Nitter fork Shitter offers JavaScript-free Twitter browsing without an account

What surfaced on Hacker News

A fork of Nitter named Shitter reached the Hacker News front page on 23 October 2026. According to the project's README on Codeberg, it is a free and open-source alternative front-end for Twitter, focused on privacy and performance, and it takes its cues from Invidious, the equivalent project for YouTube.

How the front-end works

The core idea is a proxy: every request is handled by the server running the instance, and the visitor's browser never contacts Twitter at all. The README argues this prevents Twitter from seeing a visitor's IP address or collecting a JavaScript fingerprint, because no Twitter code executes in the browser and the pages ship without JavaScript or ads.

The implementation is written in Nim, which the project credits for its small footprint: a page is listed at roughly 60KB, compared with 784KB for twitter.com. The README also claims the result is around 15 times lighter than Twitter on average, with timelines loading two to four times faster — figures that come from the project itself rather than independent testing. Other features include RSS feeds, themes and a responsive layout for mobile devices.

Notably, the front-end talks to Twitter's unofficial API, so an operator does not need a developer account. The project's wiki includes a guide to obtaining the session tokens that instances rely on to make this work.

Why the project says it is needed

The README frames the fork as a response to changes on Twitter itself. As of 2024, it says, an account is required and the site cannot be used without JavaScript. Even behind a VPN and aggressive blocker configuration, a browser can still be tracked through fingerprinting with no JavaScript involved, and the README points to Twitter's removal of user control over advertiser data sharing as a further motivation. Hosting an instance on a personal VPS is presented as the way to browse Twitter while sidestepping all of that.

Deployment and contribution model

Shitter ships with the deployment options familiar from Nitter — the install instructions still refer to the parent project in several places, a telltale of its fork origins. A prebuilt multi-arch Docker image covering amd64, armv7 and arm64 is available, but it needs a separate Redis instance for caching, or Valkey, which the README recommends because Redis is no longer open source as of 2024. Docker Compose and systemd setups are documented, and building from source requires a Nim toolchain plus libpcre and libsass. The project advises running an instance behind a reverse proxy such as Nginx or Apache, and notes that logging is currently limited to errors printed to stdout.

Development happens in an unusual way for a modern forge: the repository does not accept pull requests, and contributors are asked to submit patch files through issues instead. There is a Matrix channel for discussion, and the wiki also lists public instances, other forks of the Nitter codebase and browser extensions that redirect Twitter links automatically.

Why it matters

Shitter is evidence that demand for account-free, JavaScript-free reading of Twitter persists. For technical users, it packages a general pattern — proxy front-ends like Invidious — into something self-hostable with modest resources, and its traction on Hacker News suggests the appetite is still strong. The trade-off is structural: because it depends on an unofficial API and manually obtained session tokens, the whole approach lives at Twitter's sufferance, and a change on the platform's side could break instances overnight. For now, the fork keeps a privacy-preserving route to Twitter content alive for anyone willing to run an instance or trust one operated by someone else.