• 4 Posts
  • 273 Comments
Joined 1 year ago
cake
Cake day: July 14th, 2023

help-circle

  • I have Fedora Workstation (rpm) as my daily driver and use console ArchLinux on a sdcard for daily backups and rescue etc, so it’s in daily use too.

    Expect:

    1. Much slower bug fixes and updates. Fedora has an organised new kernel testing weekend, not point updates, before they are released to stable. So the first release of a kernel is .3 or something. Arch releases the new kernel fairly quickly, but not as quickly as it used to!
    2. Bug reports are processed much slower. Same reasons.
    3. Reversing out problem packages is much easier in Fedora. Typically three kernels versions are kept. Arch is much more flimsy but it’s getting better over time. It’s still my go-to rescue distro because it’s ‘clean’ - there is not much downstream fiddling on the original source code ( like in Debian etc ) and excessive soft-dependency lists. Fedora does introduce some excess dependencies IMHO.
    4. Fedora is more user friendly. dnf options are well documented and the dash-dash long form options exist. dnf seem to be moving to subcommands too. Pacman really is mainly single letter options which are difficult to remember, and I’m always looking at the arch-wiki Rosetta webpage far more than I should.
    5. I haven’t installed ArchLinux from scratch for a couple of years, but there’s not been a grub2 layer that I can remember. Fedora is still depending on grub long after it was necessary or useful. Things are much slower to change in Fedora. The secure kernel isn’t really as advanced on Fedora compared to ArchLinux.
    6. I’m a rolling release fan because ‘updating’ 2000 packages at once, a la Fedora, seems a recipe for problems. That said, my fears aren’t justified: I’ve only really had one system-upgrade problem with Fedora, and I’ve been using it since FC4 - not always continuously.
    7. I’ve always installed Fedora (and other distros) from the base packages not using an installer. This time I’ve installed SELinux. The miriad complaints I’ve seen about it are not justified. I had an issue with a change to the Fedora SELinux policy tripping a fail; I was impressed and the solution was offered by the setroubleshooter. A user generated problem (ie me) trying to put a systems unit into the /etc/systemd/system directory puzzled me for a while until I realised what I was doing that it didn’t like. Again easily fixed and there are plenty of documents about.

    You could probably summarise those as Fedora is corporate run and ArchLinux is user collective.

    As I said Fedora is my daily driver and I have tried to operate with ArchLinux as my daily, but it was a while ago and too fragile then. ArchLinux is my thin Rescue distro. I wouldn’t be without either.




  • deadcatbounce@reddthat.comtoSelfhosted@lemmy.worldBest phone sync
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    edit-2
    8 days ago

    If you’re using Obsidian for free then maybe try the built-in link which you’ll find in the built-in options I think. It’s a cost option but cheap. I think it eliminates the problems I’m having (below). I’m stubborn.

    I’m not having problem with Syncthing, bar dealing with the stupid attempts to deal with deleted files that Android leaves laying around. I have .stignore files with .trashed-* and .trash/ entries on the Linux machine. Still having problems with _ed directories though and Syncthing conflict files when the sync isn’t fast enough when I switch between the two.

    Sometimes it takes Syncthing a while to work out the best route between the two nodes. Sometimes days. It used to send my packets to the internet before letting them back into the local network. Eventually it found a more direct route between them. I’m not sure but I think it has something to do with local IPv6; I’m talking out of my ass though.

    I’m not affiliated to Syncthing or Obsidian besides being a happy user.

    I have decent battery life on my Pixel 7 Pro. I have the respect battery save setting on so syncing stops at 20% or so I think.









  • I did start trying to prefer the Fedora flatpaks but the flatpak priority system doesn’t seem to work. I have flathub 50 and Fedora 100 but it still chose flathub for the same version.

    Flathub’s packages are usually more up-to-date. Sometimes ridiculously different.

    Then LibreOffice said that they’re releasing on flathub/fedora not packaging so eventually I saw which way things are going and went with the flow.


  • Came to Obsidian from Logseq without prior knowledge. Logseq was my first note-taking of any use. I had Keep but I didn’t really use it, cos my data is in Google’s clutches. I’m trying to degoogle.

    Logseq doesn’t feel together like Obsidian does. I’m in my first month of use and I had to spend ages trying to understand how to take notes properly before I could even feel confidence about using it.

    Templates are a revelation. Dataview supreme. Front-matter a delight. If they exist in logseq they’re hidden. The learning curve for Obsidian is brutal.

    I have around a thousand pages of mixed tags and links because it’s not discouraged in Logseq and I have to undo that huge mess. The discord forums are not a good place to Google for help and there’s nothing like obsidian.rocks or innumerable posts helping. There are lots of logseq YouTube fans and that’s where I ended up looking, but I prefer the written word.

    Logseq was my learning phase but Obsidian is where I’ll stay. I think Logseq needs the level of financial input that Obsidian gets. I’ll have to live with the fact that it is not open source.

    I had disabled the non-open source packages in Fedora software so i didn’t see Obsidian and only saw Logseq. It was a good experience and given my time again I would probably do Logseq then Obsidian, again. I wish I had seen the articles on how to take notes on Obsidian.rockd before I started Logseq.



  • Make it look like a centralised system initially. Provide a portal to a pre vetted/chosen instance that is accepting new members in their locale/country, that is the same for everyone.

    Update: This (above) is badly written. I’m trying to say every potential new member gets presented with the same (pretend centralised) portal that is in fact an (valid long-lived) instance local to the individual potential for them to sign up with. So two local users in Oz get given a proxy to the instance local to them, and a user in Blighty an instance local to that person. The decentralised Lemmy looks centralised, but isn’t. The proxy front end should explain that they’re joining their local instance and it’s like a network of little affiliated clubs that can see each others posts globally. they log in for the first time it will become clear.

    It’s late, I’m tired, sorry everyone. Is that any better?

    I think it’s confusing (the reverse of what they’re used to) for a newbie who have been bought up in a centralised internet with single front ends of all the big players to be presented with little instances to join to access the whole.