gkak.laₛ

Not OP, but at least for me when I tried it:

There was no way to use or even just mount and migrate my existing storage (btrfs+LVM). LVM wasn’t even installed, and when I tried to install it, I got an error saying that apt was disabled on the system, which means I was basically locked out of doing anything more than what they allow you to do on your own hardware.

It seems like it’s technically open source, but having all the vendor lock-in features and lack of control of a proprietary solution

The only use case seems for it to be used as a black box appliance:

• on a new system
• with empty hard drives
• only with ZFS
• without having any control on your own system, except enabling samba etc and maybe installing the predefined Docker containers that they allow you from the web interface

I knew it is supposed to be only an appliance, but with how much people recommended it, I didn’t thing it would be this closed of a system; I think I’ve read about people doing more things with even just their Synology hardware

Do you mean that you’re trying to reproduce it? 🤔 If so:

• hello

• • one
  • two
  • three

• world

is

* hello
* > * one
  > * two
  > * three
* > world

(Btw also, you can check if your client has the option to show the source of a comment 😉)

TL;DR:

Price:

“Under $100”:

After [the preorder], it will go up to $99.

Battery is not rechargeable:

And what happens when the battery runs out? You just send the ring back to be recycled.

Runtime:

The integrated battery will power the device for 12–14 total hours of recording. The designers estimate that to be roughly two years of usage if you record 10 to 20 short voice notes per day.

• “Roughly two years” = lets say that’s 20 months
• 12 hours = 43.200 seconds = 72 seconds/day
• “10-20 short voice notes” = 3.6-7.2 seconds per note

Features:

• Records only while pressing the button

• The recording is converted to text and fed into a large language model (LLM) that runs locally on your device to take actions. The speech-to-text process and LLM operate in the open source Pebble app, and no data from your notes is sent to the Internet. However, there is an optional online backup service for your recordings.

• A model small enough to run on your phone has to focus on specific functionality rather than doing everything like a big cloud-based AI

• • Create or add to notes
  • Set reminder
  • Create alarm
  • Create timer
  • Play/pause/skip music track (via button press)

• also designed to be hacking-friendly. The audio and transcribed text is yours […] You can route it to a different app via a webhook, and the LLM supports model context protocol (MCP), so you can add new functionality that also runs locally. The AI model will also be released as an open source project.

I always set it (mobile client, Thunder), because I find it pretty annoying when I see posts in my feed that I don’t understand (so it’s only fair that I don’t cause it to others)

Fortunately it hasn’t been much of an issue on Lemmy, but Mastodon is pretty much unusable for me partly for this reason (last time I tried to curate my feed, ~50% of the posts I saw were in languages I cannot understand – and I don’t follow language-specific topics or people)

It seems it has now been “solved”, with a popup for users posting from the website, reminding them to select a language: https://github.com/mastodon/mastodon/issues/25568 I think users (including me) will always make mistakes, and, as you note, not all clients support this setting, so I don’t think relying on the UX of everyone’s clients is a permanent solution 😕

In the meantime, the best I can do is set the tag manually when I’m posting 😔

That’s great! (Sorry, couldn’t resist: 😅)

The OSM dataset is huge, and keeps growing every day. Great news, of course, but sometimes the sheer volume can be overwhelming – there are just gobs and gobs of data!

For people who have difficulty reading this (small screens, larger screens, screen readers, etc)

Hi @everyone!

Time for a pretty big update! Behind the scenes, we’ve been quietly cooking up something exciting, and we’re finally ready to share it: the Jellyseerr and Overseerr teams are merging into one team called Seerr! This has been in the works for quite some time, and we couldn’t be happier to officially join forces.

What does that mean for you? A single unified codebase where all the latest Jellyseerr features will make their way in, plus the combined effort means we can move faster on new features and keep things more up to date.

We’re sharing this news a little early because we need beta testers before our first release. If you’d like to help shape the future of this project (and move us towards a quicker first release), now’s your chance!

To test, you can switch from our official image to fallenbagel/jellyseerr:preview-seerr

We do not recommend using this on a production instance, but if you do, please back up your data before switching. For any questions or feedback, please post in our ⁠#seerr-beta channel!

source

I’m not a gemini user, but this seems like a nice tool!

I just came to share this blog post, written by the developer of curl about their opinion on gemini, for anyone interested :D

(PS: thank you for using codeberg instead of centralized alternatives!)

(They changed the name apparently)

https://l.opnxng.com/r/selfhosted/comments/14pdu0m/introducing_crackpipe_your_decentralized/

Embracing “alternatively obtained” games

we want to prevent others from profiting off our hard work by selling our software without our consent.

TIL; for people like me who just found out:

https://gamevau.lt/blog/2023/07/13

For a self-hosted app like GameVault, we believe it’s crucual to disclose the source code. We want you, our users, to have full transparency and control [?] over the software you use on your servers.

our desire to protect our code from unauthorized use and commercial exploitation. While we absolutely encourage you to copy, modify, and share our code for personal use […] we want to prevent others from profiting off our hard work by selling our software without our consent.

As a small business with just two members, we strive to provide you with a valuable product but cannot continue to do so as volunteers indefinitely.

(I’m a AGPL kind of guy, but) btw at least there are licenses specifically for software:

https://www.mongodb.com/licensing/server-side-public-license

https://opensource.stackexchange.com/questions/12070/allowed-uses-of-a-software-licensed-under-cc-by-nc-license

Copyright and the CC-BY-NC license do not regulate mere use, such as executing a program.

Ok proprably we’re at least allowed to run it (That’s not a given, e.g. iirc if someone publishes their code on github without a license, it doesn’t mean that people can fully and legally use it, except for what some Github ToS clause defines that you agreed to)

I was interested in checking it out for personal use; anyone has any experience with alternatives? (I can look them up, I’m just curious about peoples’ recommendations)

For a simple links-only page I’m using Linkstack, and there’s also littlelink

I probably wouldn’t go with a relatively new project that isn’t guaranteed to stick around long-term

Oh of course, I just shared it because I don’t think I’ve seen anything similar and simple, just in case anyone wants to check it out and experiment etc

Oh also a friend of mine is developing this, that uses passwordless magic links or passkeys https://github.com/dzervas/magicentry I haven’t looked much into it though!

You can also check https://kanidm.com/ and https://goauthentik.io/ as well!

Yes, it’s pretty good! I’m a DevOps engineer, and have experience with Ansible, Docker, etc, but I just couldn’t find time to deploy services the best way that I wanted™ for my personal server

So, even though it e.g. doesn’t even use Docker, yunohost really helped me start using the many services I wanted/needed, which otherwise might take e.g. a few hours to a couple of days for each of them to research and configure

So I have one “production” yunohost server, one “testing” yunohost server to test services that I don’t know if I’ll use yet (and I wouldn’t want them to interfere with production e.g. by using too many resources)

and one server without yunohost for mailu, Docker, traefik, etc, which I can use to deploy services the correct way™ as I figure out the services that I really use and find the time to migrate them one-by-one

Even when using yunohost, there are so many things to do after deploying a service (e.g. DNS, configure the server and client software), so it has been really useful to save time when deploying and configuring.

I think it gets you ~80% there, makes self-hosting accessible to everyone, and helps democratize the Internet a bit 💚 It’s more important to have many people setting up e.g. Immich or Nextcloud for their family photos, than only a few Linux people being able to learn how to do it perfectly (Docker/kubernetes high availability, reverse proxies, etc) and have everyone else to need to resort to using centralized services

Element (the matrix chat app) has a “live location” feature (I haven’t had the need to look into anything else)

traccar.org

I’m also pretty sure that I’ve seen more than one relevant Nextcloud apps, but I can’t find one right now 🤔