is intended to be a drop-in replacement for existing users of the C++ driver seeking minimal disruption to their current workflow and code. The build system, physical structure, C++ API, and other aspects of the driver will change minimally in this stream. We will backport selected fixes from the v2.6 server repository branch as needed, but conservatively. Enhancements and improvements will be minimal or non-existent.
The ‘legacy’ branch and release stream will more aggressively diverge from the server sources. While this release stream will retain the overall look and feel of the existing driver, we are likely to make small breaking changes to the physical structure, build system, API, and testing model for the driver. New driver features will only be made in this release stream. Our goal in this release stream is to provide meaningful improvements to the C++ driver, while requiring only modest changes on the part of users.
I see a ‘legacy’ branch in the new repository, but no ‘26compat’ branch. Why?
The initial release candidates for the 26compat release stream will be made from the ‘legacy’ branch. When MongoDB 2.6.0 is released, the 26compat branch will be forked from the legacy branch to track the MongoDB 2.6 point releases. Until that split, the legacy branch is effectively the 26compat branch. After the 26compat branch is formed, the legacy branch will begin to receive improvements and changes that are not compatible with the goals of the 26compat branch.
What license will apply to the C++ drivers in this repository?
Unlike the driver embedded within the server repository, which is a mix of Affero GPL and Apache 2.0 licensed files, all files in the new repository will be Apache 2.0 licensed.
How will existing users of the C++ driver be affected?
Our aim is to minimize disruption, but also to enable more rapid evolution and improvement of the driver. Since these goals are in conflict, there will be a period where there are several valid ways to obtain and build the C++ driver, all yielding slightly different results. We will be updating the documentation to reflect these alternatives in more detail, but as an overview, the initial options are to:
- Switch to the new C++ driver repository and use the 26compat release stream. This should offer as close to a drop-in replacement for your existing workflow and application code as can be managed. Your help in testing the release candidates of this release stream will help us ensure that our first production release of this stream is an effective replacement for the older mechanisms of obtaining the driver.
- Continue to use the server repository build of the driver. This process is valid for releases of the driver up to and including 2.6.0-rc0, but will not advance beyond that release. If you are using this process and wish to continue with minimal change, we encourage you to move to the 26compat stream.
- Continue to use the tarball versions of the driver. This process will remain valid for releases of the driver up to and including 2.5.2, but will not advance beyond that release. If you are using the C++ Client Driver tarball, we strongly encourage you to move to any of the other processes listed above.
Later, once the 26compat stream has been stabilized, there will be an additional option to track the legacy release stream. Our goal is that all C++ driver users will eventually be able to migrate to the legacy release stream.
What about JIRA, SERVER tickets, etc.?
There is a new JIRA project for the C++ driver: https://jira.mongodb.org/browse/CXX. We will be evaluating existing SERVER tickets filed under the “C++ Driver” component or label and migrating or copying them to the new project as appropriate.
New C++ driver issues should be filed in the CXX project.
Why is it called the ‘legacy’ driver or the ‘legacy’ branch?
We are saving the master branch of the repository for something new and exciting a bit further down the road. In the meantime, we want to make the experience with the existing C++ driver better, but it isn’t going to diverge radically from the ‘legacy’ of its origins in the server codebase.
Can I help?
Absolutely. We are very interested in feedback on the plan outlined above, and your feedback will help us solidify the details. We will be releasing an RC of the 26compat release stream sometime in the next week or two, and testing of that release candidate will be much appreciated.
Similarly, we will be releasing a series of unstable releases of the legacy driver leading to a stabilized version in the near future. Experimenting with these releases will give us the best opportunity to improve the driver for the community.
I have other questions…
Please post to the mongodb-user or mongodb-dev mailing lists with any questions about this announcement and the C++ Driver team will be happy to answer.
Thank You,
The C++ Driver Team原文地址:Upcoming Changes to the MongoDB C++ Driver, 感谢原作者分享。

热AI工具

Undresser.AI Undress
人工智能驱动的应用程序,用于创建逼真的裸体照片

AI Clothes Remover
用于从照片中去除衣服的在线人工智能工具。

Undress AI Tool
免费脱衣服图片

Clothoff.io
AI脱衣机

AI Hentai Generator
免费生成ai无尽的。

热门文章

热工具

记事本++7.3.1
好用且免费的代码编辑器

SublimeText3汉化版
中文版,非常好用

禅工作室 13.0.1
功能强大的PHP集成开发环境

Dreamweaver CS6
视觉化网页开发工具

SublimeText3 Mac版
神级代码编辑软件(SublimeText3)

推荐使用 MongoDB 最新版本(当前为 5.0),因为它提供了最新特性和改进。选择版本时,需考虑功能需求、兼容性、稳定性和社区支持,例如:最新版本具有事务、聚合管道优化等特性。确保版本与应用程序兼容。生产环境选择长期支持版本。最新版本有更活跃的社区支持。

Node.js 是一种服务器端 JavaScript 运行时,而 Vue.js 是一个客户端 JavaScript 框架,用于创建交互式用户界面。Node.js 用于服务器端开发,如后端服务 API 开发和数据处理,而 Vue.js 用于客户端开发,如单页面应用程序和响应式用户界面。

机器之能报道编辑:吴昕国内版的人形机器人+大模型组队,首次完成叠衣服这类复杂柔性材料的操作任务。随着融合了OpenAI多模态大模型的Figure01揭开神秘面纱,国内同行的相关进展一直备受关注。就在昨天,国内"人形机器人第一股"优必选发布了人形机器人WalkerS深入融合百度文心大模型后的首个Demo,展示了一些有趣的新功能。现在,得到百度文心大模型能力加持的WalkerS是这个样子的。和Figure01一样,WalkerS没有走动,而是站在桌子后面完成一系列任务。它可以听从人类的命令,折叠衣物

MongoDB 数据库的数据存储在指定的数据目录中,该目录可以位于本地文件系统、网络文件系统或云存储中,具体位置如下:本地文件系统:默认路径为 Linux/macOS:/data/db,Windows:C:\data\db。网络文件系统:路径取决于文件系统。云存储:路径由云存储提供商决定。

MongoDB 数据库以其灵活、可扩展和高性能而闻名。它的优势包括:文档数据模型,允许以灵活和非结构化的方式存储数据。水平可扩展性,可通过分片扩展到多个服务器。查询灵活性,支持复杂的查询和聚合操作。数据复制和容错,确保数据的冗余和高可用性。JSON 支持,便于与前端应用程序集成。高性能,即使处理大量数据也能实现快速响应。开源,可定制且免费使用。

MongoDB是一款面向文档的、分布式数据库系统,用于存储和管理大量结构化和非结构化数据。其核心概念包括文档存储和分布式,主要特性有动态模式、索引、聚集、映射-归约和复制。它广泛应用于内容管理系统、电子商务平台、社交媒体网站、物联网应用和移动应用开发等领域。

在 Linux/macOS 上:创建数据目录并启动 "mongod" 服务。在 Windows 上:创建数据目录并从服务管理器中启动 MongoDB 服务。在 Docker 中:运行 "docker run" 命令。在其他平台上:请查阅 MongoDB 文档。验证方式:运行 "mongo" 命令以连接并查看服务器版本。

MongoDB 数据库文件位于 MongoDB 数据目录中,默认情况下为 /data/db,其中包含 .bson(文档数据)、ns(集合信息)、journal(写入操作记录)、wiredTiger(使用 WiredTiger 存储引擎时的数据)和 config(数据库配置信息)等文件。
