I've made some big changes to Jinear's shape. The short version: I cut features that pulled focus away from the core, and I split the old web app into two purpose-built frontends. Here's what changed and why.
What I removed
To keep Jinear focused on tasks, calendar, and file storage, I retired these features:
- Projects. The project layer is gone altogether. Jinear now centers on tasks, calendar, and files without the extra project grouping on top.
- Project Pages. Each project used to get a public/restricted feed, like a project's own profile or blog. The backing service is archived.
- Custom domains on project feeds. With Pages gone, the on-demand TLS and Cloudflare DNS machinery that served custom project domains went with it. The gateway now runs stock Caddy.
- Chat. Real-time messaging and its websocket service have been removed.
They were the wrong fit. Self-hosted small teams tend to already live in an existing messaging platform, and expecting 3 to 5 people to move to another one just isn't realistic, so chat had to go. The same logic applied across the board: I'd rather do less, well. Instead of piling on more features, I want to polish the ones that are left, the core flows and the design, until they feel right.
Two frontends instead of one
The old jinear-web was a single Next.js app doing two very different jobs:
serving the logged-in product and the public marketing pages. I split it.
- jinear-app is the product, a Vite + React 19 single-page app, served as an installable PWA. It works offline, installs to your home screen, and ships its own manifest. The PWA install also replaces the old WebView mobile wrapper, so there's one less thing to maintain.
- jinear-site (this site) is the public face, marketing pages, pricing, and the blog, built as a fully static Next.js export. Real pre-rendered HTML per route means better SEO and friendlier crawling for AI bots.
The blog also moved here, from the old blog.jinear.co subdomain to /blog.
I didn't find moving old blog posts here helpful since they are mostly about
features that are gone, but they can still be accessed on Jinear's X page
or my X.
Why it's better this way
Splitting the apps lets each one do its job well. The product app can lean into offline-first PWA behavior without dragging marketing SEO concerns along; the marketing site can be pure static HTML without bundling the app's auth and state machinery. Self-hosters get a smaller, simpler stack, fewer services, stock Caddy, no archived dependencies.
jinear-web is still around during the transition, but new work happens in
jinear-app and jinear-site. Leaner, faster, and easier to run yourself.