TL;DR
Get business pricing on networking and server gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A developer writing on dbushell.com on Oct. 3, 2026, described returning to Node.js after using Deno for years. The author said a static-site generator migration required few code changes and produced builds that were 15% faster in their test, while also citing problems with Deno tools and services.
A developer says they moved a SvelteKit client project and a static-site generator from Deno to Node.js. In an account published Oct. 3, 2026, the author reported that the generator needed few code changes and built 15% faster in their test. That result applies to one project and is not a general benchmark.
The author said the static-site generator needed two main adjustments: replacing Deno’s file-system API with Node’s node:fs and replacing Deno.serve with Hono’s Node adapter, which wraps node:http. They also switched from Deno’s @std/path package to Node’s built-in node:path import. The author wrote that the codebase still favored Deno conventions, and said further performance work might be possible.
For package and version management, the author chose Fast Node Manager (FNM) to switch Node versions and pnpm as the package manager. They said pnpm’s controls for install scripts and release timing informed that choice, and described setting a one-day minimum release age after a one-month setting caused dependency resolution problems. The author also used npm and npx aliases pointing to pnpm and pnpx to accommodate scripts and copied installation instructions.
The report also described a limitation the author encountered when using TypeScript with Node: Node’s documentation says type stripping is unsupported for TypeScript files inside node_modules. The author said that meant bundling packages written in TypeScript, and chose tsdown for that task. They also reported configuring pnpm’s trust policy after self-hosting a Forgejo instance, saying their packages no longer had registry provenance.
A Migration With Few Code Changes
The account offers a practical example of how a developer’s runtime choice can shift as the tools change. The author’s experience suggests that a project built around Deno APIs may not require a broad rewrite to run on Node.js: in this case, the reported work centered on file-system and HTTP-server interfaces, plus a path import. That may be useful to developers weighing maintenance costs, though one migration cannot establish how easy the change would be for other codebases.
The reported 15% faster builds give the switch a measurable result, but the source does not describe a controlled comparison, the hardware, repeated runs, or the exact measurement method. It should be read as the author’s observation about one static-site generator, not evidence that Node.js generally builds projects faster than Deno. The author also said the project retained Deno-oriented code, leaving open whether further changes would alter the result.
The package-management choices highlight another concern for teams: dependency security and reproducibility. The author described using pnpm settings to delay newly published releases and restrict downgrades. Those are the author’s precautions, not proof that pnpm or any other package manager prevents malware. The report’s value is in making the trade-offs and setup work visible rather than establishing a security verdict on either ecosystem.
As an affiliate, we earn on qualifying purchases.
The author said Node.js had become more capable since their earlier experience with it, citing support for modern ECMAScript features and improved APIs. They had continued using Deno partly through habit and familiarity, but said a SvelteKit client project during the month before publication gave them a reason to work with Node again. The report reflects that developer’s assessment of Node.js v26.10.0, the version named in the account.
The author’s criticism of Deno was based on personal experience and opinion. They cited weeks of broken ZSH integration, repeated HTTP 429 responses from JSR, and bugs they said affected concurrent HTTP requests. They also criticized Deno Land Inc.’s direction and reported layoffs, but the article does not independently document those claims. The author said they removed Deno with Homebrew and requested deletion of their JSR account; they noted that older package versions remained installable.
“Node got a glow-up, wow!”
— The author of the dbushell.com report
As an affiliate, we earn on qualifying purchases.
What the Migration Test Cannot Show
The report does not provide benchmark details for the 15% build-time difference, including the test conditions, baseline run, or whether the result was repeated. It is not clear whether the result would hold for other workloads or after the author optimizes the remaining Deno-style code. The account also does not compare Node.js and Deno across a wider set of applications.
The reported Deno service and integration issues are the author’s account; the report does not include responses from Deno or JSR, or independent confirmation of the scope and duration of those problems. It also does not establish whether the issues have since been fixed. The author’s criticism of Deno’s company strategy is opinion, while specific claims such as the extent of layoffs are not supported in the supplied report with independent documentation.
As an affiliate, we earn on qualifying purchases.
Further Node.js Tuning Planned
The author said they may explore additional Node.js built-in APIs to improve the migrated generator, but did not give a schedule or promise a follow-up. The immediate next step, according to the report, is continued use of Node.js on the project and possible performance work. Whether that changes the build result remains unknown.
For readers considering a similar move, the account identifies the areas that required attention in this case: file-system calls, HTTP serving, path utilities, TypeScript packaging, and dependency-management settings. It does not offer a universal migration estimate. Any decision to switch runtimes would depend on a project’s APIs, dependencies, operational needs, and independently tested performance.
As an affiliate, we earn on qualifying purchases.
Key Questions
What prompted the developer to switch from Deno to Node.js?
The author said they used Node.js for a SvelteKit client project and found its current APIs and ECMAScript support better than they remembered. They also cited problems they experienced with Deno’s ZSH integration, JSR rate limits, and concurrent HTTP requests.
How much of the static-site generator had to change?
The author reported replacing Deno’s file-system API with node:fs, moving from Deno.serve to Hono’s Node adapter, and swapping Deno’s @std/path for node:path. The report describes the work as limited but does not publish a full code diff.
Does the report prove Node.js is faster than Deno?
No. The author reported 15% faster builds for one migrated static-site generator. The report does not provide benchmarking methods or evidence that the result applies to other projects.
Why did the author use pnpm instead of npm?
The author said they chose pnpm for controls including blocking post-install scripts and delaying installation of newly released packages. They configured a one-day minimum release age after a one-month setting caused dependency resolution problems. These are the author’s stated reasons and configuration choices.
Is the developer continuing to use Deno?
The author said they had uninstalled Deno with Homebrew and described the switch as a goodbye. The report does not say whether they will use Deno again for future projects.
Source: hn
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
