Modernise website technology stack

[achrinza]

The current loopback.io website stack leverages Jekyll. Although it has, and continues to, meet our needs for hosting a simple website, there are certain issues which makes it unsustainable in the long-term:

1. Unfamiliarity of Ruby and Jekyll among the maintainers
2. Slow build times
3. Inability to use component-based, modern front-end development workflows
4. Legacy baggage

In alignment with the earlier blog modernisation effort, this issue is to track rewriting the current website in a modern, Node.js-based stack.

Currently, prototyping work is being done with Next.js.

[dhmlau]

I wonder.. since our blog site is using Docusaurus, would we also consider that for this site as well?

[GrooveCS]

@achrinza

1. Is this still an issue that we want to address?
2. If yes, could you let me know if your proposed solution is still valid today, or would you like to modify it?

Please let me know what you think. P-)

[achrinza]

Hi @GrooveCS! I don't think the discussion has expanded beyond what's stated above.

It's still an valid issue, as I think the LoopBack.io codebase has a lot of baggage from hacks and scripts that pull data from elsewhere.

The main objective is to move away from Ruby Jekyll and onto JS so that it's easier for us to extend in the future.

I initially proposed Nextjs at the time as that's what I was familiar with. However I wonder if using Docusaurus or even Eleventy would be more suitable?

Great thing about Eleventy is the granular control over the output, which means we don't have to adapt our existing documentation structure to fit a model.

Docusaurus uses React under the hood, which may mean access to more comptehensive pipeline tools. But I'm not certain if we need the baggage of a frontnend framework.