In other non-news, I just "dnf5" updated my Asahi Linux install from Fedora 42 to 43. The only caveat I can provide is that this mac (m2 pro mbp) still runs un-updated from Sonoma 14-something, since I (quite) never use the Dark Side of this machine & never bothered to touch it. Having had to fiddle with a brand new such laptop recently, I can confirm that the more modern your apple device is, the greater the pain in your lower parts it is. #Asahi #Gnome #foss #linux

There is a webcomic called strong female protagonist that i want to persevere(in case the website is ever lost) but not sure how.

The image you see above is not a webpage of the site but rather a drop-down like menu. There is a web crawler called WFDownloader(that i am using the window's exe file inside bottles)that can grab images and can follow links, grab images "N" number of pages down but since this a drop-down menu i am not sure it will work

There also the issue of organizing the images. WFDownloader doesn't have options for organizing.

What i am thinking about, is somehow translating the html for the drop-down menu into separate xml file based on issues/titles, run a script to download the images, have each image named after its own hyperlink and have each issue in its own folder. Later on i can create a stitch-up version of the each issues.

i had cachyOS installed for a couple of months but was plagued with random system freezes (only hard reset possible, no leads in journalctl). i tracked it down to an issue with the combination of wayland, KDE plasma and the kernel or at least that's what i could gather from web searches. i had at least one of those freezes per week, often more.

i am now on kubuntu which basically has the same combination of things (wayland and KDE) that should cause the problem but it has been running fine for three weeks, no freezes. so something with the cachy kernel didn't agree with my system.

i was now told i could use the arch kernel on cachyOS, which was news to me. i tried switching to the cachy LTS kernel but the issue persisted. i now wonder how does the compatibility of the linux kernel work? is it compatible because it is both arch linux? or would the kubuntu kernel also work on cachyOS?

Some of you need to watch this video, and hang your head in shame.

Dylan Taylor has been receiving constant harassment, including threats to his life and safety, for actions done collectively by SystemD. The article by Sam Bent was explictly mentioned as part of the harassment campaign, and rightfully so.

I don't think enough people realize that this is catastrophically bad. It'll discourage people from becoming open source developers, it'll discourage people from using Linux, and it'll discourage legislators from taking the Linux community seriously.

If you ever wished ill upon another human being for complying with a relatively inconsequential law, you are better off never touching a computer again. The Linux community has collectively gone so far beyond what is acceptable here.

I finally separated some overkill hardware from my server rig (an RTX 3080 and a Ryzen 5800X3D) and put them in their own chassi. Installed Arch, again, of course. The server got a "new" Ryzen 5700G , which apparently is a repurposed laptop APU.

A few years ago, I used to fire up my desktop environment (dwm on X11) and launch Steam from a pseudo terminal.

On my reborn gaming PC, could I skip installing a desktop environment and still play games on Steam? Will it run directly from the console like - for instance - mpv does?

The purpose would be to minimize any overhead from dwm and X11. Which sounds ridiculous now that I "say" it out loud, because it's negligible, but still. My question still stands.

cross-posted from: https://lemmy.ml/post/44022276

Jolla may not be a household name, but for more than a decade the Finnish company has positioned its Linux-based Sailfish OS as an alternative to the mobile software duopoly that is Google’s Android and Apple’s iOS.

Now, 13 years since it tried to cut through the market with the Jolla Phone—a device which remarkably received software updates through 2020—it's back with a successor of the same name.

This time, the company is positioning its handset as the “European phone.” This bit of marketing caters to the growing distrust in US digital services and platforms that has arisen since Big Tech sidled up to the second Trump administration.

The new Jolla Phone (pronounced “Yolla”) costs €649, mimics the Scandinavian design of the original, and has secured more than 10,000 preorders since its preview in December 2025. Those orders are expected to begin shipping at the end of June. At Mobile World Congress 2026 in Barcelona this week, the company divulged more details about the phone's hardware.

Alt Android

Jolla has had a turbulent history. After the company floundered the launch of its Jolla Tablet in 2015, it nearly went bankrupt and pivoted to licensing Sailfish OS to automotive companies and governments, including Russia. After the invasion of Ukraine, Jolla had to cut ties with Russia, and a corporate restructuring meant that Jolla's assets were acquired by the company's former management under a new company called Jollyboys.

It got back into the smartphone game in 2024 with the Jolla C2 Community Phone, made in collaboration with a local Turkish company, and it was this experience that gave Jolla the courage to jump back into the hardware business with the new Jolla Phone. Unlike the C2, this device is completely assembled in Salo, Finland, where Nokia phones were manufactured more than a decade ago.

“Europeans want more European technology,” Sami Pienimäki, CEO of Jolla Mobile, tells WIRED. “People want to go away from Big Tech, and the other trend is that European people want sovereign tech—it makes it possible for our kind of company to have a position in the market.”

Building a smartphone from scratch was also much harder over a decade ago, but today, Pienimäki says the operation can be fairly lean without having to “pay too much up-front.”

The components are sourced from various vendors and countries. The MediaTek Dimensity 7100 5G chip hails from Taiwan; the 50-megapixel main and 13-megapixel ultrawide camera sensors are from Sony; the 8 or 12 GB of RAM is from SK Hynix in South Korea.

“There are Chinese components as well—we are totally open about it—but the key is that, as we compile the software ourselves and install it in Finland, we protect the integrity of the product,” Pienimäki says.

What makes Sailfish OS unique over competitors like GrapheneOS and e/OS is that it's not based on the Android Open Source Project, but Linux. That means it has no ties to Google—no need for the company to “deGoogle” the software; meaning there's a greater sense of sovereignty over the software (and now the hardware). Still, it's able to run Android apps, though the implementation isn't perfect. Another common criticism is that it's not as secure as options like GrapheneOS, where every app is sandboxed.

There's a good chance some Android apps on Sailfish OS will run into issues, which is why in the startup wizard the phone will ask if you want to install services like MicroG—open source software that can run Google services on devices that don't have the Google Play Store, making it an easier on-ramp for folks coming from traditional smartphones without a technical background. You don't even need to create a Sailfish OS account to use the Jolla Phone.

Jolla’s effort is hardly the first to push the anti–Big Tech narrative. A wave of other hardware and software companies offer a deGoogled experience, whether that’s Murena from France and its e/OS privacy-friendly operating system or the Canadian GrapheneOS, which just announced a partnership with Motorola. At CES earlier this year, the Swiss company Punkt also teamed up with ApostrophyOS to deploy its software on the new MC03 smartphone. Jolla is following a broader European trend of reducing reliance on US companies, like how French officials ditched Zoom for French-made video conference software earlier this year.

Murena CEO and founder Gaël Duval wrote in a statement emailed to WIRED that the company believes it has a different mission from the Jolla Phone as it's trying to bring the existing mobile app ecosystem—minus the permanent data collection by Google and third-party trackers—without a learning curve for the average person. “We want to make privacy possible for the everyday person without the need for technical expertise or a development background,” he says.

The Phone

A common problem with these niche smartphones is that they inevitably end up costing a lot of money for the specs. Take the Light Phone III, for example, a fairly low-tech anti-smartphone that doesn't enjoy the benefits of economies of scale, resulting in an outlandish $699 price. The Jolla Phone is in a similar boat, though the specs-to-value ratio is a little more respectable.

It's powered by a midrange MediaTek Dimensity 7100 5G chip with 8 GB of RAM, 256 GB of storage, plus a microSD card slot and dual-SIM tray. There's a 6.36-inch 1080p AMOLED screen, the two main cameras, and a 32-megapixel selfie shooter. The 5,500-mAh battery cell is fairly large considering the phone's size, though the phone's connectivity is a little dated, stuck with Wi-Fi 6 and Bluetooth 5.4.

Uniquely, the Jolla Phone brings back “The Other Half” functional rear covers from the original. These swappable back covers have pogo pins that interface with the phone, allowing people to create unique accessories like a second display on the back of the phone or even a keyboard attachment. There's an Innovation Program where the community can cocreate functional covers together and 3D-print them. And yes, a removable rear cover means the Jolla Phone's battery is user-replaceable.

Pienimäki says that while the device doesn't have FCC approval, you can theoretically import it into the US, and it should work with the major US carriers, though compatibility is rarely a given. Jolla is considering a separate US launch, though right now it's focusing on the European Union, the UK, Norway, and Switzerland.

Antti Saarnio, Jolla Group’s chairperson, reiterates that the Jolla Phone will be a niche product. “Most of the people using Android or iOS will not switch, but we should treat this as a stepping stone for something new,” Saarnio says. The “path to real volume” will come from the mobile market breaking down into new form factors, powered by artificial intelligence.

He's likely referring to Jolla's Mind2, a privacy-focused AI computer, which is still in active development. It plugs into a PC and connects Jolla's AI assistant to apps like email and calendar locally—no cloud access required. The chatbot-like interface lets you ask it questions about your data, whether you're fishing for something from an email or a private message. While the new Jolla Phone won't have any AI capabilities at launch, Saarnio says an integration will be an option users can enable later this year.

Jolla has street cred for supporting its devices for a long time, but we'll have to wait and see how the fresh hardware holds up and just how much the company has polished the Sailfish OS experience, especially since it's much easier today to get started with a deGoogled Android alternative.

My system is running Debian 13, and has been running Debian great for well over a year however, recently when I went to reboot my computer KDE Plasma (X11) froze and didn’t want to log in, I found it odd and rebooted as per usual but it repeated itself yet again.

I jump into another TTY and start checking the journal, nothing out of the ordinary, obviously annoyed I start reinstalling packages.

kde-full, kde-standard, kde-plasma-desktop, sddm, nvidia-driver, linux-generic-headers xorg and so on. No luck. I figured I would give Wayland a try even though a lot of my software still does not support it, and to my surprise loaded up instantly, so I got some hope my system isn’t borked, I tried X11 again but instead of rebooting or shutting down after it froze I just left it to see if anything at all changes and after a while it decided to load my desktop!

So after a few more days of trying to catch something in my journal I finally noticed this 3 minute gap in these entries of my journal.

3/23/26 9:44 PM systemd systemd-timedated.service: Deactivated successfully. 3/23/26 9:45 PM systemd-timesyncd Timed out waiting for reply from 84.16.67.12:123 (2.debian.pool.ntp.org). 3/23/26 9:45 PM systemd-timesyncd Contacted time server 217.147.208.1:123 (2.debian.pool.ntp.org). 3/23/26 9:48 PM systemd Reload requested from client PID 2681 ('startplasma-x11')... 3/23/26 9:48 PM systemd Reloading...

I don’t have much to work off of but I’m guess this is what is kicking my system back in order, is there a way I could reduce the timeout of the above systemd request?

Update

So i manage to capture the time it takes for me to reboot and land back at my desktop, it's roughly ~30 seconds to reboot and land at sddm but another whopping 5-6 minutes to actually load the X11 desktop.

I captured the logs within this time frame to hopefully weed out the issue I'm encountering.

Log File (via CopyParty)

cross-posted from: https://reddthat.com/post/62484268

Please explain to me. I moved away from Big Tech and installed - even on my old MacBook Pro 2015 - Linux Mint. I use open source software and my social media is on Fedivers. I tought I was "safe" by using Linux, but the Linux Foundation is sponsord by a lot of money by Microsoft, Amazon, Meta, Google, etc... etc... the exact companies I try to take some distance off. Can somebody please explain me if Linux is "sold" to US Big Tech now?

I have a Canon Maxify 2750.

Up until now I've been using the standar linux drivers for it but they have a serious limitation I really need to solve which is I can only print one page at a time, on one sheet of paper. Printer is capable of full duplex printing.

If been able to install the drivers before but now I'me being confronted with an unexpected _"("_ error. Nothing more, nothing less.

I run the install script and it just fails.

Can someone lend a hand on this, please?

This is a small guide to setup your Linux desktop to view 360 degree interactive photos - also called photospheres - and panoramas in Linux.

Regular image viewers generally don't support 360 images. There are very few native Linux viewers out that that supports them. The best I know of are:

• SphereView
• Lux

SphereView is available as a Flatpak for amd64 and arm64 platforms (I use both myself), while the author of Lux provides packages and AppImages but only for amd64.

What's more, I prefer the way SphereView uses the mouse to pan and tilt. So I recommend SphereView.

To install it, open a terminal and simply install the Flatpak from Flathub:

flatpak install io.github.dynobo.sphereview

And that's pretty much it for the basic installation: when you want to view a 360 image, right-click on it, select Open with, then select the SphereView application (your mileage may vary depending on the particular file manager you use, but this seems fairly universal).

But what if you want to automatically open a 360 image in SphereView and a flat image in your regular image viewer?

Unfortunately, SphereView doesn't render flat images correctly, so you can't use it as your default image viewer. But it's possible to write a small "shim" script that replaces the default image viewer, that inspects the image(s) the viewer is supposed to open, determines those that are flat and those that are spherical, then opens the flat images in the regular viewer and the spherical images in SphereView.

To do this:

• Install zenity and exiftool. On a Debian-based system for example, do:

sudo apt install zenity exiftool  

• Create a text file called auto_open_image_as_normal_or_photosphere.sh in your path with the following content:

#!/bin/bash  

FLATIMG_VIEWER="gtk-launch org.gnome.gThumb.desktop"  
PHOTOSPHERE_VIEWER="gtk-launch io.github.dynobo.sphereview.desktop"  

if [ $# -eq 0 ]; then  
  IFS='|' read -ra ARGS <<< $(zenity --title "Choose one or more JPEG images to view" --multiple --file-selection --file-filter="*.jpg *.jpeg *.JPG *.JPEG")  
else  
  ARGS=("$@")  
fi  

FLATIMGS=()  
PHOTOSPHERES=()  

for FILE in "${ARGS[@]}"; do  
  if exiftool -X -xmp:ProjectionType "${FILE}" | grep -i ProjectionType > /dev/null 2> /dev/null; then  
    PHOTOSPHERES+=("${FILE}")  
  else  
    FLATIMGS+=("${FILE}")  
  fi  
done  

if [ ${#FLATIMGS[@]} -gt 0 ]; then  
  ${FLATIMG_VIEWER} "${FLATIMGS[@]}"  
fi  

if [ ${#PHOTOSPHERES[@]} -gt 0 ]; then  
  echo TOTO  
  ${PHOTOSPHERE_VIEWER} "${PHOTOSPHERES[@]}"  
fi  

I like to put all my scripts in a ~/scripts directory in my home directory that I added to my PATH. What follows assumes the script resides in /home/user/scripts.

Also, the script - and the explanations below - assume spherical images are only in JPEG format. I only use JPEG for photospheres personally. If you use other formats, adapt the script and the installation as needed.

This script assumes you have Gtk installed, and your default image viewer is gThumb. Replace the xdg launcher gtk-launch and/or the viewer org.gnome.gThumb to the launcher and image viewer of your choice.

If you want to reuse the default image viewer to view JPEG images, you can find out which one it is currently set to by doing:

xdg-mime query default image/jpeg  

• Make the script executable:

chmod +x ~/scripts/auto_open_image_as_normal_or_photosphere.sh  

• The script needs a .desktop entry so it can be used as the new default application for the image/jpeg mimetype: create ~/.local/share/applications/auto_open_image_as_normal_or_photosphere.desktop with the following content:

[Desktop Entry]  
Name=Automatically open image as a normal image or as a photosphere  
Exec=/home/user/scripts/auto_open_image_as_normal_or_photosphere.sh %U  
MimeType=image/jpeg  
Terminal=false  
Type=Application  

• Finally, change your default viewer for the image/jpeg minetype to the script:

xdg-mime default auto_open_image_as_normal_or_photosphere.desktop image/jpeg  

And that's it! Now when you open an image in your file manager, SphereView will be used to view it if it's a properly-formatted 360° image, as shown in the video.

cross-posted from: https://lemmy.ml/post/44893731

The 6.11 release for Qt Framework is now available, with improved performance, newly supported techniques and capabilities on graphics, connectivity and languages, not to mention a whole new approach to asynchronous C++ coding.

• Hardware-Accelerated 2D Rendering: A new module, Qt Canvas Painter, based on the HTML Canvas 2D Context, provides performance & productivity gains.
• 3D Improvements: New rendering techniques Screen Space Global Illumination (SSGI) as an option for lightmap baking, and Screen Space Reflections (SSR). Also imrovements on the Temporal Anti-aliasing algorithm with motion vectors. New user-defined render passes for post-processing effects, color picking, layer masks, etc. directly in QML.
• Interactive Graphs: You can now implement custom graphs where a user-defined delegate renders each data point. There's a new Qt example, the Wind Turbine Dashboard, and many improvements, e.g. new ways to style line graphs, and multi-axis support on 3D graphs.
• Declarative Approach to C++: Qt Task Tree brings a whole new approach to asynchronous coding and C++ API design in Qt. In addition, various APIs have been unified to allow adapting any asynchronous task to work with the new module.
• Other Improvements: Improvements on vector graphics, controls, and accessibility. Connecting to web servicers is now easier with the new module, Qt OpenAPI. Navigating in an IDE between QML and C++, and making data available from C++ backend code to Qt Quick have gotten easier. A wealth of other improvements, such as for multimedia, Android, and API documentation.

I feel pretty good about this as I have only had Linux installed as my daily driver since late October 2025. This machine is the only exposure to Linux I get, as I work as a Windows sysadmin. I run openSUSE LEAP 16.0 with KDE and while I can't say I'm comfortable or even within spitting-distance of being comfortable with it, I feel like today moved the needle a bit more towards that.

This started a few days ago with my three displays. I run an LG 34" curved display as my main monitor and two 27" CRUA curved displays on the sides of it. Previously, I had experienced no issues with this setup when using Bricklink Studio 2.0 via wine. However, on Thursday night I quit Studio and boom, my side monitors wouldn't stay on or detect a signal, and my main display kept freaking out and blinking every 5-7 seconds. I could get one of the two side monitors to work, but not both with the main monitor.

Long story short (DP->HDMI adapter swaps, cable changes, port arrangements with the graphics card, etc.), I used DuckDuckGo searches (lots of the results came from the Arc forums, my consolences) and was pointed toward log files for kwin. I used the Logs app on my machine to check the important logs that would appear when I tried to have both monitors plugged in. That showed me that it was having trouble finding or removing some reference object. I looked in the Display Configuration settings and noticed the monitors would pop up, last for about 5-7 seconds, then get disconnected within the same time frame as the logs. I also noticed that when they would be visible, the 'Enable' checkbox would be unchecked.

So with my trusty vertical mouse in hand, I studied the placement of the buttons and checkbox and after a few fails, successfully selected the checkbox to enable one of the displays, apply the change, and select keep before it could fully disconnect the monitor. Boom! The monitor turned back on and stayed on. I had to adjust it's position in the layout, but after that, it had no issue being on! I repeated this for the other monitor and now, I am happy to say, all three of my monitors are on and my system is running exactly as before!

I really appreciate the openness to information that I see in many of the Linux communities, and thank you to those of you who have contributed, or will contribute to that knowledge. Because of people keeping that information open and available, a complete and utter Linux-n00b like myself can take a shot at investigating and fixing my own system woes.

Best regards!

P.S. I have a theory about what happened with wine and why the issue wouldn't happen with one of the side monitors plugged in, and only happen when both were. But I'll save that for a comment if someone asks.

Hey there folx, I'm getting ready to go back to school and I was curious of any distro that have good pen support. The basic use see is likely just reading studies and being able to highlight in the PDF.

I'm looking at either a new framework12 or trying to find a surface to meet these needs. Likely won't be until next year I'm in school so I have time to tinker and troubleshoot.

I have installed noto-fonts-cjk but the Linux kernel terminal won't display Japanese/Chinese characters.

Environment:
ThinkPad T480
Arch Linux

Edit: I managed to track down this page https://wiki.archlinux.org/title/Linux_console and realize that I need to setfont. Now I just need to find which one in /usr/share/kbd/consolefonts can display Japanese, I suppose.

Edit2: looked through https://adeverteuil.github.io/linux-console-fonts-screenshots/ to no avail.

Edit3: this thread https://askubuntu.com/questions/193391/how-do-i-display-chinese-japanese-characters-in-a-linux-vt-console hints at this not being a possibility, because - paraphrased - the framebuffer can't hold 2000 characters, which is what I would like it to be able to display.

Edit4: as pointed out in the above askubuntu thread, and as @mina86@lemmy.wtf secondarily recommends, fbterm seems to be the solution: https://wiki.archlinux.org/title/Fbterm

Edit5: kmscon also seems to be able to display CJK: https://wiki.archlinux.org/title/KMSCON

Marking as SOLVED, but I have not yet tried replacing the Linux kernel terminal. That'll have to wait until next weekend.

I read on PC, how can I save power and make PC last longer by under powering the PC ?

On Windows I put max CPU state to 1%, PCIe and WIreless Power to power saving mode.

Also related question, How to disallow software from inhibiting power management?

I’m planning to switch from Manjaro to pure Arch, but I really like the easy setup Manjaro provides. Are there any guides, wikis, or personal tips to replicate that setup on a fresh Arch install? Looking for something that covers the essential packages, configurations, and tweaks to get a similar out‑of‑the‑box feel. Thanks in advance!

Hi,

I want to use openntpd to sync my clock

I'm using

ntpd -ds

I see in my firewall that the dns resolution is working, and I get a server IP from the pool 👍

but anyhow I get

ntp engine ready no reply received in time, skipping initial time setting
no reply from x.x.x.x received in time, next query 300s

Weird my nftable config file should allow it:

# extract
chain OUT {
type filter hook output priority 0; policy drop;
udp dport 123 accept
}

chain IN {
type filter hook input priority 0; policy drop;
ct state established, related accept
}

Any ideas, or which lemmy community to cross-post ?

Thanks.

California's law, as written, contributes vastly to privacy and free speech. It is a fundamentally fair and democratic law, which additionally acts as a safeguard against fascism.

Point 1: Age-Verification is coming, whether you like it or not. Multiple states are pursuing such laws. Approaches adopted by Discord and others include techniques such as AI facial scanning with severe privacy concerns. California's law will set the standard for an 100% robust solution that completely eliminates any pretense that facial scanning, uploading copies of identifying documents to every web service you use, etc. are necessary to implement age validation. This standard explicitly leaks the least possible amount of information about you as humanly possible to achieve this goal. This legislation preempts further attempts to invade user privacy to "protect the children", by creating an actually effective and pragmatic privacy-preserving age verification scheme, which will undercut any future laws using "protect the children" as an excuse for privacy violations.

Point 2: Good legal standards are beneficial to everyone. Many complain this helps Meta, etc. This helps everyone creating websites, including federated social media like Lemmy. If there is a concern federating with epstein.ml, at least one less concern is threat of legal repercussions potentially affecting large parts of the community-run open internet. This law reduces the burden significantly, avoiding the need to adopt (paid, proprietary) age verification services. This standard absolutely obliterates such commercial offerings in favor of a non-commercial common standard. It even goes so far as to forbid third-party sharing of this data (e.g. for tracking and advertising) and collection of additional information beyond what is necessary to implement the law.

Point 3: You already leak more information that your age bracket through regular web browsing - IP determines approximate location, your cookies tell a whole bunch more, and sophisticated ad trackers have a whole profile on you. In terms of Shannon's information theory, this new law is leaking less than 3 bits of information about you. Unless you are in the 0.1% of people deliberately avoiding such pervasive tracking, major corporations tracking you online already know your age bracket to fairly high accuracy (but not to sufficient standard to satisfy legal obligations).

Point 4: Fighting fascism has preventative and active measures. Making fascist-resistant technology is an important part of that. This technology has zero potential for fascist abuse. In fact, it undercuts the existence of other tech solutions crafted by your friends at Meta and the NSA. Ensuring quick, near-universal adoption of this standard will solidify an explicitly fascist-resistant piece of technology as a defacto standard.

Point 5: The harm this law aims to address is grave and real. For the 99% of the population who aren't compiling their own kernels, the ability to "age-lock" a child account to prevent young children from accessing doomscroll brainrot on Instagram is an amazing and valuable feature. Lack of such protections is a compelling enough concern that it has time and time again had enough popular support for sway legislators to take up the issue. The principled "linux source code is free-speech, and no government mandates can compel changes" stance is quite divorced from reality. Are crypto-exchange founders likewise free to implement whatever fraudulent schemes they like, as their source code is their speech to freely dictate? People are sick of strangers shoving content down their children's throats. In a democratic society, when the wise and ever-just "free market" fails to solve a pressing issue or exacerbates it beyond recognition, it is fair for the state to step in and solve it striking a fair balance, not trampling on anyone's rights.

Point 6: "Slippery slope" does not apply here - maybe focus more of your attention on ICE, which appears to be slippery when it comes to constitutional rights like due process and privacy and equal protection. We should be proclaiming this law as a paragon, in the way it codifies a clear line beyond which privacy invasions are unjustified. It is a work of legal genius, and society and all of you will come to appreciate how prescient it really is next time an actual bullshit law to "protect children" comes to the table. This law itself is precendent-setting, but in a way that is quite favorable to both privacy and free-speech rights long-term.

The above considerations are not made lightly. I am someone who has put considerable thought into extricating corporate control of communication media, who cares deeply about the ecosystem which can support humanity or drive us into deadly peril and destruction. I honestly thought I would see more nuanced discussion here, but it seems I am alone as of this moment in my perspective.

cross-posted from: https://lemmy.ml/post/44815211

Two-way file sync, no remote agent needed

Today Synchi is finally public! It's designed for syncing files between two locations (local or over SSH). It detects conflicts, and lets you decide what to do.

Why not rsync/Unison/Syncthing?

• rsync has no memory between runs and is one-way
• Unison needs to be installed on both sides
• Syncthing requires always-on daemons

Synchi runs on demand, works over SSH, and only transfers what actually changed.

• GitHub: https://github.com/jakobkreft/synchi
• Docs: https://jakobkreft.github.io/synchi/

I use it daily for syncing a shared folder between my machines and an android phone. Works great in combination with Tailscale/WireGuard so that you can sync files remotely.

Not that it matters much, ultimately it's about becoming familiar with where stuff is put, even if it's in a weird grab bag of /usr, /var, /etc/etc/etc. Still, I can't help but check out Gobolinux from time to time.

They’re basically minimum-viable products that by design can be used to violate the law in California when the Act goes into effect on Jan. 1, 2027.