瓶中云:让每个人都能轻松自建服务
Cloud in a Bottle: making self-hosting accessible to everyone

原始链接: https://cloudinabottle.org/blog/launch-post

**Cloud in a Bottle** 是一个开源平台,旨在让自托管(self-hosting)变得像使用智能手机一样直观。它旨在解决现代网络软件“恶化(enshittification)”的问题,即中心化云服务往往将企业利润置于用户体验和数据隐私之上。 虽然 Nextcloud 或 Coolify 等现有的自托管解决方案常常存在易用性差、安全隐患或缺乏集成等问题,但 Cloud in a Bottle 提供了一种精简的替代方案。它在安全的无根(rootless)容器中运行应用程序,同时提供类似于移动操作系统的统一体验,包括集中式身份验证和应用间权限管理。 通过让用户掌控自己的数字环境,Cloud in a Bottle 旨在在不依赖监控或厂商锁定的前提下,重新获得云软件带来的便利(如可访问性和协作性)。该项目开源且透明,并提供托管服务选项以确保其可持续性。其目标是推动新一波高质量、用户友好的开源网络应用浪潮,使每个人都能轻松使用,而不仅仅是技术极客。

关于“瓶中云”(Cloud in a Bottle)的 Hacker News 讨论凸显了人们对现代云服务商(AWS、GCP、Azure)日益增长的复杂性、厂商锁定以及高昂成本的挫败感。 **主要观点:** * **问题所在:** 参与者认为“云原生”开发变得愈发繁琐。一些简单的任务——例如基础的密码保护或静态网站托管——现在却需要复杂的 IAM 配置文件、容器编排以及昂贵且不可移植的专有服务。 * **“自托管”运动:** 许多用户开始回归自托管,以重获数据主权并降低成本。讨论区中人们在争论这对于大众来说是否是一条可行之路,还是仅仅是技术发烧友的“套件车”式爱好。 * **准入门槛:** 虽然“瓶中云”旨在简化部署,但用户指出真正的障碍仍在于基础设施层面:网络可访问性、DNS 管理、端口转发,以及对持续更新和备份的需求。 * **人工智能的作用:** 一个显著的共识是,人工智能体正在通过自动化处理那些“困难”环节(如配置安全的 Kubernetes 集群或管理复杂的 Linux 环境),从而使自托管变得触手可及,而这些环节在过去对非专业人士来说过于耗时。 总体而言,社区对于“交钥匙”式的个人云是否真正可行,或是复杂性是否为可靠性所必须付出的代价,仍存在分歧。
相关文章

原文

TLDR: Cloud in a Bottle is your open-source personal cloud: containerized apps, unified auth, good UX. Self-hosting should feel like using a smartphone that serves webapps, not a sysadmin side job.

screenshot of the dashboard

I'm frustrated with the state of the digital world. The software we use is so misaligned with our own incentives (intentionally addicting, loaded with ads, tracking you and selling your data, enshittifying). Software is easier than ever to make; how did we end up here? I think a big reason is the shift of software into the cloud. Software served from the cloud (think Google Docs vs Microsoft Word) delivers a great experience - no install, accessible on all your devices, immediate sharing/collaboration. But to serve something in the cloud costs money, and so just about everything on the modern web has a company behind it - open source authors naturally don't want to pay to serve their free software to the world. And those companies have a financial incentive fundamentally misaligned with our own.

In the pre-cloud days, open source worked - authors distributed their software, and everyone "hosted" it for themselves (i.e. ran it on their own PC). Nowadays, cloud software is centralized, with one instance of the software serving the world, and the costs borne by the creator. This is because individuals don't have an accessible way to put software into the cloud for themselves.

Self-hosting exists already, but it's niche and unapproachable. When I looked at the existing options, none of them felt right:

  • sandstorm.io: very close in spirit (their article from 2014 makes basically the same argument I'm making here), but unfortunately long abandoned. It also required significant changes to existing software to run on their platform.
  • nextcloud: slow and unreliable, and more enterprise-targeted now
  • yunohost: apps run directly on the host, no sandboxing. one insecure app compromises the entire server.
  • coolify: hosts containerized apps, but every app is an island with its own login and little integration with the host or other apps.

We created Cloud in a Bottle to take another stab at this. At the core, it's just an Ubuntu machine with a web server that hosts a dashboard and routes HTTP(s) requests to containerized apps. Apps are (rootless, hardened) containers, so it can run existing software with minimal changes - and in a reasonably secure sandbox. But beyond just hosting containers (like e.g. coolify), there are various opt-in platform features that apps can take advantage of to make the experience better. In a way, I want it to feel like your "cloud smartphone" - something that you own, where finding and installing great apps is trivial, and where the user experience is as good as the closed source products we're trying to replace. As a basic example, if you're logged into your instance, you're automatically logged into all your apps - no separate accounts in each app. And there's an interface to allow permissioned access to data and capabilities between different apps in your instance. More generally - think of all the different APIs that Android or iOS provide to enable apps to deliver a great, integrated experience (sensor access, notifications, health data sharing, etc). We've put a lot of time into thinking about the right analog of those APIs in the context of a personal cloud.

It's open source, self-hostable, zero telemetry, and aims to be as simple and "magic-free" as possible. We (Imbue, the company I work for) also offer a managed version, which I think is really important to making this widely accessible - and it gives us a straightforward business model to support the project. But the open-source/self-hosted path runs exactly the same code and will always be first-class.

We're launching now after building and testing it privately for 6+ months, now that the core platform is mostly complete and stable. I've been slowly moving my digital life into my own Bottle instance, and it's felt great. That said, there's a chicken and egg problem - without a broadly accessible way to self-host open source web software, the audience for such software is small, and the available software is limited. A big goal of ours is catalyzing the creation of more great open source web software - things we can recommend to friends and parents, not just other software nerds. Our curated app catalog is one step in this direction - we aim to maintain a high bar, only adding things that actually deliver a great user experience, so it's pretty small right now. But we're adding things every week - and we hope to grow it faster with the help of the community. For now, early users will likely need a bit of technical familiarity (or a coding agent) to find/create/adapt the things you want to run on your instance (it's really easy!). Over time, as the catalog gets deeper and we polish the experience further, we hope for it to become increasingly accessible to everyone.

Want to check it out?

联系我们 contact @ memedata.com