Ah I see, I guess at least that would help with the main UI, but I’m already using collabora through the collabora code server in next cloud so it sounds like I’ll probably have the same document editing experience with OCIS/opencloud. I used to use onlyoffice but after I tried out their mobile app, it started blocking me from editing documents using the next cloud app (which seemed to use the only office web UI) so I was forced to switch unless I started paying for onlyoffice.

What are the apps that you would miss? I basically only use my NC as a Google drive and docs replacement, so all it has to do is store docx files and let me edit them on desktop or mobile without being glitchy and I’ve really wanted to consider OCIS or similar.

That second requirement for me seems hard because of how complex office suites are, but NC is driving me to my wit’s end with how slow and error prone it is, and how glitchy the NC office UI is (like glitches when selecting text or randomly scrolling you to the beginning).

Isn’t the iMX 8M in the Liberty phone still made overseas? In which case it won’t be exempt from the tariffs.

Looks really interesting, I tried to download it off of hugging face though and my access request was denied, so I guess it’s not really downloadable today for regular people it seems.

After I check the usual torrent sites, if I can’t find it on Usenet then I try finding random sketchy streaming sites (just by duckduckgoing title + year + stream) because usually you can youtube-dl them with a little fiddling, then my last resort is a DHT indexer like btdig. That’s more risk with viruses but if you’re paying attention it’s not that hard to avoid, but DHT indexes have lots of ancient stuff so be prepared to wait for seeds if you find it.

Sometimes I also find what I’m looking for on like page 7 of duckduckgo results at archive.org

In case you haven’t realized, the user and pass in the docker compose are for setting the user/pass that you will enter on windows to access the share. It doesn’t have to be the same as the Linux server user account - though mine is the same because it’s easier to remember.

This very community seems to be the largest by far - around 45k followers if I’m seeing it correctly. It feels a lot more active than the average Lemmy community as well.

Hmm, well it doesn’t seem to be any problem with the docker compose then as best as I can tell. I picked a random ext4 flash drive and replicated your setup with the UID and GID set and it seems to work fine:

# /etc/fstab
/dev/sda1       /home/<me>/mount/ext_hdd_01  ext4    defaults 0 2

~/mount % ls -an
total 12
drwxr-xr-x  3 1000 1000 4096 Mar 27 16:22 .
drwx------ 86 1000 1000 4096 Mar 27 16:31 ..
drwxrwxrwx  3    0    0 4096 Mar 27 16:26 ext_hdd_01

~/mount/ext_hdd_01 % ls -an
total 6521728
drwxrwxrwx 3    0    0       4096 Mar 27 16:26 .
drwxr-xr-x 3 1000 1000       4096 Mar 27 16:22 ..
-rw-r--r-- 1 1000 1000 6678214224 May  5  2024 PXL_20240504_233345242.mp4
drwxrwxrwx 2    0    0      16384 May  5  2024 lost+found
-rwxr--r-- 1 1000 1000          5 Mar 27 16:27 test.txt

# ~/samba/docker-compose.yml
services:
  samba:
    image: dockurr/samba
    container_name: samba
    environment:
      NAME: "Data"
      USER: "user"
      PASS: "pass"
      UID: "1000"
      GID: "1000"
    ports:
      - 445:445
    volumes:
      - /home/<me>/mount:/storage
    restart: always

I was able to play the PXL.mp4 video from my desktop and write back the test.txt file

Have you checked the logs with docker logs -f samba to see if there’s anything there?

Also you could try to access the HD from within the container, using docker exec -it samba bash and then cd into /storage and see what happens.

I would suggest adding “UID” and “GID” environment variables to the container, and set them to the numeric values for user and group numbers that show in place of your name when you use “ls -an” inside of the “mount” folder (they will probably be the same number).

For example, if inside your mount folder you see:

ls -an
total 12
drwx------ 2 1001 1001 4096 Mar 27 13:54 .
drwxr-xr-x 3 1000 1000 4096 Mar 27 13:51 ..
-rwx------ 1 1001 1001    0 Mar 27 13:54 hello.txt
-rwx------ 1 1001 1001    4 Mar 27 13:54 test.txt

Then set UID: 1001 and GID: 1001

I get the same error as you when I copy your docker-compose and try to access a folder owned by my user. When I add the UID and GID of my user id to the docker-compose (1001 for me), the error goes away.

What did you set UID and GID to and what is the output of “ls -an” when run inside of the shared directory? You can remove the file names for privacy. I just tested the docker container and it seems to work between my Linux laptop and my windows 11 desktop using this docker compose:

services:
  samba:
    image: dockurr/samba
    container_name: samba
    environment:
      NAME: "Data"
      USER: "samba"
      PASS: "secret"
      UID: "1000"
      GID: "1000"
    ports:
      - 445:445
    volumes:
      - ./samba:/storage
    restart: always

The files in my shared folder are owned by UID/GID 1000/1000 which is why I put those as my UID/GID, and when I logged in from Windows I entered samba and secret as the password and I was able to access and modify the files in the shared folder.

Have you done the steps under “How do I modify the default credentials?” and “How do I modify the permissions?” from the readme?

The bigscreen beyond used 1 inch displays so this could be useful for getting much higher resolutions in a smaller form factor. I don’t think my Index displays are even as big as 4 inches so I guess that explains why all the super high res headsets are so huge.

Edit: my index screens are under 3", just did a lens swap so I have the old eye tubes handy

I think this is the tool you are looking for: https://github.com/luigi311/JellyPlex-Watched

I haven’t used it myself but I plan on it eventually when I do make the switch.

Gotcha, I’ve never actually considered the bandwidth limits. It looks like digitalocean includes 1TB per month and I used 242GB last month. If I ever get close to the limit I will just spin up another droplet. I don’t think I would even need to load balance unless the first one is struggling since the bandwidth allowance across all droplets is pooled together.

If you aren’t already using a reverse proxy, then do you currently just port forward or use the Plex relay? The only reason I use one is because of CGNAT. Before I moved to a place with only CGNAT I port forwarded for both Plex and Jellyfin.

We’re seeing this at work too - our public git frontend is constantly getting scraped as well as our self hosted issue tracker. We had to spend days working on fail2ban and other kinds of tools to mitigate all the traffic that’s adding tons of load to our instances, which otherwise would easily be able to handle the handful of employees who actually use these systems.

In case this helps as a reference point, I use a $5 digital ocean droplet as my Plex and Jellyfin reverse proxy and it seems to handle the traffic of 3-5 simultaneous streams just fine. I use Haproxy in tcp mode (so no http interpreting, just passing packets) in an attempt to keep the CPU load minimal and just make it a pure I/O task.

Then we’ll have to disagree about that - imo this is the perfect place to discuss Plex alternatives and what features are keeping us on Plex. I think this discussion needs to happen if we want to learn how to create viable alternatives.

I especially want to talk about this because I personally want nothing more than to switch myself and everyone who I share my library with onto Jellyfin, and I don’t think that will happen unless we talk about what’s missing. I’m personally invested in Jellyfin enough to donate to apps I don’t even use in hopes that they will improve.

Yes? Is that odd to you? If jellyfin supported it then that would be one less reason against switching which would be a good thing, wouldn’t you think? If you advocate for using jellyfin then shouldn’t you want such basic features to be supported for those who want to use them?

Even though I still use Plex full time, I very much want Jellyfin to succeed (I run it and offer it to everyone I share with), and so I want Jellyfin to be usable for people of all skill levels. I can’t get my parents to use an app that requires them to know anything about file sizes or codec compatibility or converting anything. That is why Plex is as successful as they are.

If you’re satisfied with Jellyfin lacking certain features, that’s your perogative. But I don’t think it’s that hard to empathize with someone wanting more feature parity, especially if the motivation is to make Jellyfin accessible to more people and increase adoption.

It’s honestly kind of silly to suggest that only technically minded users care about file sizes. We’re lucky enough to even know why the file is so big. My regular friends will just complain that it won’t fit, blame jellyfin, and then go back to Netflix.

You know that regular people with 64GB phones exist right? Suggesting that a non technical person should just know that they need to convert a 30GB remux using ffmpeg is absurd.

I give all my friends the choice between Plex and jellyfin (I run both containers side by side pointed to the same media folders) and they all invariably choose Plex. I think it has a lot to do with the jellyfin UI, and I think an overhaul like jellyfin-vue or something that looks like findroid needs to happen in order for jellyfin to really appeal to regular people.