Unknown parent

lemmy - Link to source

InnerScientist

From the mailing list I'm reading that kernel maintainers have heard a few companies looking for something like this, so yes?

Edit:

However, to be clear, the Hornet LSM proposed here seems very reasonable to me and I would have no conceptual objections to merging it upstream. Based on off-list discussions I believe there is a lot of demand for something like this, and I believe many people will be happy to have BPF signature verification in-tree.
This entry was edited (2 months ago)
in reply to ikidd

They probably named it HORNET for a reason - think Japanese Murder Hornets... What Could Possibly Go Wrong??

It will probably start out as little glitches and slowdowns to destroy faith in your system ("Windows works right all the time") a random 2 second pauses. Finally one day every Linux box in the world crashes, all at the same time, because some 'dummy' in Microsoft deleted the private signing key.

This entry was edited (2 months ago)

Latest image of u-blue has removed firefox, several gnome support packages and some pipewire and gstreamer packages


Removed: firefox-136.0.1-1.fc41.x86_64 firefox-langpacks-136.0.1-1.fc41.x86_64 freerdp-libs-2:3.12.0-1.fc41.x86_64 gnome-bluetooth-1:47.1-1.fc41.x86_64 gnome-browser-connector-42.1-8.fc41.x86_64 gnome-classic-session-47.4-2.fc41.noarch gnome-initial-setup-47.2-1.fc41.x86_64 gnome-remote-desktop-47.3-1.fc41.x86_64 gnome-shell-extension-apps-menu-47.4-2.fc41.noarch gnome-shell-extension-background-logo-47.0-1.fc41.noarch gnome-shell-extension-common-47.4-2.fc41.noarch gnome-shell-extension-launch-new-instance-47.4-2.fc41.noarch gnome-shell-extension-places-menu-47.4-2.fc41.noarch gnome-shell-extension-window-list-47.4-2.fc41.noarch gstreamer1-plugin-libav-1.24.11-1.fc41.x86_64 gstreamer1-plugins-bad-free-1.24.11-1.fc41.x86_64 pipewire-utils-1.2.7-7.fc41.x86_64 uresourced-0.5.4-2.fc41.x86_64

The latest build of u-blue:main has removed these packages. Is anybody also experiencing this? I could find no information regarding this on the discourse forum.

Edit : the issue has been resolved in the next build image. It was a small error that crowd in due to shifting to dnf5. Thanks for to maintainers for doing all the hardwark

This entry was edited (2 months ago)
in reply to Sentau

Didn't know what uBlue was, so here: universal-blue.org/

"The Universal Blue project builds a diverse set of continuously delivered operating system images using bootc. That's nerdspeak for the ultimate Linux client: the reliability of a Chromebook, but with the flexibility and power of a traditional Linux desktop.

These images represent what's possible when a community focuses on sharing best practices via automation and collaboration. One common language between dev and ops, and it's finally come to the desktop.

We also provide tools for users to build their own image using our templates and processes, which can be used to ship custom configurations to all of your machines, or finally make the Linux distribution you've long wished for, but never had the tools to create.

At long last, we've ascended."

in reply to Leaflet

Nice post, ~~but your title is misleading~~: the blog post is actually titled "Supply Chain Attacks on Linux distributions - Overview" - the word "attacks" as used here is a synonym for "vulnerabilities". It is not completely clear from their title if this is going to be a post about vulnerabilities being discovered, or about them actually being exploited maliciously, but the latter is at least not strongly implied.

~~This lemmy post however is titled (currently, hopefully OP will retitle it after this comment) "Supply Chain Attack found in Fedora's Pagure and openSUSE's Open Build Service".~~ edit: @OP thanks for changing the title!

Adding the word "found" (and making "Attack" singular) changes the meaning: this title strongly implies that a malicious party has actually been detected performing a supply chain attack for real - which is not what this post is saying at all. (It does actually discuss some previous real-world attacks first, but it is not about finding those; the new findings in this post are vulnerabilities which were never attacked for real.)

~~I recommend using the original post title (minus its "Overview" suffix) or keeping your more verbose title but changing the word "Attack" to "Vulnerabilities" to make it clearer.~~

TLDR: These security researchers went looking for supply chain vulnerabilities, and found several bugs in two different systems. After responsibly disclosing them, they did these (very nice and accessible, btw - i recommend reading them) writeups about two of the bugs. The two they wrote up are similar in that they both involve going from being able to inject command line arguments, to being able to write to a file, to being able to execute arbitrary code (in a context which would allow attackers to perform supply chain attacks on any software distributed via the targeted infrastructure).

This entry was edited (2 months ago)

On Zorin OS, can't get XClicker custom location to work [Solved]


Sorry if this isn't the place for this, I couldn't find a forum for the program. I have 2 monitors and when I try to use a custom location it seems like XClicker doesn't understand how to handle that. When getting a location the coordinates behave as I'd expect, treating all of my screen space as one with the numbers increasing as you go right and down, the top left of my left monitor currently being 0,142. However when I activate to start clicking it seem as though it treats the top left corner of whichever monitor the cursor is on as though it were 0,0, meaning it doesn't actually click where I want.

For example, if I Get a location on the right monitor and position the mouse on the right monitor it'll actually click a location 1920 pixels to the right, since the left is 1080p, and I think 142 pixels down. With the same coordinates, if I position my mouse on the left monitor then activate It'll go close to where I want it to on the right monitor for the first click, then shift to the right again since it's now on the right monitor.

Edit: Oh I figured it out and it's definitely newbie stuff. I was poking around and for shits tried to upgrade my OS. While I couldn't do that, it seems like it did upgrade a lot of other stuff and now the clicker works normally. I was hoping it'd be a miracle and fix another issue, too, but it didn't.

This entry was edited (2 months ago)
in reply to mina86

Note: for readers who aren't aware, the notation ^X means hold down the ctrl key and type x (without shift).

ctrl-a though ctrl-z will send ASCII characters 1 through 26, which are called control characters (because they're for controling things, and also because you can type them by holding down the control key).

^D is the EOF character.
$ stty -a | grep eof
intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>;
$ man stty |grep -A1 eof |head -n2
       eof CHAR
              CHAR will send an end of file (terminate the input)


Nope, Chuck Testa: there is no EOF character. Or, one could also say there is an EOF character, but which character it is can be configured on a per-tty basis, and by default it is configured to be ^D - which (since "D" is the fourth letter of the alphabet) is ASCII character 4, which (as you can see in man ascii) is called EOT or "end of transmission".

What that stty output means is that ^D is the character specified to trigger eof. That means this character is intercepted (by the kernel's tty driver) and, instead of sending the character to the process reading standard input, the tty "will send an end of file (terminate the input)".

By default eof is ^D (EOT), a control character, but it can be set to any character.

For instance: run stty eof x and now, in that terminal, "x" (by itself, without the control key) will be the EOF character and will behave exactly as ^D did before. (The rest of this comment assumes you are still in a normal default terminal where you have not done that.)

But "send an end of file" does not mean sending EOT or any other character to the reading process: as the blog post explains, it actually (counterintuitively) means flushing the buffer - meaning, causing the read syscall to return with whatever is in the buffer currently.

It is confusing that this functionality is called eof, and the stty man page description of it is even more so, given that it (really!) does actually flush the contents of the buffer to read - even if the line buffer is not empty, in which case it is not actually indicating end-of-file!

You can confirm this is happening by running cat and typing a few characters and then hitting ^D, and then typing more, and hitting ^D again. (Each time you flush the buffer, cat will immediately echo the latest characters that had been buffered, even though you have not hit enter yet.)

Or, you can pipe cat into pv and see that ^D also causes pv to receive the buffer contents prior to hitting enter.

I guess unix calls this eof because this function is most often used to flush an empty buffer, which is how you "send an end of file" to the reader.

The empty-read-means-EOF semantics are documented, among other places, in the man page for the read() syscall (man read):

RETURN VALUE
      On success, the number of bytes read is returned (zero indicates end of
      file), and the file position is advanced by this number.


If you want to send an actual ^D (EOT) character through to the process reading standard input, you can escape it using the confusingly-named lnext function, which by default is bound to the ^V control character (aka SYN, "synchronous idle", ASCII character 22):

$ man stty|grep lnext -A1
       * lnext CHAR
              CHAR will enter the next character quoted
$ stty -a|grep lnext
werase = ^W; lnext = ^V; discard = ^O; min = 1; time = 0;

Try it: you can type echo " and then ctrl-V and ctrl-D and then "|xxd (and then enter) and you will see that this is sending ascii character 4.

You can also send it with echo -e '\x04'. Note that the EOT character does not terminate bash:

$ echo -e '\x04\necho see?'|xxd
00000000: 040a 6563 686f 2073 6565 3f0a            ..echo see?.
$ echo -e '\x04\necho see?'|bash
bash: line 1: $'\004': command not found
see?

As you can see, it instead interprets it as a command.

::: spoiler (Control characters are perfectly cromulent filenames btw...)

$ echo -e '#!/bin/bash\necho lmao' > ~/.local/bin/$(echo -en '\x04')
$ chmod +x ~/.local/bin/$(echo -en '\x04')
$ echo -e '\x04\necho see?'|bash
lmao
see?

:::
This entry was edited (2 months ago)
in reply to Arthur Besse

Which is why I haven’t wrote ‘EOF character’, ‘EOT’ or ‘EOT character’. Neither have I claimed that \x4 character is interpreted by the shell as end of file.

Edit: Actually, I did say ‘EOF character’ originally (though I still haven’t claimed that it sends EOF character to the program). I’ve updated the comment to clear things up more.

This entry was edited (2 months ago)

Here's an exercise in extreme masochism:


The Linux Ship of Theseus


  1. pick any distro and install it.
  2. Then, without installing another distro over the top of it, slowly convert it into another distro by replacing package managers, installed packages, and configurations.

System must be usable and fully native to the new distro (all old packages replaced with new ones).

No flatpaks, avoid snaps where physically possible, native packages only.

EDIT: Some clarification on some of the clever tools brought up here:

chroot, dd, debootstrap, and partition editors that allow you to install the new system in an empty container or blanket-overwrite the old system go against the spirit of this challenge.

These are very useful and valid tools under a normal context and I strongly recommend learning them.

You can use them if you prefer, but The ship of Theseus was replaced one board at a time. We are trying to avoid dropping a new ship in the harbor and tugging the old one out.

It may however be a good idea to use them to test out the target system in a safe environment as you perform the migration back in the real root, so you have a reference to go by.


Easy: pick two similar distros, such as Ubuntu and Debian or Manjaro and Arch and go from the base to the derivative.

Medium: Same as easy but go from the derivative to the base.

Hard: Pick two disparate distros like Debian and Artix and go from one to the other.

Nightmare: Make a self-compiled distro your target.

This entry was edited (2 months ago)

Switching to linux for newbies.


Honestly, at this point I’m so done with window’s bullshit. Their operating system is damn near unusable. All the convenient stuff is hidden in weird places. The constant whining about having to buy their crap. Or worse trying to get me to use the horrible software that comes with the new versions.

My excuse used to be, but I can’t play games on it. This is no longer the case for the games I love. So Windows can suck it. At this point I’m switching away from a lot of stuff I used to use. (x-box became Steam-deck, twitter became blue-sky and reddit is becoming Lemmy) As a kind of computer illiterate person, this has been a worth while transition but a difficult one. Let’s just say I had to learn a lot of new stuff.

So I’m a total Linux newbie but thanks to my Steam deck I’ve become somewhat used to using it. Not like an expert, but I have run wine to create separate environments for running pokémon fangames. And have taken a look around the Linux environment. I like it and think I’ll be able to get used to it with practice. It reminds me weirdly of windows XP in how easily I can get everything to work the way I want. It takes a bit of doing and some research, but it works. Which is all I want in an operating system.

I am looking for tips as to where to start searching, because I am converting my windows computer to Linux. I just don’t know what version.

Any user experience is welcome, I have no idea where to begin. I mostly use the computer I’m installing this on as a glorified typewriter, that I play movies, music and retro-games on.

A user friendly version is preferred, I find it hard to parse out from the various versions I have seen so far how easy they actually are to use. Extra points if a large amount of the information has easy to find tutorials on the internet. I don’t always know where to start looking and as I learned while getting wine to work, some of the names/terms are completely different. (And kind of a lot at once if you are just getting started).

Any resources you might think are useful for a newbie are also highly appreciated.

tl;dr: I (a Linux noob) am looking for a recommendation for what version of Linux to use for my needs. And any tips tricks or other info that I might need to know before I switch. Because windows sucks.

I’m sorry if this has already been asked and answered. I did try to find an answer through searching, but as I already mentioned. My lack of terms and knowledge is holding me back.

Email client for Linux


I have been looking for an email client on Linux after being tired of Gmail and Outlook web clients.

I had Thunderbird installed on my system and thought I'd give it a spin. I set up POP for my email accounts and it worked fantastic... For a total of 2 hours, after which I realised that searching in Thunderbird is simply not going to work for me. I need to search by attachment name and sometimes even by text inside attachment and unfortunately Thunderbird can't do that (I think I tried an extension too but it made the UI super clunky to the point that I couldn't even understand how to navigate it anymore).

Does Betterbird or any other email client fix this problem? I'm willing to try other options if they are FOSS.

Thanks

in reply to marauding_gibberish142

Yes but I can’t search by the name of the attachment.


I just searched for text thats in an attachment filename and it worked - with a caveat. I have a filename called "PMASUP236 - Operate Vehicles In The Field.pdf" on an email. There is no reference to the PMASUP236 in any other part of any email.

If I search "PMASUP236", it returns the email as a result.
If I search "SUP236" it does not.
If I search "Operate Vehicles" it returns that email (along with a heap of others containing the word "Operate" and "Vehicles" in any order).

Admittedly this is on Windows at work, though I do run Thunderbird on Linux at home. Will have to try it there to confirm.

in reply to marauding_gibberish142

I just checked on Linux (Thunderbird 128.5.2esr, Opensuse Tumbleweed) and the behaviour is the same.

If I search “PMASUP236”, it returns the email as a result.
If I search “SUP236” it does not.


This is using the normal search function (top of screen in current version). Quick Filter does not look at attachments at all by the looks. The "Attachments" toggle is only a has / does not have attachment filter.

This entry was edited (2 months ago)
in reply to dontblink

sure. first, configure sudo to be passwordless, or perhaps just to stay unlocked for longer (it's easy to find instructions for how to do that).

then, put this in your ~/.bashrc:

alias sudo='echo -n "are you sure? "; for i in $(seq 5); do echo -n "$((6 - $i)) "; sleep 1; done && echo && /usr/bin/sudo '

Now "sudo" will give you a 5 second countdown (during which you can hit ctrl-c if you change your mind) before running whatever command you ask it to.

This entry was edited (2 months ago)

Need help with setting up full disk encryption (FDE) where /home is on another drive.


Any distro would do, but for my case, it is Arch because I have more control over the partitions. I would like the OS, so root, swap and others on 1 drive. The /home should be on a separate drive. The tricky thing is to have everything encrypted, except /boot and /efi of course.

Now, here is what I can do

1) FDE on 1 drive. This is easy: you create /efi, /boot and then create a large LUKS partition. From there, you create LVM on that LUKS partition and get your: /, /home and swap. Then mount everything correctly and install.

In the grub config, you only need to set it so it knows the LUKS partitom and where the root is. For eg, if your LUKs partition is /dev/sda3, you do:

  • cryptdevice=UUID=: cryptlvm rootfs=/dev/vg/root.

2) Unencrypted /home on another drive. This is like 1) but /home is mounted on a separate drive. Still need to do the grub config, but nothing is needed for /home. It is automatically mounted when you login.

Now for my case: Encrypt /home

The encryption and mount part is easy. But how to get the OS to recognize it? The Arch wiki has this weird thing where you create an encryption key, they called it home.key, using cryptsetup. You then store the key in /etc and then in your /etc/crypttab, you specifiy the drive with /home and location of the key. No need for any passphrase.

The problem I have with this is that keys are stored in root. So if my root system is corrupted, I cant even decrypt home....

Any advice is welcome..

This entry was edited (2 months ago)
in reply to mazzilius_marsti

Well thanks everyone. I finally managed to get it to work on Arch. System has separate encrypted root and swap in LVM, and a separate encrypted home. It can suspend and hibernate. Below are my steps

DISK PREP

  • partition the main drive for your swap and root first. For me, it is a boot partition + an EFI + a LUKS container with LVM on top. Create your volumes. I use Arch, so format and mount them appropriately before pacstrap. Leave out mount point for /home.
  • Go to your other drive, follow: cyberciti.biz/hardware/cryptse…

to create a LUKS container that is encrypted with: a keyfile and a password. Test both to make sure you can open the locked drive. Format and mount it at /mnt/home or where you want the /home to be.

  • Pacstrap and then genfstab.
    Important: Make sure to copy the keyfile from your archiso environment to your chroot environment aka your system. Otherwise, when reboot, the keyfile is gone. I put it in /root and set permission so only root can read.

AUTOMATIC UNLOCK

  • First, fstab. When you do genfstab, things should be fine. But just double check the UUID is correct for /home. Note in fstab, the UUID is the unlocked one: so the one with /dev/mapper/home. Change to noatime if you desire.
  • Second, crypttab. Assume you decrypt your LUKS home as "home". Add this:

home uuid of the unencrypted home drive location of the keyfile luks

The link above said to just use /dev/sda, but imo UUID is safer if you have a removable drive.

  • Third, grub. Edit your /etc/default/grub and append the following to GRUB_CMD_LINUX:

"rd.luks.uuid=UUID of the locked luks home drive"

FOR HIBERNATION

For some reasons, hibernation doesnt work out of the box. It works when I have everything in 1 drive, i.e 1 boot, 1 efi, 1 lvm on luks for /home, swap and /. The fix is simple:

  • add "resume" to /etc/mkinitcpio.conf. Add before "filesystems" . Rebuild your initramfs with mkinitcpio -P.
  • add to /etc/default/grub: "resume= uuid of the unlocked swap partition". Or if you do LVM, just use "resume=/dev/vg/swap".

Special thanks to bodaciousFern@lemmy.dbzer0.com and Lemmchen@feddit.org for giving me correct ideas about "rd.luks.uuid" and that LUKS can do both pass and keyfile.

This entry was edited (2 months ago)

My (satisfying) experience moving to Linux!


Just wanted to share my experience moving to Linux from MacOS. Very satisfying, but of course not at first. I think my patience has improved a lot too lol.

I started out trying live bootables on my 2012 MBA. 4GB RAM, 60GB HDD. Not a beast really, but it is my only computer. I obviously couldn't risk ending up without a working OS, so the only option was dual boot from an external drive. Bought an SSD connected via USB and started trying to install distros. Initially Fedora Workstation. Was a mess. Slow, wifi was not working well, odd crashes etc.. Decided to start over with something lighter, but all other installers crashed halfway through. I kid you not I shot my back again bent over my small laptop i
without working peripherals trying to install different distros. My doctor was not happy when I came back and told her I fucked up my back again because of my posture lol. Apparently, a shitty USB leads to crashes on most installers. I knew Anaconda worked tho, so I went back to a lighter DE with Fedora, XFCE. Set up an install on the SSD with a shared partition I could access from both MacOS and fedora. No big permission issues yet.

Then fixing network drivers. There is a lot of info about what chip needs what driver, a lot of which is incorrect apparently, because my chip which was supposed to work with bcma needed broadcom-wl. The joy when I remembered USB tethering was a thing.. For a laptop with no ethernet plug this was a godsend. Got the drivers, got wifi.

And since then, many "issues" I encountered where simply things that generally happened behind the scenes on MacOS I didn't even know where happening. Learning about these things has been very gratifying, and gives a lot of respect for a polished OS that just works like magic. Eventually, an issue on MacOS I could not solve due to it being a walled garden made me switch to Linux as a daily driver, and once I got over CMD and CTRL being swapped it sped up my workflows and runs better overall. More tweaking tho of course.

There are odd quirks but I found fun solutions for some, and began planning and learning to remedy others. Mostly, everything is working really well. I am having a lot of fun!

My tip for anyone struggling with getting started with linux, set up a log function so you can easily log any relevant changes you make, and have it accessible from somewhere else (like a shared partition or external drive for example). This way you know what you have done and can use that to fix whatever you fuck up. Also, make a knowledge base with the sources you find useful. I have a small kb in UpNote now so I can look up how some things were done instead of having to search and find the right guides over and over.

This entry was edited (2 months ago)
in reply to wewbull

Nice I didn't know that ^^ should probably learn at least the basic bash operators, I am just hacking together the different commands I happen to know at the moment really

Edit: why echo instead of printf?

This was causing a lot of issues with newlines, like when I fetched the log to view it my $ was right after the log entry so I switched it back. But it is probably useful in the future to use >> instead :)

This entry was edited (2 months ago)

TIL you can send commands to terminal (in vscodium)


I've been using the terminal for 5 years.

99% of attempts at keeping valuable code snippets failed. Or having a useful command history.

Finally, I looked up how to send a line to terminal.

I use vscode since Atom was discontinued (looking forward to zed). But in the meanwhile, I just added

{
    "key": "ctrl+enter",
    "command": "workbench.action.terminal.runSelectedText",
    "when": "terminalProcessSupported && editorTextFocus && editorLangId =~ /shell|bash|zsh|fish/"
  }

Maybe it helps someone :)

With this, I can finally keep a history of all useful commands I run. Like I can have a systemd shell file where I store the commands that I ran. Or a file to store all my podman commands. Now I do not have to type them anymore, I can just run them.

This entry was edited (2 months ago)

Torvalds Frustrated Over "Disgusting" Testing "Turd" DRM Code Landing In Linux 6.15


The big set of open-source graphics driver updates for Linux 6.15 have been merged but Linux creator Linus Torvalds isn't particularly happy with the pull request. In particular, he's unhappy with some new "hdrtest" testing code being built as part of full kernel builds and the "turds" it leaves behind and this code "needs to die" at least from the perspective of non-DRM driver developers.

Mint Cinnamon Issues (Solved)


So following on from my last post I'v now tried again, this time using Rufus to make my boot USB rather than balena etcher which despite being reccomended is apparently problematic. I also downloaded and verified a fresh install of the cinnamon ISO.

Result is uh....

Failed to open \EFI\BOOT\mmx64.efi - Not Found
Failed to load image (windows symbol): Not Found
Failed to start MokManager: Not Found
Something has gone seriously wrong: import_mok_state() failed: Not Found

This happens when trying to boot from the USB so its actually worse than last time where the demo environment worked but it kept crashing on attempting a proper setup.

PC in question is a small notebook with no discreet GPU to speak of (Intel integrated graphics) that I've been hoping to use as a testbed before comiting to Linux on my main PC.___

This entry was edited (2 months ago)
in reply to CheeseNoodle

Sorry to link to reddit, but have you already tried the suggestions in this thread from a year ago: old.reddit.com/r/pop_os/commen… ?

Large (1920x9750, ~3MB) screenshot for posterity + those who absolutely do not want to access reddit at all: files.catbox.moe/mqsdxh.png

Edit: (Related links)
- linuxquestions.org/questions/l…
- forums.linuxmint.com/viewtopic…
- unix.stackexchange.com/questio…

I'd try the "copy \EFI\BOOT\grubx64.efi to EFI\BOOT\mmx64.efi" solution personally and see what happens. If that works, you might just have an obnoxious BIOS on that computer.

This entry was edited (2 months ago)

The fediverse has a bullying problem


So check it out: Mastodon decided to implement follower-only posts for their users. All good. They did it in a way where they were still broadcasting those posts (described as "private") in a format that other servers could easily wind up erroneously showing them to random people. That's not ideal.

Probably the clearest explanation of the root of the problem is this:

Something you may not know about Mastodon's privacy settings is that they are recommendations, not demands. This means that it is up to each individual server whether or not it chooses to enforce them. For example, you may mark your post with unlisted, which indicates that servers shouldn't display the post on their global timelines, but servers which don't implement the unlisted privacy setting still can (and do).

Servers don't necessarily disregard Mastodon's privacy settings for malicious reasons. Mastodon's privacy settings aren't a part of the original OStatus protocol, and servers which don't run a recent version of the Mastodon software simply aren't configured to recognize them. This means that unlisted, private, or even direct posts may end up in places you didn't expect on one of these servers—like in the public timeline, or a user's reblogs.


That is super relevant for "private" posts by Mastodon. They fall into the same category as how you've been voting on Lemmy posts and comments: This stuff seems private, because it's being hidden in your UI, but it's actually being broadcasted out to random untrusted servers behind the scenes, and some server software is going to expose it. It's simply going to happen. You need to be aware of that. Even if it's not shown in your UI, it is available.

Anyway, Pixelfed had a bug in its handling of those types of posts, which meant that in some circumstances it would show them to everyone. Somebody wrote on her blog about how her partner has been posting sensitive information as "private," and Pixelfed was exposing it, and how it's a massive problem. For some reason, Dansup (Pixelfed author) taking it seriously and fixing the problem and pushing out a new version within a few days only made this person more upset, because in her (IMO incorrect) opinion, the way Dansup had done it was wrong.

I think the blog-writer is just mistaken about some of the technical issues involved. It sounds like she's planning on telling her partner that it's still okay to be posting her private stuff on Mastodon, marked "private," now that Pixelfed and only Pixelfed has fixed the issue. I think that's a huge mistake for reasons that should be obvious. It sounds like she's very upset that Dansup made it explicit that he was fixing this issue, thinking that even exposing it in commit comments (which as we know get way more readership than blog posts) would mean people knew about it, and the less people that knew about it, the safer her partner's information would be since she is continuing to do this apparently. You will not be surprised to discover that I think that type of thinking is also a mistake.

That's not even what I want to talk about, though. I have done security-related work professionally before, so maybe I look at this stuff from a different perspective than this lady does. What I want to talk about is this type of comments on Lemmy, when this situation got posted here under the title "Pixelfed leaks private posts from other Fediverse instances":

Non-malicious servers aren’t supposed to do what Pixelfed did.

Pixelfed got caught with its pants down

rtfm and do NOT give a rest to bad behaving software

dansup remains either incompetent for implementing badly something easy or toxic for federating ignoring what the federation requires

i completely blame pixelfed here: it breaks trust in transit and that’s unacceptable because it makes the system untrustworthy

periodic reminder to not touch dansup software and to move away from pixelfed and loops

dansup is not competent and quite problematic and it’s not even over

developers with less funding (even 0) contributed way more to fedi, they’re just less vocal

dansup is all bark no bite, stop falling for it

dansup showed quite some incompetence in handling security, delivering features, communicating clearly and honestly and treating properly third party devs


I sort of started out in the ensuing conversation just explaining the issues involved, because they are subtle, but there are people who are still sending me messages a day later insisting that Dansup is a big piece of shit and he broke the internet on purpose. They're also consistently upset, among other reasons, that he's getting paid because people like the stuff he made and gave away, and chose to back his Kickstarter. Very upset. I keep hearing about it.

This is not the first time, or even the first time with Dansup. From time to time, I see this with some kind of person on the Fediverse who's doing something. Usually someone who's giving away their time to do something for everyone else. Then there's some giant outcry that they are "problematic" or awful on purpose in some way. With Dansup at least, every time I've looked at it, it's mostly been trumped-up nonsense. The worst it ever is, in actuality, is "he got mad and posted an angry status HOW DARE HE." Usually it is based more or less on nothing.

Dansup isn't just a person making free software, who sometimes posts angry unreasonable statuses or gets embroiled in drama for some reason because he is human and has human emotions. He's the worst. He is toxic and unhinged. He is keeping his Loops code secret and breaking his promises. He makes money. He broke privacy for everyone (no don't tell me any details about the protocol or why he didn't he broke it for everyone) (and don't tell me he fixed it in a few days and pushed out a new version that just makes it worse because he put it in the notes and it'll be hard for people to upgrade anyway so it doesn't count)

And so on.

Some particular moderator isn't just a person who sometimes makes poor moderation decisions and then doubles down on them. No, he is:

a racist and a zionist and will do whatever he can to delete pro-Palestinian posts, or posts that criticize Israel.

a vile, racist, zionist piece of shit, and anyone who defends or supports him is sitting at the table with him and accepts those labels for themselves.


And so on. The exact same pattern happened with a different lemmy.world mod who was extensively harassed for months for various made-up bullshit, all the way up until the time where he (related or not) decided to stop modding altogether.

It's weird. Why are people so vindictive and personal, and why do they double down so enthusiastically about taking it to this personal place where this person involved is being bad on purpose and needs to be attacked for being horrible, instead of just being a normal person with a variety of normal human failings as we all have? Why are people so un-amenable to someone trying to say "actually it's not that simple", to the point that a day later my inbox is still getting peppered with insistences that Dansup is the worst on this private-posts issue, and I'm completely wrong and incompetent for thinking otherwise and all the references I've been digging up and sending to try to illustrate the point are just more proof that I'm horrible?

Guys: Chill out.

I would just recommend, if you are one of these people that likes to double down on all this stuff and get all amped-up about how some particular fediverse person is "problematic" or "toxic" or various other vague insinuations, or you feel the need to bring up all kinds of past drama any time anything at all happens with the person, that you not.

I am probably guilty of this sometimes. I definitely like to give people hell sometimes, if in my opinion they are doing something that's causing a problem. But the extent to which the fediverse seems to like to do this stuff just seems really extreme to me, and a lot of times what it's based on is just weird petty bullying nonsense.

Just take it it with a grain of salt, too, if you see it, is also what I'm saying. Whether it comes from me or whoever. A lot of times, the issue doesn't look like such a huge deal once you strip away the histrionics and the assumption that everyone's being malicious on purpose. Doubly so if the emotion and the innuendo is running way ahead of what the actual facts are.

This entry was edited (2 months ago)
in reply to PhilipTheBucket

IMO, Dan has some responsibility but more of it lies with Mastodon and other microblogging software that labels this post type as "private", "followers only" or similar without any further explanation. It needs to be clear that it's dependent on good faith and competence of remote servers that may collect that information.

Moreover we need to do a better job of letting users know that anything posted on the internet, and especially anything posted to the fediverse where it's backed up on potentially thousands of servers, should be assumed to be publicly-visible and eternal. If nothing else, it will be backed up on the internet archive. If you want to communicate privately, this is the wrong place.

I wish there was a private social media platform but it seems like the closest we're going to get is Signal.

Also "the bullying problem" has nothing to do with the Fediverse and everything to do with people in general and the erosion of nuance.

This entry was edited (2 months ago)
in reply to PhilipTheBucket

Re: The fediverse has a bullying problem


Let's clarify something here. Mastodon follower only posts don't have the "public pseudo user" addressed, do they?

That's the important piece that this whole thing hinges upon

If it is present, Mastodon is a fault. If not, Pixelfed messed up.

Am I mistaken?

For example, a followers-only post is addressed to the followers collection of the author, at minimum. When NodeBB sees this, it doesn't even consider it public, because the public pseudo user isn't addressed. We also have no concept of "followers only", but we handle them just fine.

This entry was edited (2 months ago)

Is it rude to reply using English under posts in a language you can’t speak?


I often reply under Japanese posts, and I always assume users will use a translator as I do, but maybe in the context of a Japanese instance or conversation this may look rude?
This entry was edited (2 months ago)
in reply to Jumuta

nah, it’s better for information integrity to reply in the language you understand imo, comments translated using translator services are very obvious anyway and some people are multilingual


Sure, I agree? Maybe there's a misunderstanding here and I should add that it simply would never even occur to me to enter a conversation if I didn't natively understand the language that's being used.

This entry was edited (2 months ago)

Interstellar (mbin app) added piefed support!


Interstellar (!interstellar@kbin.earth) has added Piefed support!

How do i try?


Install interstellar from the play store/F-droid, and create an account on preferred.social/ (only instance with the API enabled, get join code from invite). Then, login from interstellar and you're done!

What's next?


Well, andrew is working on porting thunder to piefed (the main dev of it is interested in it) Admiral patrick said he might port tesseract to piefed, once the API is stable, and ae_harding said he may add it after lemmy 1.0.

Thank you @andrew_s@piefed.social, @olorin99@kbin.earth and @jwr1@kbin.earth for working tirelessly to implement this :D

This entry was edited (2 months ago)

Kann man das Mastodon Konto verknüpfen?


Bin wirklich noch etwas überfordert hier. Wie kann man das feddit Konto und das Mastodon konto verknüpfen?

Wo finde ich die Adresse von meinem feddit Konto und muss ich es auf Mastodon einbinden? Hilfe.... 😱 Bitte...

This entry was edited (2 months ago)

I propose we storm Toronto. Search every building until we find Nicole. The fediverse chick!


Maybe then she'll SHUT UP! Oh my god, is there a single person on this platform who doesn't get these messages every month???

Guys, you all KNOW what I'm talking about, right? Gonna have to go get Tim Misney on this case! And you KNOW what he DOES! eyebrows

in reply to Lost_My_Mind

One huge motive behind creating !nicole@feddit.org was to concentrate all the posts about that spam, which, in turn, are also spam, to one place, which can easily be blocked by users that - understandably - don't want to see that spam. Especially when they don't get the original nicole spam in the first place, such as me - which is highly ironic, but an interesting phenomenon. I'm pretty active as a commentator, and not inactive as a poster, yet only got exactly one message a few months ago.
This entry was edited (2 months ago)

advanced38 doesn't like this.

Isn’t GPLv3 the ultimate solution to all broligarcy problems?


I always believed that problems around technocratic broligarchy could’ve been solved if the software ecosystem using GPLv3 was strong. So when I suddenly see a flurry of articles stating that the EU is trying to come up with its own spin of Linux to boycott monopolies by US companies, I think it is either a waste of time or self defeating because the ultimate power of computer codes isn’t the code itself, but the license attached to it.

Is there something that I have missed in this matter because I barely see anyone talking about the need for more libre software. I find the latest transition from GLP to MIT license for rust-based coreutils to be concerning because we’re basically adopting “doormat” licenses that will enable companies to easily embrace, extend and extinguish the freedoms we cherish so much in this space.

in reply to Maroon

We as a society have to enlighten ourselves such that we have a standard and this standard is open source.

The more people refuse to use proprietary stuff, and promote open source tools, the more powerful those tools get.

As long as we pay 100 bucks for a shitty game and refuse to donate 1 buck to a foss calendar we are far away from where we'd like to be.

If people would pay good for foss, companies would do it

This entry was edited (2 months ago)
in reply to AbnormalHumanBeing

I've started doing notes in the terminal as well. I used Obsidian and Logseq for a while, mainly because I wanted something with a GUI so that I could recommend it to people who aren't comfortable with the terminal. But eventually I figured that a terminal solution was the right one for me, since I have a terminal open all the time anyway.

I switched from vim to kakoune a while ago ( kakoune.org ), so I use nb ( xwmx.github.io/nb ) instead of vimwiki.

nb is a terminal application that will open whatever your default text editor is.

This entry was edited (2 months ago)
in reply to AbnormalHumanBeing

Not a fan of vim, it's improper implementation of the ex command set and the way it ads line feeds, when you cut-n-paste between windows, makes it basically useless. I much prefer the BSD derived nvi, even on Linux. Like VIM it also handles multiple byte character sets, but UNLIKE VIM it is a COMPLETE and CORRECT implementation of vi/ex not a half-assed kind of sort of implementation.
in reply to CheeseNoodle

If you want to try to get Linux on the main drive working (since USB works but isn't ideal) there are a few things I encountered.

-some distros just didn't like my hardware. Failed to install, or installed but boot would get errors and halt. The remedy was using an rpm distro rather than deb based (I tried about 10 debs, the rpm ones acknowledged the bios error and moved on)

-secure boot can be a bit of a pain. If you don't want to deal with it, Turnoff secureboot, and in some cases EFI and use legacy BIOS mode.

-if you want Secure boot and EFI. Allow USB boot in BIOS, do the install and ensure it is building a GPT disk with an EFI partition. At the reboot stage it should ask if you want to enroll keys, say yes.

If during reboot it does nothing or boots to windows(assuming you have windows drive). Go into BIOS and choose secure boot option where you can pick which Secure OS it found and move that to top of boot list.

-if it is not those things it is often nvidia on Wayland or X issue on laptops. If you don't want to mess with installing a GPU switcher, you can often set your laptop to discrete graphics before install and bypass the two GPU issues

This entry was edited (2 months ago)
in reply to CheeseNoodle

As others say, it can be done. If you want more normal umpf, you'll need to mount parts of the filesystem to your ssd. You can mount /home or / on ssd, or have an overlay file system as a file on an ssd/hdd, or use bcachefs with back propagation to the usb, or similar fancy setups.

So you'll boot linux kernel from the usb, but most disk activity will be on your ssd. Fun project, but not super easy/practical if it isn't done automatically.

My old HP microserver is 'made' to boot from a usb-stick inserted on the mb.

Anyway, perhaps an AI can suggest a script to do what you want ?

This entry was edited (2 months ago)

What do you think might be some fun, positive ways for instances to distinguish themselves?


Instances of any fediverse software, to be clear.

Can we please make a viable (federated!) amazon alternative? I have an idea!


Hi folks!
I'm here with another idea. Let's make an amazon alternative. I know! I know! That was asked for a couple times already but lets discuss some details.

Amazon is basically glorified dropshipping by now. What if we just made federated (not sure if over activitypub would work) ads and sales, powered by fediseer (the "trust" network of the fediverse).

Example 1:
So you buy at toms groceries, you trust them. they have experience with tina's hardware store and they trust them. so you can buy both toms and tinas wares on both sites.

Example 2:
So for example, I run a small business that sells computers. You run a small business that sells mice and keyboards. I have worked with you before so I mark you as trusted in my local website, which federates with yours, showing your products in my shop. If a customer buys my computer and buys your keyboard on top, my site sends you a buy order with customer address and payment. I get a small fee for my electricity of say 1%.

Can someone try and poke holes in this idea? It feels like this could work!

Have a nice weekend.

in reply to haui

A version of this focussed on a gift economy/trading platform (e.g. like freecycle, or the buy nothing groups on facebook) would also be cool.

Also person-to-person buying/selling, rather than business-to-person would be nice to have, like craigslist, reverb.com, gumtree, or used items on ebay.

If this was focused on a craigslist/gumtree style of selling, where most of the actual trade is done off-site/in-person with cash or bank transfers, it would completely side-step the payment processor problem.

This entry was edited (2 months ago)

Bridges: The Last Network Effect


This talk discusses how, through bridging, Mastodon and Bluesky can create a cross-protocol network for the open social web.
This entry was edited (2 months ago)

Federated wiki software?


I absolutely love wiki walking through random obscure fan wikis, but I hate how most are on Fandom.

I think a federated wiki solution makes sense. I could see it as an evolution of the interwiki concept.

in reply to lgsp@feddit.it

Also:

  • gift economy/trading platform (e.g. like freecycle)
  • buying/selling (e.g. like ebay)
  • local community/bioregionalism networks (e.g. what nextdoor should be)

These seem kind of ideal for a federated network, IMO.

I actually think Lemmy would be a pretty decent format for something stackoverflow like - just maybe needs to UI tweaks to minimise the visual space that replies take up, plus maybe answered post flair

in reply to ProfHillbilly

I think you're starting too big, to use your phrase.

Get used to the basics first: what's different (to whatever you're running now), what's the same. They Linux distributions almost all have GUIs (KDE, GNOME are the main ones but there are many others).

Run a live USB version from a usb stick to get used to it until you have the confidence to install it on an old pc. Personally I do not recommend dual booting; data gets lost that way. Install it on an old pc and learn how to restore your backups to a Linux filesystem (not the fs of what you're used to on Linux platform). I write that because you said that want to end up with a Linux server.

Choose one of the top few from distrowatch.com/

Your aim is to understand what's going on under neither the GUI; how permissions work.


I started by installing a VoIP product into a VM on Windows 2000, but there are better ways now.

Good luck. You shouldn't find it that difficult.

This entry was edited (2 months ago)

Configuring A PeerTube Instance


This entry was edited (2 months ago)

What is going on behind the scenes when searching for a lemmy post in a search engine?


I was looking for some collection of posts earlier about Proton Mail and the whole controversy with the CEO, and I opened a post the lemmy instance that was suggested was lemmy.zip but the community and the poster were from lemmy.world so that made me ask myself a bunch of questions. Reference link

Note: I used duckduckgo

Here are some questions I have:

  • How does the search engine decide which instance to link you to as you could in theory show every instance for the same post?
  • Could you get a result where all the results are the same post just different instances?
  • Do you think that could deter new people finding out about lemmy through search results?
  • How can an instance make themselves more visible in the search results (for exposure)?
  • I did not get any results from lemmy clients such as vger.app the only results were direct instances, will this always be the case?

I remember learning about search engines a while back but I don't know how relevant that information is any more. Having crawlers and the more a website is linked in other websites the higher up in the search result will be and the whole robot.txt thing.

I know if I wanted to search for something specific in lemmy I could just use its own search function, but what about people who ask general questions and that happens to be answered in a lemmy post. I wanted to know how exposed we are/ will be to people who don't yet know about lemmy.

Meltdown doesn't like this.

in reply to JustAnotherKay

So domains are recursive tree structures. So u have node com which has a bunch of nodes below it of different domains which can have sub domains etc. Its like this cos that's how DNS was designed.

Search engines gives scores to domains and pages. When u say that the content origin of ur page is some other domain the search engines will use that knowledge to adjust rating accordingly.

Say u have 2 domains site1.example.com and site2.example.com the search engine will have a rank for site1, site2, and example.com where the rating of both site1 and site2 effect the rating for example.com.

If the content origin for all federated content has the same origin say instance.fediverse.com then the rating for all federated content will be classified as part of fediverse.com and all content will be working together across all instances boosting the fediverse as a single entity.

This entry was edited (2 months ago)

PieFed supports auto merging threads so you see all the comments in one post


cross-posted from: piefed.social/post/538685

No more duplicate posts

One of the things that the recent addition of the Feeds feature highlighted was how many cross-posts / duplicate posts there are. When you display posts from linux@lemmy.world, linux@programming.dev, linux@lemmy.ml, etc all the cross-posts make it get repetitive, really fast. The same thing happens on the home feed too although it's a bit less obvious because there's a wider range of subjects involved.

Except now, it doesn't, because PieFed de-duplicates your feed! And your home page, and your topics. Attached to this post is a screenshot showing how it works out - an article posted to 7 different places is only shown once despite me having joined most of those communities.

We're still figuring out whether it's a good idea to merge all the comments from all the cross-posts into one page and how to do that in a way that respects the different culture/rules in the communities that the posts were made in. It's a tricky UX and social question.

I've held off on adding a cross-post function to PieFed until now but it'll be added soon.

I believe that the "Nicole" images being sent to Threadiverse users may be intending to deanonymize accounts


For those not familiar, there are numerous messages containing images being repeatedly spammed to many Threadiverse users talking about a Polish girl named "Nicole". This has been ongoing for some time now.

Lemmy permits external inline image references to be embedded in messages. This means that if a unique image URL or set of image URLs are sent to each user, it's possible to log the IP addresses that fetch these images; by analyzing the log, one can determine the IP address that a user has.

In some earlier discussion, someone had claimed that local lemmy instances cache these on their local pict-rs instance and rewrite messages to reference the local image.

It does appear that there is a closed issue on the lemmy issue tracker referencing such a deanonymization attack:

github.com/LemmyNet/lemmy/issu…

I had not looked into these earlier, but it looks like such rewriting and caching intending to avoid this attack is not occurring, at least on my home instance. I hadn't looked until the most-recent message, but the image embedded here is indeed remote:

lemmy.doesnotexist.club/pictrs…

I haven't stored and looked through a list of these, but as I recall, the user sending them is bouncing around different instances. They certainly are not using the same hostname for their lemmy instance as the pict-rs instance; this message was sent from nicole92 on lemmy.latinlok.com, though the image is hosted on lemmy.doesnotexist.club. I don't know whether they are moving around where the pict-rs instance is located from message to message. If not, it might be possible to block the pict-rs instance in your browser. That will only be a temporary fix, since I see no reason that they couldn't also be moving the hostname on the pict-rs instance.

Another mitigation would be to route one's client software or browser through a VPN.

I don't know if there are admins working on addressing the issue; I'd assume so, but I wanted to at least mention that there might be privacy implications to other users.

In any event, regardless of whether the "Nicole" spammer is aiming to deanonymize users, as things stand, it does appear that someone could do so.

My own take is that the best fix here on the lemmy-and-other-Threadiverse-software-side would be to disable inline images in messages. Someone who wants to reference an image can always link to an external image in a messages, and permit a user to click through. But if remote inline image references can be used, there's no great way to prevent a user's IP address from being exposed.

If anyone has other suggestions to mitigate this (maybe a Greasemonkey snippet to require a click to load inline images as a patch for the lemmy Web UI?), I'm all ears.

This entry was edited (2 months ago)

Help with systemd (I think)


SOLVED If any lost souls find themselves here in a similar situation let it be known that the file that worked for me was creating a file at '%h/.config/systemd/user/google-drive-ocamlfuse.service' with the following content:

[Unit]
Description=FUSE filesystem over Google Drive
After=network.target

[Service]
ExecStart=google-drive-ocamlfuse %h/googledrive
ExecStop=fusemount -u %h/googledrive
Restart=always
RestartSec=300
Type=forking

[Install]
WantedBy=default.target

Howdy, I have very recently installed Opensuse Tumbleweed alongside Windows 10 (On a seperate drive) and am trying to get things setup to where I can fully transition to linux. One of the first things I have hit a wall on is getting a file to execute on boot using systemd.

I am trying to use this package to be able to access my google drive from Dolphin. And so far it works okay. Except that it doesn't survive a reboot. I have to run the command:

google-drive-ocamlfuse ~/googledrive

after each reboot in order for the google drive directories to be accessible. So I googled how to make that happen automagically on boot and found this guide that helped me get a startup script going.

I created /usr/local/bin/ocamlfuseStartup.sh as a file that contains the command from before:

google-drive-ocamlfuse ~/googledrive

and verified that it works as intended when I enter ./ocamlfuseStartup.sh from that directory.

I then created another file at /usr/lib/systemd/system/startup.service that contains the following:

[Unit]
Description=Startup Script

[Service]
ExecStart=/bin/bash /usr/local/bin/ocamlfuseStartup.sh

[Install]
WantedBy=multi-user.target

I have no idea what the /bin/bash portion is for because I found it from a googling but without it I get the following error:

startup.service: Main process exited, code=exited, status=203/EXEC

However with it I get this error:

startup.service: Main process exited, code=exited, status=2/INVALIDARGUMENT

which I take to mean that there is something wrong with my ocamlfuseStartup.sh file maybe? But since it works when I manually execute the file I'm kind of at a loss.

I found this thread where it seemed like someone else was having a similar issue but I didn't really grok what they were talking about.

Any and all help is greatly appreciated!

This entry was edited (2 months ago)
in reply to Marafon

The 203 error you got is because your script isnt a valid executable, it needs to have a shebang at the top, you can change it to something like this and set the executable bit with chmod +x <file>

\#!/usr/bin/env bash
google-drive-ocamlfuse ~/googledrive

this tells it to run using bash as the interpreter.

Im not familliar with this google drive software, but im figuring that its exiting with an error code cuz its running as a system service, and $HOME probobly isnt set so ~ doesnt expand and the software gets an invalid path.

But I recommend using a user service for this, it will run when you login, you should be able to copy the service file you already have into ~/.config/systemd/user/ and run systemctl --user daemon-reload and systemctl --user enable startup.service --now, this will enable and start the service in one go.

I also recommend adding the following lines under [Service]

Type=simple
Restart=always
RestartSec=60

idk if the software will exit if it loses network or wifi or anything, but this will have it automatically restart after 60 seconds, should it exit for any reason.

If you need it to run before login, it is possible to do with a system service, but it will need a bit more setup

This entry was edited (2 months ago)
in reply to Marafon

i would install cron if its not installed, then add a file under

/etc/cron.d/someNiceToRememberName

and with an editor write the following in it:

@reboot root /bin/bash -x /usr/local/bin/ocamlfuseStartup.sh >> /tmp/output 2>&1 ; echo returncode=$? >> /tmp/output

where root is the user it should run with and the redirection (>> /tmp/output 2>&1) is for seeing what is wrong, and echo returncode= prints the exit error indicator.

you might be able to do this with systemd too, but it often messes things up and i like i.e. clean log outputs.

for testing you can add a line that would start it without a reboot.

22 * * * * root /path/to/your/command parameters >> /tmp/output 2>&1

where the 22 is the next minute it should run. with this cronjob it would be run at every x:22 after each hour 09:22 pm 10:22 pm aso. i do this sometimes for testing

as systemd starts cron with a broken environment too, it might as well not work using cron, but then you know its the broken environment systemd starts it with.
if thats your problem, you might want to add:
PATH=$PATH:/usr/sbin:/usr/local/sbin

somewhere at the top of the shellscript depending on what is missing, which by itself would be the directories the commands executed are in.
in terminal use :

which sed

for example to find where the "sed" binary is located (if the script cannot find it)
you could change every occurance of "sed" in the script with the full path to it, or add the directory using the PATH variable as a colon separated list of directories to search commands in (where order matters)

to maybe more easy see what is missing i added the -x as a parameter to bash which makes it print out a line for every command it executes to sdterr so you can see what the last commands were.
you can try it in the terminal (if your terminal is bash or compatible

$ set -x
$ echo foo
+ echo foo
foo
$ 

i'ld guess systemd starts the shellscript with a broken environment and thats maybe why it doesnt work, maybe try adding this to the script too:
echo PATH: $PATH

somewhere at the top of the script and use the redirection i showed for the cron job.
run it once in your terminal and compare the output with what it prints during automated script run.

good luck.

transition to linux. One of the first things I have hit a wall on is getting a file to execute on boot using systemd.

I am trying to use this package to be


that bash thing is the interpreter. the script ocamlFuseStartup.sh is written as bash (possibly the same as your terminal) if that script has a "shebang" (#!/bin/bash) as the very first line AND the file has the execute permission set (chmod +x /path/to/file), then you dont need /bin/bash at the beginning to start it.

This entry was edited (2 months ago)

Raspberry Pi Announces rpi-image-gen To Help Craft Custom Software Images


This entry was edited (2 months ago)

Following the article about old school forums dying, a reminder that there is !fedibridge@lemmy.dbzer0.com that is dedicated to promote the Fediverse. Pinned posts have resources everyone can use.


in reply to solrize

The fediverse is more like the evolution of those old school forums, but diversified into more than just forums. Decentralized, user ran shit that could be a forum, a microblog, a YouTube alternative, a Discord/IRC alternative, an image host, etc. And then they could possibly share content with each other directly, but that's kinda up to the individual app/site.
This entry was edited (2 months ago)

Pixelfed user count has gone vertical.


Editing to let people know that I will be blocking anyone who feels the need to tell me why this graph is inaccurate. I truly don't care, but feel free to chime in with your useless take and land a spot on my block list! 🙂
This entry was edited (4 months ago)

Madbrad200 doesn't like this.

Unknown parent

Yeah I was on Fosstodon first. Moved to Sakurajima (Sharkey) and now I'm on Sakurajima (Mastodon).

You're right, the instance makes a big difference. Very important to find one you enjoy.

This entry was edited (2 months ago)

Subscribing to mastodon


Given the different way they apps are structured I don't even know if it's a sensible question, but is there a way to subscribe to something on a Madison server?

The server as a community, or a tag as a community? Does mastodon have some other concept that I could subscribe to?

Mastodon.online invitation if anyone wants it


For anyone who wants an invitation to join mastodon.online, let me know! Like mastodon.social, it is one of the two official communities run by the Mastodon non-profit. The difference is that mastodon.online is invite-only.

Joining an invite-only server helps keep out spammers who target mastodon.social and lets someone be part of a more exclusive community. This can help someone have a better-quality mastodon experience, overall.

Of course, to keep out scammers, the invitation isn't open to just anyone who asks, they will have to have a posting history of some sort.

in reply to Sibshops

Quick question for you. I signed up to Mastadon a while back, have been endeavouring to use it more regularly, and have been trying to sign on to as many local or often-subscribed orgs (on other media platforms). What I'm finding is that remarkably few of them are on Mastadon, and often those that are have been using them as a kind of alt-Twitter content splash with little engagement.

Are the actual active profiles/communities sort of walling themselves off from bots and bad actors in the way that you're describing? Invite-only groups whose activity isn't seen by non-members, or even accessible when searching for related tags/topics? Have I been casting about looking for content like someone looking for people in a neighbourhood where everyone is inside with the curtains drawn, or is there actually low engagement on Mastadon overall?

in reply to Komodo Rodeo

If you want to follow specific organizations, unfortunately, there aren’t as many on Mastodon as on other social media platforms. Some are accessible through Threads.net federation, like @theonion@threads.net, through the Bluesky bridge, or via other non-Mastodon federated communities (e.g., Flipboard.com) if that’s something you’re interested in.

However, some communities do wall themselves off from bots and bad actors. For example, the most popular Mastodon instance (mastodon.social) blocks the second-largest instance (pawoo.net). Other instances, like mastodon.art, defederate from AI art supporting instances, while others like hachyderm.io has defederated from Threads.net. Additionally, some Mastodon communities, like Truth Social, are entirely isolated and defederate from everyone else.

So, if you want access to the widest range of communities, you may need to join more than one instance.

in reply to Sheridan

The servers are hosted by Hetzner with servers across Europe. They call out Hetzner when there is an outage.

status.mastodon.online/cm722jt…

We have switched back to our original networking, and all functionality and performance is restored. Hetzner has confirmed that is was indeed a vSwitch issue that caused the network outage.


More info about Hetzner here:
hetzner.com/unternehmen/ueber-…

This entry was edited (2 months ago)

Substack open source rival Ghost is now connected to the fediverse


in reply to jackalope

The docs are super unclear.

In future, we hope to develop a deeper integration between your social web profile and followers <> and your public website and registered members. To start with, though, they operate independently from one another.

Your social web profile and followers are separate to the rest of your site and memberships, so you can think of it as a new, additional distribution channel.


Emphasis mine. I think the setup is basically federating links in Mastodon style. So if you have a post available as a subscription, it will prompt a login after clicking through.

This entry was edited (2 months ago)

Organizing files to admit different grouping depending on use cases


I have separate directories for uni courses, which are grouped in semester directory. I also TA some classes, which is stored in separate folder named 'TA'. That is, it is grouped like this:

University
| - ...
| - 2024.2
| | - Lie algebra
| | - Operator algebra
| - 2025.1
| | - Mathematical Algorithms
| | - Diophantine equations
| - TA
| | - ...
| | - 2024.2
| | - 2025.1

Oftentimes, I focus on the current semester, so I want to view courses on a same semester grouped together.
On other times, I want to group TA activities across semesters together to .
I may also do the same with grouping similar subjects.

Basically, I want to view directories with different grouping for each use case, as in the title. I hope this makes sense..
Is there any kind of directory structure or application-based solutions for this cases?

EDIT: I want both GUI and TUI solution for browsing files like this, it's great if linux filesystem supports this natively but fine if it doesn't.
Database with redirection capability would be even better.

Thanks in advance!

This entry was edited (2 months ago)

Why aren't lemmy DMs compatible with mastodon?


Mastodon dms are just posts with limited visibility, and a mention to the person recieving it.
They don't seem to be compatible with lemmy.
in reply to irelephant [he/him]🍭

They're not, they're a different kind of object on the ActivityPub level. Lemmy and some others use the unofficial extension called Litepub which includes a custom ChatMessage type, while Mastodon and others have chosen to repurpose the Note object (used for posts and comments on Mastodon and comments on Lemmy) with limited visibility and weird mention rules.
in reply to Eiri

So many AP platforms are made by a couple of guys in their garage, it's not even funny, and the mentality of "just dicking around" means what gets used is whatever the whims of the day dictate, rather than the standard.

"If it works, it's not stupid" and all that.

But that kind of work lacks real world testing, and depe concern for public expectaton or desire.

Plus, you have to keep in mind that the idea of interplatform interoperability isn't this core conceit of ActivityPub. It's a potential use case, but it's not an expectation. There's no reason anyone should expect interop like that, other than some developers wanted to try it.

But some didn't, and now that their platforms are gaining audience, they're refactoring to meet that audience's expectations.

in reply to Eiri

I recently did a library implementing pretty much the whole ActivityPub protocol and let me tell you one thing I've learned: ActivityPub is a half-assed unfinished mess.

It has so little constraints that pretty much anything is valid. You'd think that there was some logic at the core, but not really.

For example, there's the Like activity, the name is pretty self-explanatory. But it's valid to like anything. You want to like another Like activity? Why not! So you can like a like that likes a like...

Obviously, no real project implements such nonsense. But because almost everything is valid and up to interpretation of the implementer, there really is no single ActivityPub to implement that gets you running smoothly. There are many de-facto standards with various levels of compatibility with each other. Sending private messages is one such thing that many projects chose to implement in different ways.

Fun fact: Without implementing the HTTP Signatures spec, your app will not be able to send any activities to any software, because ActivityPub allows anyone to impersonate anyone, which obviously didn't really fly with people developing actual software that uses it.

In conclusion, pure ActivityPub is an unusable mess in its pure form that literally won't be able to communicate with other software implementing "ActivityPub," so everyone is kind of implementing a similar slice/hybrid of ActivityPub and other protocols, but not exactly the same ActivityPub.

in reply to Rikudou_Sage

They have conferences about ActivityPub. Why isn't W3C trying to fix this mess in newer backwards-incompatible versions? The time to do it is now, not later, because it would involve a major version and years of pushing for adoption. The lack of standardization of basic concepts is why integrations of different types of implementations is a broken mess, which is the whole fucking point of ActivityPub! Now, we have to compete with ATProto, which has different kinds of problems, and it's very possible that it just wins out and kills ActivityPub.

This reminds of the early days of SMTP, where there was zero thought behind security, and that created an entire spam industry.

PeerTube integration into Lemmy is still shit, poorly implemented, and rarely linked by Lemmy admins.

Creating A GoToSocial Instance on FediHost


Does something like this exist?


I'm asking myself if exist something inside the fediverse where you can list your profiles and other people can Send you a message throw your most used app/profile without needing an sccount on that service, unify conversations throw different platforms and notify you if one person you follow in A service has a new profile in B service.

Something like a federated meeting point compatible with all ActivityPub developments. Does it exist?

Thanks!

in reply to javilopezg

Send you a message [through] your most used app/profile without needing an [account] on that service


No, I don't think so...

However, what you're describing kind of sounds like a polycentric identity. Maybe check that out. Also, I believe ActivityPods aim to handle a little of what you're describing, but I don't know how much adoption it's had.

PostgreSQL Database Lands Initial Support For IO_uring: "Can Be Considerably Faster"


KDE Dolphin can use ripgrep and ripgrep-all for faster searching, also inside PDFs!


After Trump's decree: fight for US funding for Tor, F-Droid and Let's Encrypt


in reply to Lee Duna

That website's cookie requirements are gross, here's the whole article

The Open Technology Fund (OTF) has filed a lawsuit in the US District Court in Washington D.C. against the US Agency for Global Media (USAGM) and the Office of Management and Budget. In its lawsuit, the OTF is seeking a preliminary injunction to have the USAGM release the withheld funding. US President Donald Trump had previously issued a decree largely restricting the USAGM under the current legal situation. The OTF uses its funds to support the certification authority Let's Encrypt and the anonymization network Tor, among others.

OTF lacks 650,000 US dollars for operating costs

In its application, the OTF argues that the termination of the grant by the USAGM is unlawful, as the provision of the funding has already been decided by Congress. As part of this decision, a total amount of 43.5 million US dollars has been earmarked for 2025, which accounts for 98 percent of the OTF's funding. The USAGM oversees the financial and programmatic activities of the OTF and makes payments to the non-profit organization. The OTF had requested and not received a payment of around 650,000 US dollars for operating costs in March.

Kari Lake, executive CEO of USAGM and special advisor to the Trump administration, described the US agency in a statement as a "huge rot and burden on the American taxpayer" that also poses a national security risk. OTF Chairman Zack Cooper, on the other hand, argued that his organization is the most efficient and effective tool against censorship and influence peddling. An end to OTF projects "would weaken America's national security and keep millions of people around the world trapped behind authoritarian information firewalls", Cooper said.

Overall, the US government invests a lot of money in open source software. Last year, Let's Encrypt received around 800,000 US dollars in funding from the OTF, the Tor network received almost 500,000 US dollars and the open-source Android app store F-Droid received 396,000 US dollars. In total, the organization currently supports around 50 projects, including the development of the free VPN client OpenVPN. According to its information, the OTF has published around 2,500 patches for open-source software and the organization promotes VPNs for around 45 million people in countries with censorship. OTF President Cunningham sees the lawsuit as the only way to ensure the continued existence of these projects.

This entry was edited (2 months ago)

How do you not feel overwhelmed using Mastodon?


I feel like everyone suggests following hashtags, but depending on the hashtag, I find the content that's being posted quite overwhelming when it comes to the amount of toots, and that it's hard to get an overview. Anyone that relates?
This entry was edited (2 months ago)
in reply to Yingwu

I felt the same way every time I tried to use Twitter as I feel every time I try to use Mastodon. It's either way too much or way too little. I prefer everything about the reddit/lemmy/threadiverse style.

How would we even be having this conversation on microblogging? A bunch of reposts, with or without comments, disconnected from each other... So much nicer to have a "subject" line and a page where every relevant comment is presented.

redshift doesn't like this.

in reply to Yingwu

I think the main idea is to look at some hashtags to find people to follow, then eventually wean off those hashtags if you want.

Another key detail is that you can’t read it all. Not hashtags, not people. You’ll go nuts if you try. It’s about following people who are interesting, opening the app every once in a while to check in, then going on with your day.

in reply to beeng

With tools it is often the thing with "newer, easier" and "old reliable" versions. If you want to write and print a text, Word might be easier, but TeX is still more predictable and reliable. And with software like the literally CORE utils, I actually prefer predictable and reliable over fancy eye candy. I could easily see to have them both installed, so everything that worked before still works without the risk of breaking at some odd point, but might use the Rust version like "Rcp -g" if applicable.

Importing Facebook posts into Mastodon or Friendica?


Ok, this might be a dumb question, but is it really technically possible in any way - even a convoluted one, like manually downloading HTML pages from Facebook? - to mass import posts from one's Facebook into Mastodon or Friendica instead of painstakingly copying them one by one? I've got a 10+ year archive to port and it's a daunting task to go through them...

HiDPI XFCE - use DPI, not scaling and life is good. Better than good.


As a grumpy old man who wishes his computer would stop changing I've been trying to get on board with XFCE for a while and the big blocker has been making things work well on my 4K screen. (For the record this post is based on Debian testing = trixie, X11, and nvidia proprietary drivers god have mercy on my soul.)

For a while XFCE only supported the type of scaling that makes things smaller. Understandably IMO this confused a few people and happily this has been upgraded and now it also makes things bigger. However in my experience this also makes things blurrier.

In my latest round of testing it appears that the situation can be fixed with a single setting: font DPI.

Settings Manager > Appearance > Fonts (tab) > Custom DPI setting > I chose 150, and logged out and in to have everything take effect.

From this single change everything is looking good in both GTK and Qt apps. I did also increase the size of my panel through the panel settings, and title bars are kind of tiny, but mostly I use maximised applications so I'm not stressing about this too much.

Hope this helps anyone else who is stuck in an "ohgod why couldn't we just stop after Windows 2000" love-hate relationship with computers.

This entry was edited (2 months ago)

Pixelfed fundraised $134,139 for development


Other noteworthy pixelfed things:

loops:
- loops source code released: mastodon.social/@dansup/114188…
- hashtags pinned videos: pixelfed.social/p/loops/798412…
- website: pixelfed.social/p/loops/781981…

This entry was edited (2 months ago)
in reply to fxomt

No, I don't think he is sleazy or have ill-intentions. I think he genuinely wants to do good things.

The problem it's just that he lacks focus and he worries more about feeling validated than dealing with the daily grind of continuously improving his product. As soon as any of his projects start getting a minimal amount of interest and people start depending on what he has promised, he finds himself some "new" project to be busy with. At the same time, he still feels possessive about his creations, so it's hard for him to just delegate away any significant part of the system.

I hope that the successful kickstarter makes him realize that shit got real and that he already got the validation that he was seeking, and that the money is enough to get him surrounded by good people who are a bit more focused on "proper" project management.

This entry was edited (2 months ago)

Distro for a Lenovo ThinkCentre Tiny PC


I've recently picked up a secondhand Lenovo ThinkCentre Tiny PC which ultimately I plan to use as a home server for a variety of functions (media server being the main one) but I don't need to use it as a main desktop.

It currently has Windows 10 installed. It has a decent sized SSD but no optical drive. I want to install a suitable Linux distro on it. What's a good Linux distro for my needs? And what's the best way of doing a scorched earth on the Windows install and replacing it with said distro?

I'm not entirely a Linux noob - I used to work on Xenix and later Red Hat Linux, but that was 30 and 20 years ago respectively!

Installed Mint


Mint is great but I have to fix my screwups.

My cheap laptop has a 128GB SSD. I have 30GB available .

I installed Mint on Sunday on a 15GB partition then realized I was immediately out of room.

Now I believe I have reformat, repartition to maybe 25GB and reinstall. Any better options?

Also, I could probably reformat the entire laptop if I could only figure out how to replace the Google Drive for sync backup for roughly 15 GB of personal photos and videos.

Technically, I wouldn’t have to do anything as it is already backed up but I guess I’d need a way to copy everything over to a Linux alternative that can be backed up from Mint.

Thoughts?

EU push for open source, GIMP3 is out, Firefox gets webapps back_ Linux & Open Source News




EU push for open source, GIMP3 is out, Firefox gets webapps back_ Linux & Open Source News


Try out Proton Mail, the secure email that protects your privacy: proton.me/mail/TheLinuxEXP

Grab a brand new laptop or desktop running Linux: tuxedocomputers.com/en#

👏 SUPPORT THE CHANNEL:
Get access to:
- a Daily Linux News show
- a weekly patroncast for more thoughts
- your name in the credits

YouTube: youtube.com/@thelinuxexp/join
Patreon: patreon.com/thelinuxexperiment

Or, you can donate whatever you want:
paypal.me/thelinuxexp
Liberapay: liberapay.com/TheLinuxExperime…

👕 GET TLE MERCH
Support the channel AND get cool new gear: the-linux-experiment.creator-s…

Timestamps:
00:00 Intro
00:44 Sponsor: Proton Mail
01:42 GIMP 3 is out
04:13 Firefox nightly adds PWA support back, sort of
06:36 GNOME 48 is out
08:38 Pebble is back with 2 new models
11:00 EU push for a EuroStack that competes with the US
14:14 First SteamOS powered handheld is available for preorder
16:23 AC Shadows is good on Deck, not on Linux with Nvidia
18:38 Sponsor: Tuxedo Computers
19:44 Support the channel

Links:
GIMP 3 is out
testing.gimp.org/release-notes…

Firefox nightly adds PWA support back, sort of
omgubuntu.co.uk/2025/03/firefo…

GNOME 48 is out
release.gnome.org/48/

Pebble is back with 2 new models
linuxiac.com/pebble-is-back-in…

EU firms push for a EuroStack that competes with the US
techcrunch.com/2025/03/16/euro…

First SteamOS powered handheld is available for preorder
forbes.com/sites/jasonevangelh…

AC Shadows is good on Deck, not on Linux with Nvidia
gamingonlinux.com/2025/03/assa…


Share your Bash prompts!


I'm looking for inspiration for a custom Bash prompt^[1]^. I'd love to see yours! 😊 If possible, include both the prompt's PS1, and a screenshot/example of what it looks like.

::: spoiler References
1. Type: Documentation. Title: "Bash Reference Manual". Publisher: Gnu Project. Edition: 5.2. Published: 2022-09-19. Accessed: 2025-03-21T02:46Z. URI: gnu.org/software/bash/manual/h….
- §6.9 "Controlling the Prompt". URI: gnu.org/software/bash/manual/h….
:::


Crossposts:
- sh.itjust.works/post/34840017


Share your Bash prompts!


I'm looking for inspiration for a custom Bash prompt^[1]^. I'd love to see yours! 😊 If possible, include both the prompt's PS1, and a screenshot of what it looks like.

::: spoiler References
1. Type: Documentation. Title: "Bash Reference Manual". Publisher: Gnu Project. Edition: 5.2. Published: 2022-09-19. Accessed: 2025-03-21T02:46Z. URI: gnu.org/software/bash/manual/h….
- §6.9 "Controlling the Prompt". URI: gnu.org/software/bash/manual/h….
:::


Crossposts:
- sh.itjust.works/post/34840018


This entry was edited (2 months ago)
in reply to Kalcifer

My shit is custom and rather elaborate.

Screenshot of the prompt

From left-to-right:

  • name@server-name
  • Uptime (multiplied by 10 and rounded to the nearest integer to save space)
  • Percentage disk space available on /
  • Number on established network connections
  • Git branch : commit
  • Python virtualenv
  • [new line]
  • date and time

The code for this is on GitLab.

This entry was edited (2 months ago)

Share your Bash prompts!


I'm looking for inspiration for a custom Bash prompt^[1]^. I'd love to see yours! 😊 If possible, include both the prompt's PS1, and a screenshot of what it looks like.

::: spoiler References
1. Type: Documentation. Title: "Bash Reference Manual". Publisher: Gnu Project. Edition: 5.2. Published: 2022-09-19. Accessed: 2025-03-21T02:46Z. URI: gnu.org/software/bash/manual/h….
- §6.9 "Controlling the Prompt". URI: gnu.org/software/bash/manual/h….
:::


Crossposts:
- sh.itjust.works/post/34840018


Share your Bash prompts!


I'm looking for inspiration for a custom Bash prompt^[1]^. I'd love to see yours! 😊 If possible, include both the prompt's PS1, and a screenshot/example of what it looks like.

::: spoiler References
1. Type: Documentation. Title: "Bash Reference Manual". Publisher: Gnu Project. Edition: 5.2. Published: 2022-09-19. Accessed: 2025-03-21T02:46Z. URI: gnu.org/software/bash/manual/h….
- §6.9 "Controlling the Prompt". URI: gnu.org/software/bash/manual/h….
:::


Crossposts:
- sh.itjust.works/post/34840017


This entry was edited (2 months ago)
in reply to Kalcifer

Not Bash, but zsh.

The host is there only because I'm usually logged in to 7 machines at once. The right prompt is the bare minimum path context, and the command # for easier !-re-execute; it disappears if the command gets long enough.

There's a bare minimum of cruft; no 2-line prompts, minimum verbosity in the prompt itself, because it's otherwise just noise that interferes with the output of whatever I'm doing and takes up space. It's especially important if I'm logged in via termux on my phone or something, where space is at a premium. What am I saying? Space is always at a premium.

$PS1 is

%{$fg[green]%}%m %{$fg[blue]%}»%{$reset_color%}

$RPROMPT is
%{$fg[green]%}%(5~|%-1~/…/%3~|%4~) %{$fg[yellow]%}%h%{$reset_color%}
in reply to Kalcifer

You can see here my theme with autodetect of Python (I work a lot in Python). The orange bit is the error code returned by the previous command. Git is supported as well, and looks pretty much like powerline-gitstatus, as you'll read about below.

NostraDavid's Starship theme, based on powerline-status + powerline-gitstatus

First, lets make Bash a little better:

# throw this in your `~/.bashrc`, and then `source ~/.bashrc` for it to take effect, or just restart your terminal.

# == shopts ==
# https://www.gnu.org/software/bash/manual/html_node/The-Shopt-Builtin.html
shopt -s autocd         # cd into folder without cd, so 'dotfiles' will cd into the folder
shopt -s cdspell        # attempt spelling correcting on folders
shopt -s direxpand      # expand a partial dir name
shopt -s checkjobs      # stop shell from exit when there's jobs running
shopt -s dirspell       # attempt spelling correcting on folders
shopt -s expand_aliases # aliases are expanded
shopt -s histappend     # append to the history file, don't overwrite it
shopt -s histreedit     # lets your re-edit old executed command
shopt -s histverify     # I'm confused.
shopt -s hostcomplete   # performs completion when a word contains an '@'
shopt -s cmdhist        # save multiple-line command in single history entry
shopt -u lithist        # multi-lines are saved with embedded newlines rather than semicolons; explictly unset
shopt -s checkwinsize # update LINES and COLUMNS to fit output

Autocd is a big one here, cdspell and direxpand as well. Ensures I don't need Zsh for the same experience. With Zsh I'd just get annoyed by small stuff like having to wrap things in quotes (I think pip install some_lib[some_extra] works in Bash, but not in Zsh And Ohmyzsh just felt it kept slowing things down, so I actually dislike Zsh ^(please don't kill me) >_>

Anyway, I based this on powerline-status + powerline-gitstatus (if you want to use that instead (no, my config doesn't work), just sudo apt install powerline-status powerline-gitstatus - DO NOT USE THE PYPI VERSION (it's too outdated, and a pain to install)) because I had issues with finding the right combination of my configuration and which libraries to install for bash - there are too damn many: powerline-bash, powerline-status, powerline-rs, powerline-go, etc, etc. And they all do things just that little bit different.

Anyway, here is my ~/.config/starship.toml (archive1, archive2). It's a little long, and can probably be shortered, but that might break something again, and I'm not willing to risk anything right now.

This entry was edited (2 months ago)

ReactOS 0.4.15 Released For This "Open-Source Windows" OS With Tons Of Enhancements


EU push for open source, GIMP3 is out, Firefox gets webapps back_ Linux & Open Source News


[Solved] How to get latest Mesa drivers for RX 9070


[SOLVED] - I learned that this is a sure way to break your Debian, and no matter how you go about it, you'd wish you either waited or used a different distro altogether for this purpose. In my case, I got Nobara 41 working, which already has the latest mesa. Trying to install the latest from Debian unstable almost got me pretty turned around, and I'm glad I switched course when I did.

Original post:
Apologies for my fairly low-level question. I spent all day yesterday spinning my wheels on this.

(I do fine using Debian Linux as my daily driver, but I'm not ashamed to admit that this (and things in this area) are beyond my experience. I've never compiled anything from source. I used to be a wiz with DOS 6.22 and Windows through 7, but my brain just stopped learning these things properly some time in the past.)

My distro (MXLinux 23.x) just announced they're almost ready to include Mesa 24.2.8. I purchased an AMD RX 9070, and all my Linux games (HGL or Steam) are angry that Vulkan can't recognize a valid GPU.

I see that Mesa 25.0.2 should work, but I don't know how to either build & install from source or add a repo for that particular package only.

I see that Arch users can easily use the Mesa-git or others, but not my Debian 12.

I installed Nobara to a spare drive as a stopgap, but on that install, FH5 refuses to prompt for account sign in there no matter which Proton I use.

Edit: I'm using the AHS version which includes the liquorix 6.13.7-2 kernel, and none of the repos (testing, back ports) show a higher version of Mesa.

EDIT2:

System:
  Kernel: 6.13.7-2-liquorix-amd64 [6.13-5~mx23ahs] arch: x86_64 bits: 64 compiler: gcc v: 12.2.0 parameters: audit=0
    intel_pstate=disable amd_pstate=disable BOOT_IMAGE=/vmlinuz-6.13.7-2-liquorix-amd64
    root=UUID=<filter> ro quiet amdgpu.ppfeaturemask=0xffffffff init=/lib/systemd/systemd
  Desktop: Xfce v: 4.20.0 tk: Gtk v: 3.24.38 info: xfce4-panel wm: xfwm v: 4.20.0 vt: 7
    dm: LightDM v: 1.32.0 Distro: MX-23.5_ahs_x64 Libretto May 19  2024 base: Debian GNU/Linux 12
    (bookworm)
Machine:
  Type: Desktop Mobo: ASRock model: B650E Taichi serial: <superuser required> UEFI: American
    Megatrends LLC. v: 3.20 date: 02/21/2025
CPU:
  Info: model: AMD Ryzen 5 9600X bits: 64 type: MT MCP arch: N/A level: v4 note: check
    family: 0x1A (26) model-id: 0x44 (68) stepping: 0 microcode: 0xB404023
  Topology: cpus: 1x cores: 6 tpc: 2 threads: 12 smt: enabled cache: L1: 480 KiB
    desc: d-6x48 KiB; i-6x32 KiB L2: 6 MiB desc: 6x1024 KiB L3: 32 MiB desc: 1x32 MiB
  Speed (MHz): avg: 3175 high: 5437 min/max: 3000/3900 boost: enabled scaling:
    driver: acpi-cpufreq governor: ondemand cores: 1: 2900 2: 2972 3: 2959 4: 3018 5: 5437 6: 2819
    7: 3000 8: 3000 9: 3000 10: 3000 11: 3000 12: 3000 bogomips: 93602
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
  Vulnerabilities:
  Type: gather_data_sampling status: Not affected
  Type: itlb_multihit status: Not affected
  Type: l1tf status: Not affected
  Type: mds status: Not affected
  Type: meltdown status: Not affected
  Type: mmio_stale_data status: Not affected
  Type: reg_file_data_sampling status: Not affected
  Type: retbleed status: Not affected
  Type: spec_rstack_overflow status: Not affected
  Type: spec_store_bypass mitigation: Speculative Store Bypass disabled via prctl
  Type: spectre_v1 mitigation: usercopy/swapgs barriers and __user pointer sanitization
  Type: spectre_v2 mitigation: Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on;
    RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected
  Type: srbds status: Not affected
  Type: tsx_async_abort status: Not affected
Graphics:
  Device-1: AMD vendor: Gigabyte driver: amdgpu v: kernel pcie: gen: 5 speed: 32 GT/s lanes: 16
    ports: active: DP-1 empty: DP-2, HDMI-A-1, HDMI-A-2, Writeback-1 bus-ID: 03:00.0
    chip-ID: 1002:7550 class-ID: 0300
  Device-2: AMD vendor: ASRock driver: amdgpu v: kernel pcie: gen: 4 speed: 16 GT/s lanes: 16
    ports: active: none empty: DP-3, DP-4, DP-5, HDMI-A-3, Writeback-2 bus-ID: 4f:00.0
    chip-ID: 1002:13c0 class-ID: 0300 temp: 42.0 C
  Display: x11 server: X.Org v: 1.21.1.7 compositor: xfwm v: 4.20.0 driver: X: loaded: amdgpu
    dri: swrast gpu: amdgpu display-ID: :0.0 screens: 1
  Screen-1: 0 s-res: 1920x1080 s-dpi: 96 s-size: 509x286mm (20.04x11.26") s-diag: 584mm (22.99")
  Monitor-1: DP-1 mapped: DisplayPort-0 model: Acer XF250Q serial: <filter> built: 2018
    res: 1920x1080 dpi: 90 gamma: 1.2 size: 544x303mm (21.42x11.93") diag: 623mm (24.5") ratio: 16:9
    modes: max: 1920x1080 min: 720x400
  API: OpenGL v: 4.5 Mesa 24.2.8-1mx23ahs renderer: llvmpipe (LLVM 15.0.6 256 bits)
    direct-render: Yes
This entry was edited (2 months ago)

Summit for Lemmy is now open source


cross-posted from: lemmy.world/post/27182706

At the time of writing Summit for Lemmy should be completely open source.

After a lot of contemplation I've decided to open source the app. There are still some changes that probably need to happen before it can be considered "FOSS" but this is a huge step I think.

As I'm voiced my opinions on this previously, the main repository for Summit will not be to the public for PRs. If you encounter bugs or have any feature requests please continue to send them through the proper channels (ideally to this community).

Enjoy~~

in reply to TK420

yes. Linux phone is mature with ubuntu touch (ubports.com/), even for calls. But, the difference to postmarketos is that ubuntu touch uses old linux kernel versions which are difficult to upgrade. postmarketos uses the newest kernel and develops drivers that go in the newest kernel. -> so postmarketos is future proof.
in reply to onlinepersona

lol. Just lol.

Edit: For those unaware.

"IIt’s also worth noting that plenty of Purism users have reported that they haven’t received the company’s previous Librem 5 phone, several years after ordering it. That phone first started shipping in 2020, while users report that Purism isn’t honoring refunds for these delayed orders or that the company is dragging its feet in this regard. So if it looks fishy and smells fishy, well, it must be a fish."
This entry was edited (2 months ago)

FOSS infrastructure is under attack by AI companies


LLM scrapers are taking down FOSS projects' infrastructure, and it's getting worse.
This entry was edited (2 months ago)

Here's an exercise in extreme masochism: The Linux Ship of Theseus


The Linux Ship of Theseus


Crossposted from: lemmy.ml/post/27387345

  1. pick any distro and install it.
  2. Then, without installing another distro over the top of it, slowly convert it into another distro by replacing package managers, installed packages, and configurations.
  • System must be usable and fully native to the new distro (all old packages replaced with new ones).
  • No flatpaks, avoid snaps where physically possible, native packages only.


Difficulties:


  • Easy: pick two similar distros, such as Ubuntu and Debian or Manjaro and Arch and go from the base to the derivative.
  • Medium: Same as easy but go from the derivative to the base.
  • Hard: Pick two disparate distros like Debian and Artix and go from one to the other.
  • Nightmare: Make a self-compiled distro your target.


Clarifications


  • chroot, dd, debootstrap, and partition editors that allow you to install the new system in an empty container or blanket-overwrite the old system go against the spirit of this challenge.
  • These are very useful and valid tools under a normal context and I strongly recommend learning them.
  • You can use them if you prefer, but The ship of Theseus was replaced one board at a time. We are trying to avoid dropping a new ship in the harbor and tugging the old one out.
  • It may however be a good idea to use them to test out the target system in a safe environment as you perform the migration back in the real root, so you have a reference to go by.
in reply to TheTwelveYearOld

Hmm. Considering the number of packages in a Gentoo base system and what package.provided does, [anything]->Gentoo shouldn't be that difficult. Of course, whether "work your way through about twenty packages one at a time, then just uninstall the rest of the old system and reinstall everything using Portage" violates the spirit of the challenge is another question.

Here's an exercise in extreme masochism: The Linux Ship of Theseus


The Linux Ship of Theseus


Crossposted from: lemmy.ml/post/27387345

  1. pick any distro and install it.
  2. Then, without installing another distro over the top of it, slowly convert it into another distro by replacing package managers, installed packages, and configurations.
  • System must be usable and fully native to the new distro (all old packages replaced with new ones).
  • No flatpaks, avoid snaps where physically possible, native packages only.


Difficulties:


  • Easy: pick two similar distros, such as Ubuntu and Debian or Manjaro and Arch and go from the base to the derivative.
  • Medium: Same as easy but go from the derivative to the base.
  • Hard: Pick two disparate distros like Debian and Artix and go from one to the other.
  • Nightmare: Make a self-compiled distro your target.


Clarifications


  • chroot, dd, debootstrap, and partition editors that allow you to install the new system in an empty container or blanket-overwrite the old system go against the spirit of this challenge.
  • These are very useful and valid tools under a normal context and I strongly recommend learning them.
  • You can use them if you prefer, but The ship of Theseus was replaced one board at a time. We are trying to avoid dropping a new ship in the harbor and tugging the old one out.
  • It may however be a good idea to use them to test out the target system in a safe environment as you perform the migration back in the real root, so you have a reference to go by.