https://f-droid.org/packages/com.github.catfriend1.syncthingfork

I feel like by “scaling” they mean upgradability. So either vertical (adding more drives, ram, cpu) or horizontal (adding more boxes that loadbalance an increase of multiple parallel tasks/users) hahaha ooops

https://github.com/eikek/sharry

I mean, if you try to “scam” the gov, you can clone some codeberg repo to github, rename it, rewrite history to make the commits look like you did everything and then tell the gov “look at how much work I volunteered”. At least in germany, there are currently not enough public workers so many little things go unchecked.

afaik the Ruhr Universiry of Bochum has an intranet that connects the uni and all the dorms. And they selfhost a couple of services, like email, git and pastebin. You can see a line going to the dorms on the graph.

https://noc.rub.de/web/services/start

https://nedi.noc.rub.de/netweather/

a git history is easily fabricated. you can freely edit it, remove entries or write into it whatever you want, including impersonating other users and fabricating datetime

https://github.com/Amog-OS/AmogOS/commit/4f503a0

I use arch btw. Do you also use arch btw?

It’s really touching that you consider me to be a sysadmin, because I use Linux and know how my fs works. I’m actually kinda proud of myself. My arch install has been working for many years.

Linux filesystems exam time:

section A basics

0. what does CoW stand for?

1. evaluate through pros and cons which you personally would pick: Btrfs, ZFS, F2FS, bcachefs, OverlayFS, aufs, Nilfs2, JFFS2, UBIFS

section B btefs

2. what exactly happens when running this command, including how qgroups, compression, and recursion interact?

btrfs balance start -dusage=5 -musage=20 -c zstd \
    --bg /srv/vms && \
btrfs qgroup limit 50G /srv/vms/guests/win10

3. explain the effect of the following command on device allocation, RAID reshaping, and metadsta distribution:

btrfs device add -f /dev/nvme2n1 /home && \
btrfs balance start -mconvert=raid1 -dconvert=raid1 -sconvert=dup /home

4. describe what this snapshotting pipeline does, including send stream structure, parent selection, and how receive-side overwrites are handled:

btrfs subvolume snapshot -r /opt/app /opt/.snaps/auto-$(date +%s) && \
btrfs send -c -p /opt/.snaps/last-full \
    /opt/.snaps/auto-$(date +%s) \
    | ssh backup 'btrfs receive -f /backup/opt/incoming'

5. what actions occur on the filesystem when this defrag call is executed, especially regarding extent sharing and how compression interacts?

btrfs filesystem defrag -r -v -czstd:15 \
    /var/lib/docker/overlay2

6. analyze subvolume management sequence, including how default-subvolume selection influences mount behavior:

btrfs subvolume delete /mnt/root/@old && \
btrfs subvolume snapshot -r /mnt/root/@clean /mnt/root/@ && \
btrfs subvolume set-default 256 /mnt/root

section C zfs

[…]

/s

if the government fucks around and gets Signal banned

you can use threema or wire or something else will become available on the appstore

Yes, but how though?

We’ve stopped chat control over and over for years. But now they seem to have went through with it.

Usually the linked website has instructions. Now it simply says:

About the Vote: The Council mandate was today endorsed by the Committee of Permanent Representatives (COREPER). About the Procedure: The text will now be negotiated with the European Parliament. The Parliament’s mandate (adopted in Nov 2023) explicitly rules out indiscriminate scanning and demands targeted surveillance based on suspicion.

If there is an organized protest in front of the parlament, link it please.

Whats the difference between sideloading and installing? I don’t get it tbh.

On windows for example, it is the norm to go online to a website, download an exe and double click it - instead of going to the microsoft store. And both are ways to install software. Neither way is doing it sideways.

But also

I can easily see fork related infrastructure for end users making it really eash or automatic to keep up with that cycle.

There could be a store that installs software from github (there already is) and additionally tells you “hey this app you have got forked and the fork has more development happening, do you want to switch to that one and migrate all your settings? y/n”

The current proposal as far as I understand it is that they will tell a company/service to “voluntaraly” scan the content.

Then an open source project that is forced to comply can just stop development and be forked on the last commit and shrug shoulders.

Which will buy enough time, because legally making those requests will take time and work. I can easily see fork related infrastructure for end users making it really eash or automatic to keep up with that cycle.

But even better would be to develop open source hardware based communication solutions, that don’t have the ability to add a backdoor through a software patch. Think like a pager with all the security and privacy bells and whistles.

Cursor AI, operated by …

dang son

deleted by creator

you bought one? Like in - one time payment with infinite time ownership?

my 2 old phones were both replaced because of hardware issues.

one has a broken power button that is constantly randomly spam pressed, so as soon as it gets power if it ever manages to boot it either reboots itself or tries to call emergency services.

the other has a flaky usb c that constantly connects and disconnects unless you hold it at a specific angle and the battery is like a minute away from meltdown - it can hold charge for a couple minutes.

I’d be afraid to leave either plugged in.

I guess the article is for people that buy a new phone yearly out of fashion, not need.

arch

no

I found it pretty easy to see with thedynamic examples. It flashed “no result” for a second before showing a result for example, happened, especially the first time when searching something after a fresh boot.

For a static example - what the fuck is this result ordering? What do any of those apps except keepass have to do with “ke”?

here is another, when I want to start teamspeak, steam is the first result like WTF?

I find it is absolutely insane and deranged UX that the results where the app name STARTS with the entered substring are not at the top of the list.

I tried to reproduce those out of curiosity for a couple of minutes without much effort and I immediately saw that I have about 40% of those…

But the other 60% seem weird and my menu refused to behave like that. Maybe thats just luck, maybe it depends on version and distro. idk.

Very interesting and detailed post regardless.

never backquotes, it’s unintuitive