tubearchivist VS docker

holy hell I hadn't come across archivebox as of yet, I'll definitely be spinning this up this eve. Is the UI comfortable enough to use as a "bookmark manager"? just been setting up tubearchivist for essentially this purpose, wondering if ArchiveBox would essentially do the same

I'm using TubeArchivist to maintain this VERY large collection.

There appears to be a patch in Tubearchivest at https://github.com/tubearchivist/tubearchivist/issues/587

Personally, I use Tubearchivist.

There are others though.

https://github.com/tubearchivist/tubearchivist

I've been seeing TubeArchivist posted a lot recently.

It has been some time since the last release, but v0.4.0 is finally wrapped up. This brings a wide range of fixes and changes, particularly stability improvements, with our new file system naming convention, this should solve a bunch of previously unsolvable compatibility issues. I highly recommend reading the release notes carefully, as the filesystem migration could be a breaking change if you made changes manually there: https://github.com/tubearchivist/tubearchivist/releases/tag/v0.4.0

I saw this was a Django app so I dug around to look at their models. As far as I can tell this is all they have: https://github.com/tubearchivist/tubearchivist/blob/master/t... - just a `Account` model.

It looks like Django + SQLite is used for user accounts, but all other data storage happens in Elasticsearch.

It's an interesting design decision. I would have gone all-in on the database, and used SQLite FTS in place of Elasticsearch for simplicity, but that's my own personal favourite stack. Not saying their design is bad, just different.

Oh, so it is. I happen to run a TubeArchivist at home that grabs the videos and subtitles. Then I can search through the subtitle files for key phrases. I must have grabbed it while it was still public.

You probably want Tube Archivist. The only problem is that it’s naming convention doesn’t work with the Jellyfin YouTube Metadata Plugin.

Am I better off using this container: https://hub.docker.com/_/nextcloud/

It looks like I'm not the only person who has faced this. apache-based images require buster, for instance, and some docker images that rely on Ruby face issues too (for example, I decided to try setting up Postal but it looks like it's facing the same issues).

If you look at the info on https://hub.docker.com/_/nextcloud, you'll see that you aught to be specifying at least one volume, so that you have the data you want to not disappear on restart, and can access config files.

Yes, and this is what L passed through as the env variables in the docker compose file, as the docs state: https://hub.docker.com/_/nextcloud

You can't map binaries from host to container like this. What you need is a custom Dockerfile that installs ffmpeg. https://github.com/nextcloud/docker/blob/master/.examples/dockerfiles/full/fpm/Dockerfile

A good example to start with is at https://github.com/nextcloud/docker/tree/master/.examples/docker-compose/insecure/postgres/apache

Nextcloud is free and open source. The easiest way to install it is via docker containers. nextcloud docker