ollytom
commented
8 months ago The server fetches data from Lemmy via its HTTP API then renders HTML from templates; see https://github.com/rystaf/mlmym/tree/main/templates
Something needs to make a HTTP request to Lemmy.
mlmym could be modified slightly to run from Wasm and not start a HTTP server. In this case requests to Lemmy could be made by the browser rather than the server (just like phtn).
Hey there,
I've been using this project and running it locally, (not in the docker, just straight from source) - and I was curious whether this actually really needed a running server. What this project mostly does is serve a front-end.
It's also possible to compile go into to WASM, and just serve html from there.
I was trying to do that myself from source, but ran into some obstacles. But that's because I'm not super familiar with GO yet - though I have build a bunch websites in Blazor (C# WASM) that work in a similar faction there C#/Blazor is just generating html client size
The main reason/benefit to do it this way would be so that it can be hosted anywhere without a server, for example AWS S3 or even Github Pages in a similar fashion as React or Blazor based front-ends. - and so without any/neglectable costs
I noticed it uses the server to proxy the login request back into itself, but that isn't really necessary anymore. If you look at https://phtn.app for example, a different alternative Lemmy Front-end, it just sends the login request straight to
https://programming.dev/api/v3/user/login.Was wondering what other people though of this, specifically maintainers / contributes to this project (@rystaf) - Or if there would be reasons why this wouldn't possibly work in WASM :)