• 1 Post
  • 290 Comments
Joined 2 years ago
cake
Cake day: July 9th, 2023

help-circle


  • milicent_bystandr@lemm.eetoAnnouncements@lemmy.mlLemmy needs more donations
    link
    fedilink
    English
    arrow-up
    65
    arrow-down
    3
    ·
    8 days ago

    To those who don’t want to give because of the devs’ political views, I’d like to say this goes into improving the Lemmy software our communities here are based on, not making some political podium for the devs in particular.

    There’s a lot of rubbish on Lemmy… because it’s the internet. Like the Great Garbage Patch of the Pacific, it collects; but I still like the Ocean. Some people look down particularly on .ml, but personally I think this is unjustified.

    In fact, before joining Lemmy I heard some rumour of the controversy and went looking. I certainly disagree strongly with certain things I found, but my impression was of Dessalines and Nutomic treating the Internet as a forum for respectful debate and this Lemmy as a technology project not a personal political force - as many see Reddit as becoming. Well, that was my impression, so I was happy to join, and I’ve not been disappointed.

    On this Lemmy software we have many communities, some which vehemently hate each other, but we’re all supported by this foundational software. If it’s too awful to you, you could of course use a different platform, and still federate with the Lemmy communities you love! But we’re still here.

    Because, after all, in any real community there are people you don’t get on with, and part of life is about learning what it means to love each other and live together even so. And that goes for every software project too - somewhere, hiding in the dark recesses, is something quite unpalatable intrinsically linked in. If your ideology says you can’t use anything built with such problems, you’d better start building everything yourself.

    So, you don’t have to support Lemmy. Even if you use it. You don’t have to agree with my take on it. But I’d like to encourage most of you to consider supporting Dessaline and Nutomic to continue this development work, and not just the bare minimum they need - let it overflow with blessing, since we have received such blessing from them!

    …Or you can all send me angry replies for ranting too long on the Internet about something you disagree with ;-)



  • Edit: I just realised your Edit was saying you’ve got legacy bios! So this is all irrelevant. I’ll leave it up in case it helps someone else.


    Could it be selecting the wrong SSD to put the boot loader on?

    When I reinstalled mint the other day on my laptop with an nvme and SSD (also dual boot) it labelled the extra SSD as sda and the original nvme as sdb, so it was going to try to put the bootloader on sda.

    I set up the partitions on the third option (1. Install alongside windows; 2. Wipe everything; 3. Set it up manually) and on the manual setup there’s a selector for the bootloader device just underneath the main section where you select partitions to use for /, /home, etc.

    IIRC you set the bootloader to the full device (in my case sdb) not the EFI partition (sdb1) and it works out out.


  • Thank you!

    Alas I can’t get it to work. After some tweaking and fixing, I’m stuck that the server doesn’t seem to be talking to redis, with this error,

    missing 'error' handler on this Redis client
    microservices worker error: Error: getaddrinfo ENOTFOUND database, stack: Error: getaddrinfo ENOTFOUND database    
         at GetAddrInfoReqWrap.onlookupall [as oncomplete] (node:dns:120:26)     
    microservices worker exited with code 1    
    Killing api process
    

    I’ve been trying to learn bits of podman and docker and how to translate between the two… I think it’s just a bit much for me for now! Thanks anyway, and I’m sure I’ll come back and have another look at your instructions another time.



  • Can you give me some pointers? I’m still new to docker and podman; hoping to get this going without too much learning curve to start with!


    Edit: Thanks to InnerScientist’s instructions below, plus some extra, I got it working. A 5-day-old lemmy comment is probably lost forever to anyone else who wants help, but just in case, here’s some pointers.

    1. I cross-referenced with the official docker compose file, and two other quadlet examples/guides, neither of which worked on their own for me. I used InnerScientist’s files below as my base. Other guides:
    1. Comments need to be on their own line (or deleted), not at the end of the line as below
    2. The environment file from official docker instructions needs some more lines, including duplicating with correct names for the database container (can you put referenced ones in a quadlet like in the docker compose? I don’t know)
    3. The env file also needs DB_HOSTNAME and REDIS_HOSTNAME, which for me were systemd-immich-database and systemd-immich-redis (you can find the names with podman ps, when they’re running, as later in this thread.
    4. Podman, to my surprise, bypasses ufw (uncomplicated firewall) just like Docker!
    • To restrict access to your computer only, you can PublishPort=127.0.0.1:2283:2283 as below.
    • To open to lan, PublishPort=2283:2283
    • If, like me, you’re using a reverse-proxy, but don’t want other LAN access, it’s fairly easy to limit with iptables rules. I use nebula, which puts my vpn within a particular ip range. Script below.

    iptables script

    #!/bin/bash
    iptables -I INPUT -p tcp --dport 2283 -j DROP
    iptables -I INPUT -p tcp --dport 2283 -s 192.168.1.0/24 -j ACCEPT
    iptables -I INPUT -p tcp --dport 2283 -s localhost -j ACCEPT
    

    That puts each line at the top of iptables, before ufw or anything else. So they end up in reverse order from the script.

    • packets to port 2283 from localhost, accept!
    • packets to port 2283 from 192.168.1.*, accept!
    • packets to port 2283 from anywhere [else], drop!








  • I also think atomic distros will become the norm eventually, but I think there’s a long way to go, and not just with user adoption. When I was looking into Nix I was very excited for quite a while, but eventually I realised it’s just another way of handling the package distribution/integration problem. A brilliant one, I agree, but with upsides and downsides like any other answer. And I realised that the incredible work put in by the Debian packagers is a better fit for my needs, no matter that it’s an older approach.

    Perhaps one day, Nix or Nix-like will mature and grow to have the right options to fill my needs better. Perhaps one of the modern Atomics will be good enough for me. Or perhaps Debian et al will run out of steam and good works, or perhaps my needs will change. Or perhaps I’ll die first, after a long and happy life using traditional community package distributions.

    But I look forward to the glorious future of GUIX/HURD. Even if I never live to see it.