nextra VS typedoc

co-author here

we put in a lot of effort into our docs and we'd greatly appreciate any criticism or feedback! Langfuse is powerful but the docs should help beginners to quickly get started and then incrementally use more features.

docs are OSS, repo: https://github.com/langfuse/langfuse-docs

built using: https://github.com/shuding/nextra

However, this may just be due to the lack of proper documentation from the Nextra side of things (shoutout to Nextra though, regardless).

Nextra - Nextra is another option for creating documentation sites. While it might not be as well-known as Docusaurus, Nextra offers a modern and minimalist approach to building documentation. It is designed to be lightweight and user-friendly, making it a good choice for those who prefer a simple and clean documentation style. You can explore more about Nextra on their official website.

I have looked at https://nextra.site/ but that doesn't work with the app router yet. So I'm wondering if there's another alternative.

You could also have a look at Nextra. You can use mdx components to build your blog (including support for server-side fetching). I'm currently using their documentation template, but it seems they also have a blog template.

We write everything in Markdown, as it's the closest you'll get to a 'universal' format. Then, we use a static site generator to turn the docs into a website. Current projects are using Nextra for this. If you ever need to change site generators, you still have all the markdown docs and image files, so it's pretty easy to change.

https://nextra.site/ is nice

Today I found this tool for Next.js called Nextra. You can effortlessly create a blog post website or a documentation website. All you need is markdown. Simply export your markdown from Notion and utilize it with Nextra to enjoy all the cool features, including full-text search, syntax highlighting, dark/light mode, and even image support. Everything is generated at build time, making it a static website which is Blazingly fast. https://nextra.site/

Firstly, install Typedoc using npm:

Background:The main bottleneck with the TypeDoc default theme especially for large projects is the verbose HTML for the left-hand navigation that linearly grows for each page based on the project size and consumes a massive amount of disk space; see this TypeDoc issue. The DMT caches the left-hand navigation HTML and dynamically creates a shared web component that is utilized across all pages only making a single copy of the navigation HTML. This reduces disk space utilization by up to 90% and also makes doc generation ~80% faster. I also include some style additions and replace the main search index generation using compressed MesssagePack instead of JSON which reduces the search index size by more than 90%.

While the Prim+RPC server is expected to be JavaScript, I'd like to support other languages through JSON Schema. I wrote a tool that translates TypeDoc comments into RPC-specific documentation. My plan is to turn this result into JSON Schema that can be served with the Prim+RPC server. This means you can get typed suggestions (for instance, from an IDE that understands JSON Schema) when writing requests in JSON files (I wrote a little about this here, still a WIP). From this, you could use your favorite HTTP client in the language of your choice, like but still benefit from having typed requests.

Docusaurus has a Typedoc plugin. Also there is a typedoc markdown plugin.

Finally, JSDoc can be used to generate documentation for your code using tools like JSDoc itself and TypeDoc. These tools generate HTML or Markdown documentation based on your JSDoc annotations, making it easier for others to understand how your code works and how to use it.

Since you're using TypeScript, use TypeDoc.

Generates HTML documentation using TypeDoc.

I was thinking of using something like https://typedoc.org to do it, do you have experience with this sort of tools?

TSDoc is more consistent, has cleaner documentation, better tooling (e.g. TypeDoc or ESLint plugin) and better support for data structures (e.g. straightforward enums support).