Hello,

Last week-end, someone came to me with a peculiar issue: their Windows account was linked to a Microsoft account that somehow no longer exists, preventing them from logging in. From then, the solution appeared simple: transform the account into a local one.

Unfortunately, all the resources I found online presumes that you are logged into the account you want to unlink, which wasn't possible. And unfortunately, you can't unlink said account from an Administrator account either.

Since the computer was brand new and barely touched, I ended up enabling the Administrator account, creating a new local account and deleting the problematic account.

My question is: if the situation was to happen again, but I cannot delete spare to lose data on the locked account, what should I do?

EDIT/SOLVED: The culprit turned out to be a faulty offbrand 5V 3A DC power adapter. I tried another adapter - incidentally, one that I use to charge my smartphone - and the hissing sound was gone. My other router - also a Raspberry Pi 4B - uses the same power adapter as the one that started malfunctioning/hissing was using, but it is quiet and operates normally. At the very least I draw the following three conclusions:

1. 5V 3A is enough to run a Raspberry Pi 4B as a router with constant heavy TX/RX of at least a couple of terabytes per day.

2. I should be prepared for the other offbrand power adapter - namely the one that came with the OkDo bundle - to fail within a year or so.

3. There are really nice people here that give quick and elaborate responses! Thank you all! 🩵

The linked file is a recording of a hissing sound that my Raspberry Pi 4 started to make sometime during the last 24 hours. Another symptom is that the power on and CPU activity LEDs are out. It is not running hot. It doesn't have an internal fan. It is running OpenWRT. It has been pumping Linux ISOs in and out non stop for about four months.

I have tried cleaning it by dedusting it. I have also tried rebooting it. While rebooting, it squeals like a mouse, like it was suffering but the LEDs function properly. As soon as it loads user space though and resume routing, they turn off and the hissing sound continues.

Here are some system logs with severities warning and error . As you can see, the logs are from the 15th and I don't think any of them pertain to the hissing sound and/or the LEDs.

WARNING  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    0.196065] pci_bus 0000:01: supply vpcie3v3 not found, using dummy regulator  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    0.196201] pci_bus 0000:01: supply vpcie3v3aux not found, using dummy regulator  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    0.196257] pci_bus 0000:01: supply vpcie12v not found, using dummy regulator  
[May 15, 2026, 23:29:12 GMT+2] user.warn: [    7.225917] urandom-seed: Seeding with /etc/urandom.seed  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    8.327998] snd_bcm2835: module is from the staging directory, the quality is unknown, you have been warned.  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    8.365104] brcmutil: loading out-of-tree module taints kernel.  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    8.780919] brcmfmac mmc1:0001:1: Direct firmware load for brcm/brcmfmac43455-sdio.raspberrypi,4-model-b.bin failed with error -2  
[May 15, 2026, 23:29:12 GMT+2] kern.warn: [    8.792642] brcmfmac mmc1:0001:1: Falling back to sysfs fallback for: brcm/brcmfmac43455-sdio.raspberrypi,4-model-b.bin  
[May 15, 2026, 23:29:12 GMT+2] daemon.warn: dnsmasq[1]: no servers found in /tmp/resolv.conf.d/resolv.conf.auto, will retry  
[May 15, 2026, 23:29:18 GMT+2] daemon.warn: dnsmasq[1]: no servers found in /tmp/resolv.conf.d/resolv.conf.auto, will retry  
[May 15, 2026, 23:29:21 GMT+2] daemon.warn: odhcpd[990]: No default route present, setting ra_lifetime to 0!  
[May 15, 2026, 23:29:21 GMT+2] daemon.warn: odhcpd[990]: rfc9096: br-lan: piofile updated  
[May 15, 2026, 23:29:37 GMT+2] daemon.warn: odhcpd[990]: No default route present, setting ra_lifetime to 0!  
[May 15, 2026, 23:29:53 GMT+2] daemon.warn: odhcpd[990]: No default route present, setting ra_lifetime to 0!  
[May 15, 2026, 23:30:18 GMT+2] daemon.warn: odhcpd[990]: No default route present, setting ra_lifetime to 0!  

ERROR  
[May 15, 2026, 23:29:12 GMT+2] kern.err: [    0.307942] bcm2708_fb soc:fb: Unable to determine number of FBs. Disabling driver.  
[May 15, 2026, 23:29:12 GMT+2] kern.err: [    0.307968] bcm2708_fb soc:fb: probe with driver bcm2708_fb failed with error -2  
[May 15, 2026, 23:29:12 GMT+2] user.err: [    7.197094] insmod: module is already loaded - fat  
[May 15, 2026, 23:29:12 GMT+2] user.err: [    7.204851] insmod: module is already loaded - vfat  
[May 15, 2026, 23:29:22 GMT+2] daemon.err: procd: Got unexpected signal 1  

Judging by the squealing during reboot, I believe it to be either the CPU or memory, although I have never hears memory make sound, while CPUs, I have.

SOLVED: locking the screen put tethering in battery saving mode by default, which is totally insane. Anyway, modified the setting and now it works!

I am tethering my mobile connection from my Galaxy S23 to my Tab A9+. The A9+ looses its internet connection periodically, until I restart the hotspot on the S23.

I'm unable to figure out how to configure danneskjold onto my deltachat app. QR codes are too much for me to scan and use.

@delta@chaos.social

System sat on a shelf for about 10 years (running) and the feet melted. I need to remove the residue from a wood shelf, a cotton shirt, and the metal case. A putty knife and paper towels have been helpful, but there is still residue.

Hello!

TLDR: monitor started showing lines in the top photo, now it remains completely black except for a few pixels at the top. Can it be repaired, or do I have to buy a new one?

On Tuesday, my monitor (ASUS VA24EHE, 4.5 years old) started showing the lines that you can see in the top photo. Even though the lower 3/4 of the screen looked normal, they were frozen. After a couple of minutes, they went away and didn't come back. Everything worked as usual.

On Wednesday, it "broke" almost completely. There is a slim bit of screen visible at the very top (10-20 pixels), everything else is black. The bottom photo kinda shows that area. I took the monitor off the monitor arm to use my secondary as new primary, and after a few hours, I could magically use it again, just like on Tuesday.

Today, it remained completely unusable all day long.

What I already tried:

• a different cable (still the same issues -> it's not the cable)
• a different computer / my laptop (still the same issues -> it's not my GPU)

That makes me think it's the display itself that's the issue. It's most likely not a broken HDMI input, I think, because the same issues occur when I turn it on without any video cables attached. The bottom photo would usually show the ASUS logo, but it's completely black except for the top bit. I don't have DVI or VGA cables, nor does my computer have these outputs to test the monitor's other inputs.

I've looked online and managed to find people with similar issues, but I don't know if it's exactly the same or something slightly different. Most of them were "diagnosed" with a broken display. What I'm most interested in is whether you think this can be repaired because I like this monitor. I'd also be interested in possible causes of this damage.

I know it's hard to diagnose stuff like this with just a photo, but maybe that photo is a dead give-away that it's broken?

I'd appreciate any help before I lug it to a repair shop just to be disappointed. Even if the answer is "definitely dead, buy a new one". If you need any more info, just let me know. Thank you! And sorry for the terrible photos.

Can somebody please guide me how to go about activating my delta chat once again ??

From my laptop at home, I can access websites from Brave but not from Firefox or Mullvad Browser. I use Mullvad VPN but I am not split tunneling.

I used to upload and download my photos fine from my samsung mobile phone. This is being showed to me since the past 5 days all of a sudden.

Is it really a problem of my phone ??

The problem that l've been facing with @ente continues to persist. Sometimes Mastodon is better than email.

I was having a great time with #ente uploading photos from my Samsung galaxy phone, when one fine day, the website refused to open on my browser. And the problem has been persistent since. I've been writing to the support team frantically. I'm yet to get some productive reply.

Meanwhile, someone suggested me to download the app as that would give me a greater experience. Frankly speaking, l need to keep my phone in a usable condition and not make it a piece of an exhibition. So, l simply cannot install another app on my device.

Any suggestion anybody ??

@techsupport

1. Is it worth it to buy server MOBOs and employ one, then two or more CPUs? Does KDEnlive or other Linux apps make use of multiple CPUs? Do I need to buy i7-i9 CPUs from the start, or a couple of i5 is okay? Where is the golden middle there?
2. What about double vCards? Can any software use these dupletes or is it more beneficial to just buy the top single vcard? Do I need to purchase some compatible vcards for that?
3. If you use KDEnlive on AMD, what are lifehacks to boost it's performance? What do you use for preview, render, proxy clips, do you use another hard drive for cache, etc? What options do you choose for preview and render?
4. Have you tried VFX software like Natron (native), emulated Resolume, Nuke or Houdini via Proton or Wine, and how it went? What suffixes to use there?

My specific application is stage plays and performances by artists that I support with dozens of clips and effects played in the background. Thing is, now the base for those are mostly gifs and vids I downloaded somewhere, with some effects I added to them, rarely something I did myself from the blank.

I want to up my game by generating complex footages myself from the ground up, and for that I need a budget beefy PC I'd build on my own pocket budget. If it's possible, I would want to have a scheduled plan to grow it overtime from a half-capable initial workstation to the end game machine, while being able to work from the very start tp finish of said upgrade.

And there I need lemmy folks to help me with their knowledge if they tried that and have some suggestions.

To be transparent and honest: I'm a russian guy, currently still there, who never served and never used my skills to help promote war efforts or service, and never planned to. I want to build my skills and leave the fuck out once I figure out the situation with my gradually degrading elders I'm a caretaker of and EU immigration policies (as I'm poor). I'm paid leaves and marbles, I mean rubles, so my budget and availiability of hardware is very limited, but I'm ready to build the workstation for VFX over two or three years if it would enable me to start to work from some basic investment right now, and that I can then upgdrade it further.

I am just learning arrays and I am experimenting. This one simply won't work. I am trying to have the user define (1) how many elements an int array can hold and (2) what the value of each elements is. The program now outputs random integers. Maybe this isn't even possible, but I'm a complete beginner just playing around for the fun of it. XD

#include <stdio.h>  

int main(void) {  

int scope,elementnr,i,elementvalue,j;  
scope = elementnr = i = elementvalue = j = 0;  
int array[j];  

printf("How many elements should your ARRAY hold: ");  
scanf("%d", &scope);  
getchar();  

for (i = 0; i < scope; ++elementnr, ++i) {  
printf("Enter the value of element number %d: ", elementnr + 1);  
scanf("%d", &elementvalue);  
getchar;  
j = elementvalue;  
}  

printf("You have created an ARRAY that holds %d elements and you have assigned the following values:\n", scope);  

for (i = 0; i < scope; ++i) printf("%d\n",array[j]);  
}  

I am installing Artix Linux after three years of Arch goodness. The supported init systems are openrc, runit, s6 and dinit. I have no idea which one to choose or not to choose. Please advise! :D

I am comfortable with troubleshooting on my own, reading manuals, tweaking configs and I know a little C. In case this helps you advise me. :)

I understand that openrc is well documented over at Gentoo, but other than that, I only have experience with systemd.

Goal: a program that takes a char as input and outputs which char has been inputted and the humber of line(s). The below code makes the program output the inputted char twice: once - I assume - as a result of how getchar() works and once as a result of printf(). Is there any way to do this more elegantly, as in, not displaying the inputted char twice? If possible, keep to while, without suggesting for.

PS: the repetition of getchar() is just my ugly way of removing the trailing '\n'.

#include <stdio.h>  

int main(void)  
{  
        int c, nl;  
        c = 0;  
        nl = 0;  

        while ((c = getchar()) != EOF) {  
                getchar();  
                printf("%c, %d\n", c, nl);  
                ++nl;  
        }  
}  

I encrypted a message by:

1. Typing gpg --encrypt --armor --recipient [recipient]
2. Hitting enter
3. Typing the message to be encrypted
4. Exiting/finishing with "ctrl + d"

However, I needed to press "ctrl + d" twice, for some reason. Maybe a Termux bug? Although I haven't tried it on a normal Linux distro, so maybe this is normal?

Question: if it isn't normal, did gpg encode/include the first "ctrld" in the encrypted message? If it did, I need to contact the recipient to tell them to disregard the message... 🤣

Naturally, I cannot reverse the message since it's encrypted with the recipient's public key.

When an attacker believes that their target does not use secure passwords, they can use tools that compare the digest of the target's password to any of the precompiled lists containing the digests of the most commonly used passwords.

Question: what happens under the hood (hardware and software) when the attacker knows that their target does use secure passwords, possibly using a password manager to deploy passwords of, let's say, 30 characters, whose digests do not occur on those precompiled lists? Do they "simply" have the computer brute force every permutation? For a 30 char passwd using all the upper and lower case characters on an "English" keyboard (a-Z, 0-9, ~ - ?) (94 total), that would entail running 94ˆ30 permutations.

Am I missing something?

I'd like to process some Microsoft Word documents on my android tab. Do you have any recommendations on FOSS word editors that work with .doc and .docx on android?

Edit: my ISP says they only need the DCHP protocol and ports for basic functionality, as I have a public IP, so I will temporarily disable all the other rules and see whether that has any negative effects.

I am running a few routers at home using single board computers with OpenWrt. I never realized that OpenWRT comes with its own firewall configuration. I have firewalls set up on all my individual hosts, following this guide, and I guess the firewalls on my OpenWRT routers will serve non firewalled hosts, such as when I have friends over that connect to my network, and my own smartphones of course.

BUT!

Except for Ping, DCHPv6 and ICMPv6-Input, I don't understand what these other ones do. All of them are facing wan and I would like to close as many wan facing ports as possible if they aren't needed for normie consumer use. I do have a public IP because I'm running a Tor bridge to help our friends in China and Iran, but for that service, I have already opened and allowed dedicated ports and protocols not listed here. Otherwise, I'm just surfing the world wide web. No ssh , no telnet , no nothing.

Does, for instance, my ISP require DHCP-Renew to be able to rotate my public address when they need to (rarely happens)? Why would ICMPv6 messages need to be forwarded as opposed to simply be allowed in the input chain with the appropriate port? Why would a normie use IPsec (and what is ESP)?

• DHCP-Renew
• (Ping)
• IGMP
• (DHCPv6)
• MLD
• (ICMPv6-Input)
• ICMPv6-Forward
• IPSec-ESP
• ISAKMP

Edit: Resolved! The ISP added a BS suffix ".net" to my username, deleting it fixed the Auth issue.

Hi guys!

I'm having trouble migrating from TP-Link A7 v5 stock firmware to OpenWRT.

Specifically I'm having trouble connecting to my ISP via PPPoE.
Before the migration - the router connected w/o fancy PPPoE settings (MTU,VID...) to my ISP.

Here's my network setup:

1. FTTH
2. Nokia G-010G-Q GPON
3. TP-LINK Archer A7

Before migration - I took notice the mac address under "Internet" section in
"Status" tab under "Advanced" page on my router's admin web portal

Because I vaguely remember that my ISP technician paired the GPON to a MAC address

I have copied my PPPoE credentials from my ISP web portal,
and also cross reference them against the saved credentials in the TP-Link portal.

Post migration - I configured my WAN interface to use PPPoE , entered my credentials,
made sure that the WAN MAC address is the same as before,
and all I'm getting is AUTH_TOPEER_FAILED

I've exported my /etc/config/network here

I've also exported the output of logread -f | grep ppp here

After several trial and errors - I have given up and tried to use TFTP method to revert the
firmware back to to the TP-LINK stock firmware, also w/o much success
I used a Windows machine running tftpd64 , downloaded the latest firmware from here
renamed the file to "ArcherC7v5_tp_recovery.bin"
placed the file in tftpd64 server directory ("C:\Program Files\Tftpd64")
changed my LAN settings to 192.168.0.66 / 255.255.255.0 / 192.168.0.1
and performed the recovery boot steps.
I watched the logs of tftpd64 being filled with the download attempt,
watched it succeeds , watched the router quickly flashing the power LED,
but no matter how many times I attempted this - the router goes back to load OpenWRT.

Can someone help me either revert back to the stock firmware or help me fix my PPPoE settings?

I bought an LG C3 in 2023 and since then, I never updated it. I decide to updated it like 2 weeks ago, and, after the update, I noticed that, if I set HDMI Deep Color to 4k, almost all the clarity settings are grayed out, including the TrueMotion I'm interested in (I like it on Cinematic Movement).

Before the update there wasn't this issue. What can I do?

I tried:

• changing the cable
• change the input type to streaming box or Blu-ray player
• reset the TV

Info:

• I'm using the TV connect to a pc (with Linux if that matters);
• OS version: 33.22.95

Yesterday I noticed that one of my drives was acting up, it responded very slowly. It isn't my main drive and there was very little data on there, so I wasn't immediately worried, but I did copy some of the most important data to a different drive, just in case. I saved 2 folders and was copying a third, and that was when it stopped responding entirely.

I ran lsblk and it listed the device (/dev/sda), but no partitions (used to be sda1) and also no available space on the drive. That was when I figured I was in trouble and shut off the pc. I haven't turned it back on yet, will probably disconnect it physically first.

There is still some data on there that I would like to recover, but it's not really worth paying for professional recovery. Is there anything I can do? Even if it's a single-shot rescue mission and I trash the drive afterwards.

I am admittedly low on the learning curve for my understanding of all things IT related. Ive built machines for emulating tho. And depending on what the hardware is, ive downloaded shadder caches for certain cases to improve performance. My very loose understanding of it is they are universal when certain things match like emulator and version. Are they being included in some roms/emulators and i just didnt realize or is there a more technical reason they arent included?

I am on Windows 10 pro and have debloated and stripped as much tracking junk as I could from the OS, so Copilot had been long gone... until two days ago, when it reappeared on my start menu. I assumed it was sneakily reinstalled after an update so I uninstalled it.

Less than 30 minutes later, it was back. I uninstalled it again, and again it reinstalled after about an hour.

I've tried registry edits, I've tried changing the group policy, but it does nothing - Copilot keeps reinstalling itself in the time span of 20 minutes to an hour.

I know being on Lemmy I'm going to get a bunch of "Replace it with Linux!" replies so I'd like to preempt this by promising you that yes, that will happen eventually - there's just some compatibility issues with a couple of my daily driver programs that are pending a resolution.

For now I must stay on Windows, but Microslops incredibly aggressive Copilot reinstalls are pissing me off and I was just wondering if there are any other means in which I can get rid of this program, or at least hide the stupid ugly thing from my sight?

Edit: I suppose it's important to note that I am not nearly as tech-savvy as a lot of people here, but I try to get by. Please keep this in mind if I ask stupid questions.

EDIT: ended up being 1 faulty ram stick, 1 (supposedly) bad psu, an overcooked cpu, a dead aio cooler and (potentially) a faulty psu. Thanks Linux bros.

https://youtu.be/6LYqOmmMKnY

https://youtu.be/sXIbOk2CjWM

I did the memtest. Figured out 1 out of 4 ram sticks are bad and sent it out on an RMA. Took the whole rig to a local shop and he load tested the PSU. It failed twice. He recommended I replace the PSU then the motherboard if I still had issues. So I replaced the PSU with a new unit, still the exact same issues. So I replaced the motherboard with a used one. Still the exact same issues. I updated the BIOS. Same issues but a new, more red Asus logo (yay).

I still cannot, for the love of AntiChrist, boot the thing. At. All. Not even from a usb. Attempting to fresh install windows and i have zero input from any devices.

I am at my wits end and I need my PC back up and running.

Linux bros blocked immediately.

I'm buying a new laptop, and i need to compare some based on gpu. I heard a lot of the mainstream ones are biased, but i also need the big databases that the big ones provide, as im comparing some quite obscure gpus. thanks