marver VS comment-castles

marver https://github.com/sylv/marver

My NAS has a load of files dumped onto it over the years. Some are organised neatly, some are in a dumping ground that I promised myself I'd sort through and never did. Some are in old folders that I haven't touched in years. Instead of spending a couple weekends sorting through it like a sane person, I'm making something to index it and hopefully replace Plex in the process.

The goal for marver is to index all those files and rip as much metadata as possible from each, then display that in a pretty UI and give the user multiple ways to refine the metadata, search through their files and make the most of the piled up 1s and 0s they have.

So far I've mostly just been having fun making individual pieces - image search powered by CLIP, extracting information from hard to parse file names using llama.cpp, pulling information from EXIF on images, some building blocks for intro detection on videos. Right now I'm porting the machine learning code to rust so I can bundle it all into a single, small docker container a lot easier, and to learn how the black boxes I'm using actually work, at least on a basic level.

I've tried similar apps but they all either have mediocre UI, are painful to setup, or make me do all the indexing manually. I want something that most people can deploy in a few minutes and that can show off a lot of what it can do immediately, while still letting the user refine the results from automated approaches. The UI has to be amazing to bundle this all into something that is actually useful, but it's been really fun putting it all together, piece by piece.

It's still a bit all over the place, but my NAS has annoyed me enough recently that I'm going to be focusing on getting a version that can be deployed, even if it's missing most of the features I want, so that it can start being useful to me. I think once that happens I'll be more motivated to add features, especially with a solid base I can work from.

Slightly off topic:

I am trying to innovate on moderation systems and I run/code a whitelist moderated forum [0]. You can only see posts and comments from users that you follow. It's a very simple system and there really aren't any gaming vectors. One implication is that if a new user signs up and posts, no one will see it unless they follow. I've actually never used any typical censorship moderation.

[0] https://www.commentcastles.org

Reddit clone:

https://github.com/ferg1e/comment-castles

https://www.commentcastles.org

For moderation, a user whitelist sits in front of the typical content blacklist, which introduces some interesting properties.

I think my forum should work:

https://www.commentcastles.org

I can't think of anything that would be broken except for maybe the logo which is an emoji character.

You may want to check out Comment Castles [0] (disclaimer, I'm the creator). There are 50k comments, but I have never moderated anything. Actually, there are no moderation or admin tools built into the project.

[0] https://www.commentcastles.org

Comment Castles + Reddit thread