Cloudflare Acquires Deno and Plans Sunset as Node.js Wins Runtime War

Fireship

Cloudflare will shut down Deno after integrating its CellD architecture into workerd, leaving Node.js as the sole major JavaScript runtime free of corporate cloud control.

AI-assisted summary of The runtime wars are over... and Node won — original source. Check the original for full context.

Source published: 2026-10-15

Key ideas

Cloudflare has acquired Deno, marking the end of the runtime wars as creator Ryan Dahl admitted that attempting to re-implement Node.js from scratch was a mistake. (0:00)
The acquisition spells the sunset of the Deno platform: active development will cease after 12 months of bug and security patches, giving Deno Deploy users roughly six months to migrate. (2:46)
Cloudflare purchased Deno primarily to acquire CellD, an open-source implementation of Durable Objects that will be merged into the workerd runtime to enable native self-hosting. (2:46)
With Anthropic owning Bun and Cloudflare acquiring Deno, Node.js remains the sole foundational JavaScript runtime not owned by a hyperscale corporate cloud provider. (2:46)

Detailed notes

## Background Published on YouTube by Fireship on October 15, 2026, this video covers Cloudflare's acquisition of Deno and the effective conclusion of the JavaScript runtime wars. According to the description, Node.js creator Ryan Dahl has conceded that attempting to reimplement Node from scratch was a mistake, while also introducing a sponsorship for JetBrains Junie Lite.

## TL;DR The fierce rivalry among JavaScript runtimes has reached a decisive finish with Cloudflare acquiring Deno, marking an end-of-life roadmap for the alternative runtime. Although Deno originally sought to correct Node's architectural flaws with native TypeScript and built-in security, Node rapidly eliminated those competitive advantages by adopting those very features itself. Facing stalled commercial traction and shrinking deployment infrastructure, Deno pivoted to building CellD—an open, self-hostable alternative to Cloudflare's Durable Objects aimed at stateful AI agent workflows. Cloudflare acquired Deno to incorporate that self-hosting technology into its own runtime, giving existing Deno Deploy customers six months to migrate before development is discontinued entirely.

## Atmosphere & style The video carries Fireship's signature rapid-fire, cynical, and meme-literate delivery, dissecting the harsh realities of startup life and preference stacks with deadpan humor. It balances tech-insider satire about legal fights over trademarks with genuinely incisive architectural analysis of distributed state and durable objects. The narrative feels like an entertaining post-mortem, tinged with poetic irony as Node outlasts the very project created to replace it.

## Bulletin - [00:00](https://www.youtube.com/watch?v=l3L_lETiSzc&t=0s) Cloudflare has acquired Deno, marking the end of the runtime wars as creator Ryan Dahl admitted that attempting to re-implement Node.js from scratch was a mistake. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) The acquisition spells the sunset of the Deno platform: active development will cease after 12 months of bug and security patches, giving Deno Deploy users roughly six months to migrate. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Cloudflare purchased Deno primarily to acquire CellD, an open-source implementation of Durable Objects that will be merged into the workerd runtime to enable native self-hosting. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) With Anthropic owning Bun and Cloudflare acquiring Deno, Node.js remains the sole foundational JavaScript runtime not owned by a hyperscale corporate cloud provider.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=l3L_lETiSzc&t=0s) Ryan Dahl concluded that rewriting Node.js from the ground up was a mistake, leading to Deno's acquisition by Cloudflare despite Cloudflare already operating its own edge runtime. - [00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) Deno was conceived at JSConf EU 2018 to fix Dahl's regrets about Node.js by offering native TypeScript, out-of-the-box tooling, and a secure permissions sandbox. - [00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) Node.js neutralized Deno's competitive advantages by implementing native fetch, direct TypeScript execution, and its own permission security models. - [00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) Commercial traction for Deno's hosted offerings steadily deteriorated, forcing Deno Deploy to contract from 35 geographic regions down to six. - [00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) Key Deno initiatives stalled, including Deno KV remaining stuck in beta, Fresh framework updates lagging, and resources being diverted to a trademark battle against Oracle. - [00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) Deno became trapped in an unsustainable treadmill of constantly chasing Node compatibility rather than driving runtime innovation. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Deno pivoted to distributed state coordination by releasing CellD, an open-source clone of Cloudflare's Durable Objects designed for private SQLite storage. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Cloudflare acquired Deno after its own internal efforts to build a self-hostable workers implementation failed earlier in the year. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Ryan Dahl and Bert Belder will integrate CellD directly into Cloudflare's open-source workerd runtime to power stateful, long-lived autonomous AI agent architectures. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Cloudflare is phasing out the Deno product line, committing only to 12 months of maintenance patches before abandoning development, with Deno Deploy users facing a six-month migration window. - [02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Recent corporate buyouts—including Anthropic purchasing Bun and Cloudflare taking Deno and Vite—leave Node.js as the last major independent JavaScript runtime.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=l3L_lETiSzc&t=0s) The End of the JavaScript Runtime Wars

The JavaScript runtime wars have drawn to an unexpected conclusion: Ryan Dahl, the creator who brought JavaScript to the server with Node.js and subsequently spent a decade attempting to atone for its shortcomings, announced that his company, Deno, has been acquired by Cloudflare. On the surface, the transaction presents like standard Silicon Valley corporate theater, complete with founders voicing enthusiasm for the future, the acquiring enterprise expressing excitement for shared goals, and employees abruptly discovering the harsh mechanics of preference stacks. Yet the underlying premise is counterintuitive: Cloudflare already possesses its own mature edge runtime and deployment network, raising immediate questions about why it would purchase an ostensibly competing platform.

The rationale behind the purchase reveals a shift in priorities: Ryan Dahl has arrived at the realization that attempting to re-implement Node.js from scratch was a mistake, and the resulting trajectory for the Deno platform itself looks bleak. As host Fireship frames the situation: “And on paper, it looks like a pretty traditional Silicon Valley acquisition where the founder says they're excited for the future. The company says they're excited about the mission and the employees learn the hard way what a preference stack is. But if you think about it, it's a little weird that a company which already has its own runtime and deployment platform would buy a worse one.”

[00:44](https://www.youtube.com/watch?v=l3L_lETiSzc&t=44s) The Rise and Fall of Deno's Momentum

To trace how the industry arrived at this point, one must look back to 2009, when Ryan Dahl took the stage at JSConf EU to unveil Node.js, fundamentally altering backend development. Nine years later, he returned to that exact stage with his famous retrospective, “10 Things I Regret About Node.js.” Rather than merely critiquing his prior work, he revealed Deno—a runtime built from scratch designed to eliminate those original architectural regrets. Unlike Node, Deno shipped out-of-the-box with a built-in linter, formatter, and test runner, alongside native TypeScript execution, a secure-by-default sandbox permissions architecture, and a modern URL-based external module system.

```mermaid flowchart TD accTitle: Deno Growth and Decline Lifecycle accDescr: Chronological timeline of Deno from initial 2018 reveal through 2022 funding and subsequent stagnation.

subgraph Growth["Expansion (2018 - 2022)"] A["2018: Regret talk reveals Deno"] --> B["2020: Deno 1.0 Release"] B --> C["2021: $5M Seed & Deno Deploy"] C --> D["2022: $21M Series A (Sequoia) & Fresh Framework"] end

subgraph Stagnation["Decline & Node Counter-Offensive"] D --> E["Node adds native fetch, TypeScript & permissions"] E --> F["Deno Deploy shrinks from 35 to 6 regions"] F --> G["Deno KV stalled in beta; Fresh releases slow down"] G --> H["Legal battle against Oracle over 'JavaScript' trademark"] end ```

Deno 1.0 launched in 2020, followed in 2021 by the introduction of Deno Deploy and a $5 million seed funding round. Commercial adoption seemed promising: Netlify and Supabase built their serverless edge functions on top of Deno's architecture, while Slack selected it to power their revised application platform. By 2022, Deno secured a $21 million Series A led by Sequoia Capital and rolled out Fresh, its bespoke server-side web framework.

However, Deno's competitive edge evaporated rapidly as Node refused to remain stagnant. Instead of rolling over, Node incorporated native `fetch`, developed its own permission security models, and added direct execution for TypeScript files, erasing Deno's key selling points. Simultaneously, Deno’s monetization strategy faltered: - Deno Deploy contracted from a distributed presence across 35 regions down to only six. - Deno KV, launched in 2023, languished and never exited beta testing. - The Fresh framework stagnated, enduring multi-month gaps between updates. - Dahl directed resources toward a legal dispute against Oracle to contest the “JavaScript” trademark (inherited from Sun Microsystems in 2009)—a well-meaning ideological campaign, but hardly the commercial enterprise Sequoia anticipated when writing a $21 million check.

Sentiment worsened until developer David Bushell published a critique characterizing Deno as effectively dead, trapped in an endless treadmill of shipping Node compatibility patches rather than innovating. Dahl privately reached that very same conclusion.

[02:46](https://www.youtube.com/watch?v=l3L_lETiSzc&t=166s) Durable Objects, CellD, and the Cloudflare Buyout

Confronting this reality, the Deno engineering team pivoted away from runtime-level mechanics to focus on higher-level distributed systems challenges: cross-machine compute distribution, state coordination, and scaling. Their investigation pointed toward Cloudflare’s Durable Objects—lightweight, single-threaded execution instances bound to isolated, private SQLite storage engines with persistent WebSocket communication capabilities.

The fundamental limitation of Cloudflare's Durable Objects was their proprietary lock-in within Cloudflare's ecosystem. In response, Deno engineered CellD in August, an open alternative designed to execute durable object primitives on private hardware at a fraction of hosted infrastructure costs.

Far from taking legal or competitive issue with being cloned, Cloudflare welcomed the project. The creator of Cloudflare Workers highlighted that while the workers runtime had been open-sourced in 2022, the company had spent years wanting a fully self-hostable implementation, even confessing that their own internal attempts to build one the previous spring had failed. Cloudflare, wielding a $123 billion market cap, resolved the issue by acquiring Deno outright.

Under the acquisition plan, Ryan Dahl and co-founder Bert Belder will integrate CellD directly into Cloudflare’s open-source `workerd` runtime, providing native, first-party support for self-hosted workers and durable objects. Dahl identified autonomous AI agents as the primary rationale for the technology, pointing out that agentic workflows demand economical compute, long-lived local state, and persistent real-time connections.

For users invested in the broader Deno product line, the acquisition marks an end-of-life notice: - Cloudflare will limit Deno maintenance strictly to security patches and bug fixes for 12 months, followed by a total cessation of development. - Active customers utilizing Deno Deploy are given roughly a six-month window to migrate workloads off the platform. - Following Anthropic's acquisition of Bun and Cloudflare's purchase of Deno (alongside its earlier acquisition of Vite in June), Node.js remains the sole foundational JavaScript runtime not controlled by a hyperscale corporate cloud provider.

As Fireship summarizes the stark fallout for developers on the platform: “Cloudflare says they'll only continue to ship bug fixes and security patches for Dino for the next year. But after that, they'll stop development altogether. And if you're using Dino deploy today, you have about 6 months to migrate.”

[04:42](https://www.youtube.com/watch?v=l3L_lETiSzc&t=282s) Sponsor: JetBrains Junie Lite

For developers facing migrations away from sunsetting infrastructure, host Fireship introduces the video's sponsor, JetBrains. Joking about the anxiety of token billing when orchestrating AI agents, the host highlights Junie Lite, a baseline model included across all JetBrains accounts without cost.

- **Always Free:** Operates directly inside the IDE and command-line interface without countdown trial limitations. - **Automation Use Cases:** Well-suited for maintenance tasks such as renaming functions and generating test scaffolding across legacy projects without accumulating frontier API expenses. - **Workflow Integration:** Junie Lite integrates directly into JetBrains AI, enabling developers to spin up multiple parallel agents, review proposed code modifications prior to merging, and step up to frontier models when complex reasoning is required.

## Mentioned Fireship, Ryan Dahl, Bert Belder, David Bushell, Node.js, Deno, Deno Deploy, Deno KV, Fresh, CellD, Cloudflare, Cloudflare Workers, Durable Objects, workerd, Bun, Anthropic, Vite, JetBrains, Junie Lite, Netlify, Supabase, Slack, Sequoia Capital, Oracle, Sun Microsystems, JSConf EU, TypeScript, SQLite

## Video description Every JetBrains account now comes with a free coding model called Junie Lite. Try it: https://jb.gg/JunieLite

Cloudflare just acquired Deno and the creator of Node.js is finally admitting that reimplementing it was a mistake. Let's dive in.

Want more Fireship?

🗞️ Newsletter: https://bytes.dev 🧠 Courses: https://fireship.dev

[Source](https://www.youtube.com/watch?v=l3L_lETiSzc)

中文简报

Cloudflare收购Deno并计划停运，Node.js赢得运行时之争 Cloudflare将在把CellD架构整合入workerd后关闭Deno，使Node.js成为唯一摆脱巨头云厂商控制的主流JavaScript运行时。随着Cloudflare收购Deno，JavaScript运行时之间的激烈竞争迎来决定性结局，这也标志着该替代运行时进入生命周期终结阶段。尽管Deno最初试图通过原生TypeScript支持和内置安全机制来修正Node的架构缺陷，但Node自身迅速跟进并吸收了这些特性，从而消除了Deno的竞争优势。在商业化进展受阻且部署基础设施萎缩的情况下，Deno转向构建CellD——一种旨在用于有状态AI Agent工作流、开源且支持自托管的Cloudflare Durable Objects替代方案。Cloudflare收购Deno正是为了将该自托管技术融入自身运行时，并为现有Deno Deploy客户提供6个月的迁移缓冲期，之后开发工作将彻底停止。 Cloudflare已收购Deno，标志着运行时之争走向终结，创始人Ryan Dahl也承认从头重新实现Node.js是一个错误。 (0:00) 此次收购意味着Deno平台即将停运：在提供12个月的缺陷修复与安全补丁后将停止积极开发，Deno Deploy用户大约有6个月的迁移时间。 (2:46) Cloudflare收购Deno主要是为了获取CellD，这是一个Durable Objects的开源实现，将被合并到workerd运行时中以支持原生自托管。 (2:46) 随着Anthropic拥有Bun以及Cloudflare收购Deno，Node.js成为了唯一未被超大规模企业云厂商所拥有的基础JavaScript运行时。 (2:46)

Reading summary

According to Fireship, the JavaScript runtime wars have concluded with Node.js preserving its supremacy as Cloudflare acquires Deno and prepares to sunset the competing platform. The buyout ends Ryan Dahl's effort to replace Node from scratch, redirecting his team's focus toward distributed, self-hosted primitives for AI workloads. Node.js Defends Its Lead Against Deno (0:44) Ryan Dahl introduced Deno in 2018 to resolve foundational architectural regrets in Node.js, offering out-of-the-box TypeScript support, sandbox security permissions, and native developer tooling. The project initially generated strong commercial momentum, securing a $5 million seed round in 2021, a $21 million Series A from Sequoia Capital in 2022, and adoption from platforms such as Slack, Netlify, and Supabase. However, Node.js responded by implementing native fetch, direct TypeScript execution, and security permission models, eroding Deno's primary technical differentiators. Deno subsequently struggled to maintain commercial traction. Deno Deploy scaled back operations from 35 geographic regions to six, Deno KV remained stalled in beta, and the Fresh web framework saw prolonged lulls between updates while company resources were diverted into an ideological trademark dispute against Oracle. Facing developer sentiment that Deno had become trapped in an endless cycle of writing Node compatibility patches, Dahl ultimately concluded that trying to re-implement Node was a strategic mistake. The Strategic Pivot to Distributed CellD (2:46) In response to its runtime stagnation, Deno pivoted engineering toward distributed systems architecture, focusing on compute distribution, state coordination, and machine-to-machine scaling. The team targeted the capabilities of Cloudflare's Durable Objects—single-threaded execution environments paired with isolated SQLite databases and persistent WebSockets—which had previously been confined to Cloudflare's proprietary hosting infrastructure. In August, Deno released CellD as an open-source alternative capable of executing durable object primitives on private hardware. Cloudflare, which had open-sourced its workerd runtime in 2022 but failed during internal attempts to engineer a self-hostable durable objects implementation, acquired Deno to incorporate CellD directly into workerd. Founders Ryan Dahl and Bert Belder will lead this upstream integration. Dahl highlighted autonomous AI agent workflows as the principal driver for the architecture, asserting that agentic software requires cost-effective local compute, long-lived state, and continuous real-time connectivity. Corporate Consolidation and the Deno Sunset (0:00) The acquisition initiates a phase-out of the standalone Deno platform. Cloudflare plans to restrict Deno maintenance to security patches and bug fixes for 12 months before ceasing development entirely, leaving active Deno Deploy customers an estimated six-month window to migrate their workloads. Fireship notes that the transaction reflects broader consolidation across the modern JavaScript ecosystem. With Anthropic acquiring Bun and Cloudflare acquiring both Vite and Deno, Node.js remains the sole foundational JavaScript runtime governed independently from hyperscale cloud corporations. The acquisition concludes Deno's attempt to unseat Node.js, highlighting how entrenched ecosystem network effects forced a pivot toward distributed cloud primitives, while leaving existing Deno users facing mandatory infrastructure migrations.

精炼阅读摘要

Fireship 报道指出，随着 Cloudflare 正式收购 Deno 并计划在一年后终止其后续开发，长达数年的 JavaScript 运行时之争实质上以 Node.js 的最终胜利告终。面对 Node.js 快速弥补功能缺陷以及商业化困局，Deno 转向分布式状态技术 CellD，并最终被 Cloudflare 整体吸收。竞争优势瓦解与商业化困境 (0:44) 2018年，Node.js 创始人 Ryan Dahl 针对自己早期设计的不足创立了 Deno，主打内置工具链、原生 TypeScript 支持、沙箱权限管理及基于 URL 的模块体系，并在 2022 年获得了红杉资本领投的 2100 万美元 A 轮融资。然而，Node.js 迅速采取反制，直接在自身架构中集成了原生 fetch 支持、权限控制和 TypeScript 直接执行能力，从而抹平了 Deno 最初的核心竞争壁垒。与技术优势流失相伴的是 Deno 商业化路线的停滞。Deno Deploy 的节点规模从原先的 35 个萎缩至仅剩 6 个；于 2023 年推出的 Deno KV 始终停留于测试版（beta）；其官方全栈框架 Fresh 的更新陷入了长达数月的停滞；同时 Dahl 还耗费大量资源与甲骨文就“JavaScript”商标权发起诉讼。开发者 David Bushell 公开批评 Deno 已经名存实亡，彻底陷入了不得不持续追赶 Node 兼容性补丁的繁重维护陷阱中，而这一严峻现实也促使 Dahl 放弃从零重构 Node.js 的原有设想。技术突围：Durable Objects 与 CellD (2:46) 在底层运行时层面陷入同质化竞争后，Deno 研发团队将核心方向转向更高维度的分布式系统难题，重点解决跨节点的计算分发、持久状态协调与弹性扩展。团队将目光投向了 Cloudflare 的 Durable Objects 模式——一种将单线程计算实例、私有 SQLite 存储与长生命周期 WebSocket 双工通信紧密绑定的边缘状态架构，但该模式一直受限于 Cloudflare 封闭的专有生态。为突破供应商锁定，Deno 于 8 月份研发出名为 CellD 的开源替代方案，允许开发者在私有硬件与自建基础设施上运行类似 Durable Objects 的原语，极大降低了托管成本。Cloudflare Workers 架构团队虽然早在 2022 年便开源了运行时，但在自建可自托管版本的尝试中遭遇挫败。面对 CellD 提供的完整自托管方案，市值达 1230 亿美元的 Cloudflare 选择直接出资收购 Deno，以此解决自身长期的架构短板。平台停运与云巨头对运行时的收编 (2:46) 该收购案标志着独立 Deno 平台的生命周期步入尾声。按照 Cloudflare 的整合路线图，Deno 创始人 Ryan Dahl 与 Bert Belder 将把 CellD 核心组件直接合流至 Cloudflare 的开源 workerd 运行时中，使官方能够直接提供针对自托管 Workers 和 Durable Objects 的原生支持。Dahl 指出，这项技术的主要目标场景是自主 AI Agent，因为智能体工作流在本质上高度依赖廉价计算、持久的本地私有状态以及常驻的实时连接。然而，对于现有的 Deno 生态用户而言，这无异于一张官方停服通告：Cloudflare 仅为现有 Deno 平台提供为期 12 个月的纯安全补丁与错误修复，期满后将全面终止一切开发；目前仍在使用 Deno Deploy 托管业务的活跃客户，则被要求在大约 6 个月内完成向其他平台的业务迁移。继 Anthropic 收购 Bun、Cloudflare 接连收购 Vite 和 Deno 之后，Node.js 成为目前主流生态中唯一一个未被超大规模云服务巨头控股的基础级 JavaScript 运行时。如何在巨头控制的闭环服务与开源独立性之间取得平衡，成为现代 Web 基础设施演进中留给开发者的核心课题。

Original transcript

[0:00] Last week, the JavaScript runtime wars [0:02] officially ended when the guy who put [0:03] JavaScript on the server and then spent [0:05] the next decade atoning for it announced [0:07] that his company Dino was joining [0:09] Cloudflare. And on paper, it looks like [0:11] a pretty traditional Silicon Valley [0:13] acquisition where the founder says [0:14] they're excited for the future. The [0:16] company says they're excited about the [0:17] mission and the employees learn the hard [0:19] way what a Pruff Stack is. But if you [0:21] think about it, it's a little weird that [0:23] a company which already has its own [0:25] runtime and deployment platform would [0:27] buy a worse one. In today's video, we'll [0:29] find out what Cloudflare actually paid [0:31] for, learn why Ryan Doll now thinks [0:33] re-implementing Node was a bad idea, and [0:35] dive into what the future of Dino looks [0:37] like now. And spoiler alert, it's not so [0:39] good. It is October 15th, 2026, and [0:42] you're watching the code report. Back in [0:44] 2009, Ryan Doll got on stage at JSCOM EU [0:48] and introduced the world to a new open- [0:50] source runtime called Node.js that [0:52] allowed you to write JavaScript on the [0:54] server. Nine years later, he went back [0:56] to that same stage with a talk titled 10 [0:58] things I regret about Node, where he [1:00] spent most of the time tearing apart the [1:02] original project before revealing that [1:04] he recreated it from scratch, but this [1:06] time without all the regret. He called [1:07] it Dino and unlike Node, it came with [1:10] its own formatter, llinter, and [1:12] testrunner that supported TypeScript, [1:14] was secure by default, and had a [1:15] sensible way to manage external [1:17] packages. Adino 1.0 0 launched a few [1:19] years later in 2020 and by 2021 Ryan [1:22] announced their serverless platform Dino [1:24] Deploy along with a fresh $5 million [1:27] seed round and initially things were [1:28] going great and Netlefi and Superbase [1:30] built their edge functions on top of [1:32] Dino and Slack used it to power their [1:34] new app platform. The next year in 2022 [1:37] they got 21 million in series A funding [1:39] from Sequoia launched a web framework [1:41] called Fresh and it was looking like [1:42] they had a real shot at becoming the [1:44] future of serverside JavaScript. But [1:46] then their momentum died. Their biggest [1:48] problem being that Node didn't just roll [1:50] over. During those same few years, Node [1:53] added fetch, a permission model, and the [1:55] ability to run TypeScript files [1:57] directly. So, Dino lost its edge. [1:59] Meanwhile, the products that were [2:00] supposed to make money weren't gaining [2:02] enough traction to justify the [2:03] valuation. Adino Deploy went from 35 [2:06] regions down to six. Dino KV launched in [2:09] 2023 and never made it out of beta. And [2:11] Fresh started going months between [2:13] releases. At the same time, Ryan took on [2:15] a legal fight with Oracle over the [2:17] JavaScript trademark, which Oracle [2:19] inherited when it bought Sun [2:20] Microsystems the same year Ryan created [2:22] Node. It was a noble cause, but a [2:24] GoFundMe for trademark lawyers probably [2:26] isn't what Sequoia had in mind when they [2:28] wrote a $21 million check. The sentiment [2:30] got bad enough last year that a [2:32] developer named David Bushell wrote a [2:34] blog post declaring that Dena was [2:35] basically dead since every new release [2:37] was just more node compatibility fixes [2:39] in an endless game of ketchup. And [2:41] ironically, Ryan had come to the same [2:43] conclusion. He just took a little bit [2:45] longer to get there. So earlier this [2:47] year, the team started to look at hard [2:49] problems that live outside the runtime, [2:51] like spreading compute across machines, [2:53] coordinating state, and scaling. And the [2:55] best answer they found was Cloudflare's [2:57] durable objects, which if you're not [2:59] familiar with are basically tiny [3:01] singlethreaded servers that each get [3:02] their own private SQLite database and [3:05] the ability to hold websocket [3:06] connections. But the problem with [3:08] durable objects is that they really only [3:10] work on Cloudflare. So in August, the [3:12] Dino team released their own version [3:14] they called cell D, which lets you run [3:16] durable objects on your own servers at a [3:18] fraction of the cost. And surprisingly, [3:20] Cloudflare was happy to be ripped off [3:22] like this. In the announcement post, the [3:24] creator of Cloudflare Workers said that [3:26] people have always accused Cloudflare of [3:28] using workers to lock in developers, but [3:30] the workers runtime has been open source [3:32] since 2022 and that they've wanted to [3:34] build a self-hostable version for years, [3:36] but just never got around to it. He even [3:38] admitted he tried building one himself [3:39] last spring and he couldn't get it to [3:41] work. So, when the Dino team figured it [3:43] out with Cell D, Cloudflare did what [3:45] anyone with a $123 billion market cap [3:47] would do and they just bought the whole [3:49] company. The plan is to have Ryan and [3:51] co-founder Bert Belder merge celld into [3:54] Cloudflare's open- source workers [3:55] runtime is so that anyone can run [3:57] workers and durable objects on their own [3:59] servers with official support from [4:01] Cloudflare. And because we can't go a [4:03] whole video without talking about it, [4:04] Ryan says the big focus is AI agents [4:06] because durable objects are a natural [4:08] fit for agents that need cheap compute, [4:10] persistent state, and real-time [4:12] connections. But unfortunately, if [4:14] you're using any Dino product that isn't [4:16] CLD, you are screwed. Cloudflare says [4:19] they'll only continue to ship bug fixes [4:20] and security patches for Dino for the [4:22] next year. But after that, they'll stop [4:24] development altogether. And if you're [4:26] using Dino deploy today, you have about [4:28] 6 months to migrate. So in the end, with [4:30] Anthropic acquiring Bun and Cloudflare [4:32] buying Dino after it bought V back in [4:34] June, the only major JavaScript runtime [4:36] that isn't owned by a company with a [4:38] cloud is Node, the one Ryan spent the [4:40] last 8 years trying to replace. So if [4:42] you're one of the Dino developers [4:44] needing to migrate, you can at least [4:45] make the move a little cheaper with Jet [4:47] Brains. the sponsor of today's video. [4:49] Thanks to my childhood financial trauma, [4:51] every time I ask my coding agent to do [4:53] anything, my brain starts compulsively [4:55] calculating how many tokens it will [4:56] cost. So, it's great to see that every [4:58] Jet Brains account now comes with a free [5:00] baseline model called Juny Light that [5:02] lives in your IDE and CLI. There's no [5:05] trial with a countdown clock. It's just [5:06] a free way to ask questions and get [5:08] Grunt work done without accidentally [5:10] bankrupting your family. I used it in [5:12] the Juni CLI on an old side project to [5:15] rename a bunch of functions and scaffold [5:17] some tests and it knocked it all out in [5:19] a few minutes for $0. I'd still switch [5:21] to a Frontier model for more complex [5:23] work, which is really easy because Juny [5:25] Light lives inside Jet Brains Air where [5:27] you can run multiple agents in parallel [5:29] and review all their changes before they [5:31] land. Start using Junyite for free today [5:33] at the link below. This has been the [5:35] code report. Thanks for watching and I [5:37] will see you in the next one.