Ask HN：Azure DevOps 实际上已经停止维护了吗？
1 分•作者: sam256•15 天前
我并不讨厌 Azure DevOps。相反，在过去的 5 年多里，它一直是我们在组织中使用的绝佳工具。它相当容易进行安全设置，我们有大量优秀的流水线，可以检查我们的代码、进行部署、与第三方服务集成等等。我们对组织有很好的可见性。
但……我忍不住觉得它跟不上 GitHub 的发展，现在面临一个问题：如果想保持竞争力，我们是不是别无选择，只能痛苦地迁移到 GitHub Enterprise？
具体来说，我们感受到的痛点包括：
1. Copilot PR 审查在哪里？在 GitHub 中，我只需点击一个按钮，就能通过 Copilot 立即获得对任何 PR 的初步审查。当然，它并不完美，但它几乎是免费的。
2. MCP？好吧，DevOps 终于有了 MCP 服务器，而且它实际上相当不错。但它花了很长时间才推出。这会是一个重复出现的模式吗？我们是否需要等待一年才能获得 GitHub 用户立即使用的工具？
3. Claude Code。在 GitHub 上，我可以从任何地方（浏览器、桌面、手机）给 Claude Code 分配任务，它会在一个小小的开发容器中运行，然后带着一个 PR 回来。同样，它并不总是完美的，但初级开发人员的 PR 也是如此。而且 Claude Code 不会在意我是否在周五晚上给它分配任务。
所以，标题中的问题是：ADO 实际上是否已经到了生命周期终点（EOL）？
我希望答案是否定的，我们可以坚持下去，直到情况好转。但我开始担心，这仅仅是沉没成本谬误。有人能提供一些希望吗？
查看原文
I don't hate Azure DevOps. On the contrary, it's been a phenomenal tool for our organization over the past 5+ years. It's been fairly easy to secure, we have tons of great pipelines that check our code, do deploys, integrate with third-party services, etc. We get great visibility into our organization.<p>But... I can't help but feel like it's not keeping up with GitHub and am now facing the question: do we basically have no choice but to do a painful migration over to GitHub Enterprise if we want to stay relevant?<p>In particular, pain we've felt:<p>1. Where's the Copilot PR review? In GitHub I just click a button and I get an instant first level review of any PR by Copilot. Sure, it's not perfect, but it's more or less free.<p>2. MCP? Ok there finally is an MCP server for DevOps and it's actually pretty good. But it took forEVER to come out. Is this a pattern that's going to repeat? We have to wait a year for tools GitHub users get right away?<p>3. Claude Code. On GitHub I can give Claude Code assignments from anywhere -- browser, desktop, phone -- and it runs off in a little dev container and comes back to me with with a PR. Again, not always perfect, but neither are PRs from junior devs. And Claude Code doesn't care if I give it the assignment on a Friday night.<p>So the titular question: Is ADO effectively EOL?<p>I'm hoping the answer is no, and we can just hold on till things get better. I'm starting to worry, though, that this is just the sunk cost fallacy. Anyway have hope to offer?