Add wiki pages back (2nd try)

[Tobi406]

This time, most things do actually work. Content is fetched from /metadata/wiki/:page which fetches from GitHub

TODO:

• Make the scroll listeners on _page.vue useable
• Re-fetch every x hours

[Tobi406]

And I think I just shifted the problem...

[lucko]

:o would be cool if we could ditch having to use a submodule for this

[Tobi406]

Should work now. Still have to do some To-Do things but that is it then

[Tobi406]

And done now, finally 🎉

[Tobi406]

I think I went a little bit crazy, I added

• a server middleware, so we can fetch our data from that instead LuckPerms/metadata-api
• Verbose viewer works
• Tree viewer works

[Tobi406]

I'm gonna wait with the editor until #285 is merged though

[Tobi406]

Typo

[Tobi406]

Add newline

[Tobi406]

I think I just figured out serverMiddleware doesn't run on Netlify (because it isn't static), so I guess we still have a use for the other repo.

[lucko]

Could it be something that we add to the metadata API?

[Tobi406]

It's basicly just an express server (like on the metadata-api) running in nuxt. Nothing serious, can just be added to the metadata-api.

[Tobi406]

As I said already, Netlify only supports static pages.
We can generate everything static, with the exception of the editor etc.
I think we do have the following options right now, feel free to correct me:

• No longer use Netlify and use npm run build && npm run start to run it ourselves
• Just leave this Nuxt thing
• not sure about this: Change the URLs back to using a query or hash.
  • Would break versions using the old URL, even tough we might maybe be able to make a redirect (not sure again)

[Turbotailz]

What is this doing?

[Tobi406]

I think it does the same as nuxt generate. Was more for me to test. Also see my previous comment.

[Turbotailz]

Sorry I've been silent on this, kinda been lurking as you've made updates.

Overall I think you've done a great job so far and thanks for taking the initiative 💪

I would love to continue using Netlify if possible, because then I can still deploy updates to production without needing to nag Luck about it, plus they handle all the annoying stuff like bandwidth, caching, etc. I haven't really had much chance to play around with Nuxt to really get a good idea on what is the best solution here, though.

Personally not a big fan of the HTTP requests for getting the wiki contents but submodules aren't really a great solution either. Can't really add much input at this stage. Would definitely prefer to continue using the metadata API, however. If the wiki does need to be queried via HTTP then we can add it to that instead.

Change the URLs back to using a query or hash

Definitely not keen on this haha.