Thread 42637040

Previous Thread: >>42619956
https://desuarchive.org/mlp/thread/42619956

Discuss:
>Work on any pony and/or tech related projects (You) are working on or learnt about recently.
>Post (You)r pony themed technology (desktops/rices, papes, new devices, software ponies)
>What software and technology do (You) use? (Email, Git, OSes, Messaging, Monero, etc...)

>finally man up and decide to switch to freetube
>the same day all invidious instances are down
>/pts/ gets derailed

>>42637044
Anon, you don't need invidious to use Freetube
source: literally me

>>42637442
Local is broken too

>>42637449
use external player

>>42637451
Okay, that's a good start. The video is choppy however, and I don't seem to have any way of controlling the quality.

>Windows 11 is four years old
wtf? It still feels new for my Debian brain
>Win10 LTSC enters EoL in about a week
WTF

>>42637481
only LTSB 2015.
LTSB 2016 next year
LTSC 2019 in 2029 or something
2021 in 2032
2024 in 2036 or 2034, I don't fucking remember.

>>42637485
yeah just checked the WP, looks like it's just normalfag Win10 and the very first LTS build of Win10

>>42637497
It all really depends on how quickly the software will be dropped since Windows is powered by third parties. Windows 7 had a better run from XP and Vista which were abandoned rather swiftly by around 2015-16 by everything major. 7 still gets updates from some software. I wouldn't trust them security wise though, even though nowadays all you need is a good adblocker since all viruses are now adware.

>>42637520
>I wouldn't trust them security wise though
Anon, you can't trust security with proprietary shitware even if it appears supported

my puter for uni

>>42638030
I like how the placement is done. It's as if AJ jumps across the Earth. Gotta appreciate how the chinks are using earlier season renders. That S1 Luna is beautiful

>>42638030
>wangblows logo
your laptop must be scorched by the flame of freedum

>>42638082
nta but Surface laptops aren't necessarily bad if you want ARM battery life but don't want to use Mac. Their build quality is also better than your average Dell or HP nowadays. The only issue I see for now is the lackluster ARM support for mainstream Linux distros like Mint on it.

>>42638136
>The only issue I see for now is the lackluster ARM support for mainstream Linux distros like Mint on it.
windows arm also sucks
buying arm for general computing is a bad idea desu

>>42638136
https://fedoraproject.org/kde/download

fedora kde has arm support

lol, you might be in luck that the best just works distro has support for it. oh and also gentoo too as well

>>42638160
I don't like KDE.

>>42638187
so go with gnome or some other spin, lol?

>>42638191
I don't like Fedora.

>>42638229
go for armian or arch then
>I don't like them
gentoo
>no
slackware arm
>also no
LFS
>nope
rape yourself

>>42638234
I just want Linux Mint man. If I can't get it, I'll be using winshit 11 ltsc.

>>42638234
nta but gentoo would be the second best choice after fedora for an arm notebook

>>42638250
...if you want them to die after coompiling FireFox

>>42638253
emerge --getbinpkg www-client/
firefox-bin

also, by configuring make.conf one can make gentoo only install binaries instead of source packages if the user doesn't want to compile

>>42638260
>also, by configuring make.conf one can make gentoo only install binaries instead of source packages if the user doesn't want to compile
what's the point of running Genpoo then?

>>42638247
>linux mint
it's basically ubuntu (it's literally based on it) with cinnamon
and iirc ubuntu has arm support

cinnamon is a really downdated looking desktop environment
>>42638266
to create your own system, it's like linux from scratch but with a package manager

>>42638277
Ubuntu enforces snap which I don't like

>>42638277
>cinnamon is a really downdated looking desktop environment
Cinnamon is literally the best GTK-based environment available right now

>>42638282
ubuntu is garbage even without snaps, and systems using it as a base probably have a shaky future as they will eventually have to move to a debian base
>>42638287
qt is better imo

>>42638316
>qt is better imo
Cinnamon being a based GTK environment doesn't contradict Qt's better quality
t. KDEchad
>and systems using it as a base probably have a shaky future as they will eventually have to move to a debian base
Mint has a rather stable Debian base "in case Ubuntu dies" kek

>>42638247
atleast use win 10 why use win11?

>>42638277
also mint has no snaps and LMDE exists


I'm installing gentoo right now, although it is 30 past midnight in my timezone, much past my bedtime
>>>/g/106818514
emerge --ask sys-kernel/gentoo-kernel
just finished running finally

>>42638401
10 has bad touch support

>>42638401
win11 has winget. it actually has a useful powershell now. massive improvement for terminal users, even though I don't use windows but fedora kde instead
>>42638359
>Cinnamon being a based GTK environment doesn't contradict Qt's better quality
this lol

>>42638401
>ponyfag makes a thread on /g/
>there are 0 replies from the fags who make the rest of that board miserable to read
Interesting. Good luck with your install anon, maybe you'll be the blessed one who'll finally build the promised s/Linux/Lunix/g kernel.

page 10 bump

>>42638578
>s/Linux/Lunix/g kernel
I just used whichever kernel management method it said was recommended for beginners in the handbook

also yeah /g/ is so bad I lurked it for the first time in a while yesterday and there was thread flood of ebassi which was kinda funny I guess but the other threads weren't interesting to read anymore. I guess overall it's a good thing because I end up spending my time doing more useful things

>>42638466
doesn't win10 also have winget, I never used it so I'm not sure I'm just waiting for XFCE to compile right now

>>42637040 (OP)
Oh no, brainfuck poner

>>42639511
>doesn't win10 also have winget
Nope

anon with the collection of ponified software come back

>>42638401
neat, you're doing LTO, graphite?
also here: PORTAGE_NICENESS="19" so you don't fry your laptop.
and here FEATURES="candy news fixlafiles unmerge-orphans parallel-install parallel-fetch sandbox usersandbox network-sandbox ipc-sandbox pid-sandbox"
so portage isn't complete vomit and sandboxes a few things you'll have to install those sandbox packages though
Finally, please for the love of Luna, Celestia and mom get eix and setup some aliases for emerge --update @ world and flaggie / whatever it's called nowadays. also don't use grub, it's not worth it. Limine is comfy, systemdboot works too hell even efi stub the whole thing just no more crusty grub
Oh and if they haven't fixed that debian-kernel magic number bullshittery if you're using the dist kernel you'll be in for a hell of a ride.

>>42640216
>no mpv pony
...

>>42640425
Name: Spinwave or Violet Loop

(A play on “media player” and the spinning playback circle motif)

Color Palette (based on MPV logo):

Primary coat color: Deep violet / plum (#9b59b6-like tone)

Mane and tail: Gradient of dark purple magenta white, mimicking the MPV logo’s glowing circular gradient

Eye color: Bright lavender or light pink, echoing the highlight ring

Cutie mark: A stylized play button inside a soft circular glow, referencing the MPV logo directly but pony-ified (rounded, magical shimmer around it)

Species: Unicorn

Why? Because MPV is sleek, minimalistic, and powerful — a unicorn horn can symbolize precision and control (like the command line or advanced playback features).

Personality / Theme:

Analytical, calm, and a little mysterious — often lost in thought “buffering” ideas.

Loves harmony (pun intended) — always finding the perfect rhythm in life.

Known for tinkering with enchanted devices and experimenting with magical “filters.”

Voice has a slightly echoing, modulated quality (like audio reverb).

Design details:

Horn glows faintly in rhythm with a pulsing animation — like MPV’s visualizer.

Tail and mane have soft, blurred edges — slightly ethereal, like an aura or shader effect.

Accessories: Wears sleek silver headphones or a floating glowing ring that follows her (the MPV circle motif).

Aura magic looks like a gradient shader — when she casts spells, a circular pulse radiates outward.

>>42640485
sovl nuke I'm about to emerge and compile mpv literally right now

>>42640398
what does that mean ponyGOD
also I'm already using GRUB it only works in the backup fallback mode thing lol I couldn't get it to work the normal way

>>42640691
LTO is some GCC compiler optimization, same as graphite. It's mostly autism, so if you're with the default march=native and O2 useflags keep at it it's fine. though I recommend you don't use -march=native but -march=Ryzen or whatever the wiki says your processor needs. as native needs gcc to guess and that's extra work for nothing
Everything I sent you can put in your make.conf as is.
Portage niceness uses Unix nice to not run every compile job at full priority, so you can use your computer comfortably as it compiles stuff for you.
The features one is mostly quality of life and eye candy.

>>42640691
As for grub, try Limine. It's really straightforward. The Gentoo wiki is also a fantastic resource, use it as much as possible for everything. I'd say it's even better than the arch wiki as it actually explains things and doesn't adopt the idea that you already know everything and just need a refresher.

>>42640691
Finally here. https://wiki.gentoo.org/wiki/Portage_with_Git this will make portage faster to sync.
Sorry for spamming, it's been a few years since I used Gentoo.

>>42640713
>>42640718
>>42640720
thanks will try to do all this tommorow
goodnight thread

Bump

I will try niri now that I have xfce to fall back on and maybe a few other de/wms too

>yalter, creator of niri, no longer uses kirin for his pfp on github
:,(

>>42642079
oh thats the only reason I was going to try it lol

>>42642079
I wonder what was the reason for that change, whether some HR bullshit about looking unprofessional or something else.
>tfw biggest motivation to git gud is to become that guy who's decent in some tech niche and has "inappropriate" pony pictures as avatars
>>42642094
The Kirin (hot) is still barely there on the preview picture in readme.

>>42637520
i am using win7 ultimate right now. Modern hardware, SSD, all drivers working and
my god it's so fast, firefox plus ublock for browsing, and supermium.
It buries XP because 64bit support but all the video editing and crap that takes hours on win10 are done so faster. Security is ok as long the network is safe and all bad services are off.

>>42642370
why use win7 when linux and BSD exists

I generally used Linux on this X260, but my autism got the best of me and I wanted a computer that could be plugged into a CRT TV through SCART HDMI adapter and have custom scaling. It's a bit odd to use W7 on a 2016 laptop, but it's fully supported so why not challenge myself and see how far can the baby duck OS go 5+ years after Microshaft discontinued supporting it. It kind of reminds me of the time when I used to daily drive a Pentium 4 PC with Windows XP and AGP ATi card by mid 2016 (although it was more horrid, W7 on the other hand is perfectly usable and smooth as butter).

>>42643147
fair enough I only got one device I don't think of SOVL like that

>>42643157
I've got two, a T470 and X260. I was very reluctant about moving it to Windows 10 LTSC, but I guess it won't hurt to have a spare Windows PC at home.

It is a little strange to have full speed on a 50hz TV but I guess today's computers and the signal converter aren't as restricted as a video game console that was purposefully downclocked to be able to play properly.
My Trinitron has been acting up since the last year and it needs a slap to correct the picture from time to time. I also have a Daewoo which works fine but has some color bleeding (I think that's what they call it?) And thus I settled on this chink CRT called Electro Co. Couldn't get any proper details about it, but I think it's from mid 2000s, maybe even around 2007-2008? It lacks the remote, but I was able to access the menu by pushing two buttons together and I adjusted the colors to be better. The OS is barely readable but what matters are the movies and games (Win 10 was less readable, I guess its font rendering is worse).
I don't think there is a way to manually scale your picture on any Linux DE though. This is where Windows for a change wins for me, as it practically gets dealt with overscan issues.
The tv has a few scuffs on the plastic, I'll try wiping them down but I think I might put on some stickers instead. I'm glad the glass is in perfect condition. I got this TV for free about 6 years ago, it was meant for the garage but it was never configured.

>>42643147
>>42643555
>Kosz
but enough about your waifu

>>42643559
Trixie isn't for everyone and I completely get that. She's an asshole just like me. She sometimes regrets it just like me. She's cool for that.

>>42638030
sweet. which one? ive got the surface pro 7 and decked out in stickers (on the right)

>>42638466
>>42639734
>https://learn.microsoft.com/en-us/windows/package-manager/winget/
>The WinGet command line tool is only supported on Windows 10 version 1809 (build 17763) or later
and 1809 is a windows build from 7 years ago, which i doubt anyone is even still on, but the latest versions of 10 (21H2 & 22H2) should support it. The difference is that w11 comes with it already and on 10 you have to manually install it.
>>42640485
cute mare! i really like her mane ^:)
>>42642370
>>42643147
based. i know very few people still using windows 7, especially maining. Only time I ever mained 7 was on an asus laptop a few years ago before upgrading it to 10 with a 7 skin due to the laptop not officially supporting w7 (officially advertised win8). I wonder if that spike in windows 7 users seen on statcounter will have some momentum behind it or not. A 2% to 9% jump is two months is some significance given w7 has been EOS for almost 6 years.

>>42643562
NTA She has the cutest design in the show, that's it. I genuinely don't understand how anyone can like her. More power to you, if she makes you happy then I'm happy.

>>42644100
It's the LCD version of surface 11. Got massive edu discount thanks to polish gov. After installing 11 ltsc works like charm

>>42644240
very nice. i got mine off ebay for a decent price. i have 10 iot ltsc running on it like all my other systems, but the only downside of that is that the latest drivers for the surface 7 will only work on 22H2 and nothing else, so I cant use the official driver package

>>42640485
This is just Amethyst Star

Ante up!

>>42644689
hello fellow xfce god

>>42644689
blackjack spotted!

we're up baby!

>>42640691
I have a problem. I wanted to try KDE so I asked chatgpt like the retarded zoomer I am,
THEN I opened the docs while the command GPT told me is already running:

[7] default/linux/amd64/23.0/desktop/plasma (stable)
sudo eselect profile set 7
sudo emerge --ask --update --deep --newuse @world

but if I had read the wiki page I could've read the USE flags and changed the ones I care about before emerge-ing

it's so over ponybros

>>42648941
>dash cloud engineer
>clickops everything
>lazy as fuck
>asks celly to just throw more money towards marezure when stuff is slow
>"what's a kernel?"

>>42647441
>sudo emerge --ask --update --deep --newuse @world
crashed during sleep or something I woke up and the screen was black and caps lock and num lock unresponsive,
I pressed fn to see if it makes any beeps but no luck I even got their stupid app to prepare for it,
so then I hoped it would beep on start up but it didn't beep it just booted normally
then I tried logging in to kde wayland and x11 neither works fully however x11 is "useable" well for debugging atleast
i'm running the same command and just going to use my laptop so it doens't go to sleep
I also took out and put back in the ram once emerge is done running I will then let it sleep and wait a while and see what it does
sudo emerge @world

>>42644355
/g/fwt has the instruction to upgrade 10 IoT LTSC to 22H2.
I know because I told them.

>>42650805
Dear /mlp/pts/
I think I have been filtered but I don't want to install something else, I want to be a /g/entoopony

current problems:
KDE doesn't work luckily XFCE works just fine
I will try to use sddm instead of lightdm to see if that fixes anything
If that doesn't work I will recompile kde

>sudo emerge --ask --update --deep --newuse @world
this recompiled the linux kernel and didn't seem to fix kde so that's great I've been waiting 5 hours for it to complete

wpa_gui requires me to run:
wpa_supplicant -B -i wlp3s0 -c /etc/wpa_supplicant/wpa_supplicant.conf
each time I restart to be able to connect to wifi

>>42651551
I'm installing NixOS, it's so over
postan from live iso

>>42651837
posting from install
#filtered
I'll try again in mid december

>>42651953
Just install Mint.

>>42651999
that's bloated or something

nixos allows me to fresh install in like 20 minutes by saving one file to a USB and running one command post install and as long as I keep all data on NAS/USB drive I can move to a new device whenever I want

Install Debian

>>42655673
I really wish that some mainstream Linux distro would take its version naming after ponies.

I wish you could change your filepicker more easilly, nemo takes ages when you open big directory for the first time after boot

>>42657813
Is the directory on a HDD? Have you considered just hibernating your system?

>>42657813
>pic
I imagine Twi would have this kind of "special" mind that finds great joy in filling address books with as much info about her friends as possible. Imagine, none of those fields empty.

>>42657833
>Is the directory on a HDD
Yes
>Have you considered just hibernating your system?
No, I forgot hibernation existed, I only ever used sleep mode. I'll try it, but man, that is such a hacky workaround
>>42657880
Kind of unrelated but I wonder if ponies would stay in the innocent 70s-90s era where security was sending passwords over plaintext with unlimited login attempts, their emails and contact info publicly on their hoof servers (finger protocol).

>Of course we send data in plaintext, how else would we read them later silly
>Why do you keep talking about sniffing on wires? They don't smell any good

>>42659456
HDDs are a bothersome bunch considering that they go to bed whenever there's no activity (which is good for their health if they're not a CCTV drives!). I personally don't do anything about it, but if you must insist on 24/7 activity you're going to have to make some script that writes or reads from the drive every minute or so, so it stays up all the time.
I don't know how hibernation might actually affect the performance, but I believe that it might spin up the drive faster since the data on it is already somewhat cached as opposed to a full system reboot. I might be wrong on that however.
Good luck.

>>42659530
Oh I think there was misunderstanding, I have my /home/ directory on the same HDD I doubt it goes to sleep ever, I don't think that's the problem. My problem is that nemo, the cinnamon filepicker, takes ages to list contents of the directory only on the first time I open it after boot. Then it caches it somewhere non persistent and all subsequent openings of the directory are much faster until I reboot PC

>>42659456
If you HAVE to use a HDD, use an enterprise high capacity one.
Relatively speaking, they run so much faster.

>>42650871
thats possible? i thought iot ltsc was capped at 21h2

>>42659530
>they go to bed whenever there's no activity (which is good for their health
No, it's actually terrible for HDD longevity. Look it up, letting your hard disk run 24/7 will make it last longest. At the very least turn off spindown when idle (can be done with APM or EPC depending on the model).

>>42662003
My bad. I've never had the need to do so and I've been using 3.5 drives on USB adapters since 2012. My oldest drive has over 40k hours on it. Generally everything is nice as long as it's not Seagate.

>>42650805
Just use ufed for use flags, it's a lot easier.
>>42652009
Oh well, have fun with nyx, I know nothing about it short of the nyx file and flakes.

Speaking of Gentoo, I spent the anni installing it. Took about a day give or take. Only thing that sucks dicks and got me to go back to arch out of sheer frustration was signing the Nvidia kernel drivers for secure boot.
Holy shit, the Kernel's sign-tool just sucks so hard it's unreal.
I am fully blaming the gentoo dist-kernel for this as well as my dumb as shit UEFI for not noticing the exact same db key and MOK key when I used mokutil was used to sign to god damn Nvidia modules.

>>42662877
Based... wait... what'd ya watch the anni on?

>>42663831
Thinkpad
brand new

>>42663831
I have a tv and a pc, 'non. And mostly on Gentoo itself too.

i put a tube amp for my desktop speaker for the anni

FreeTube works again!

>>42662877
Okay I figured it out.
Holy shit this is retarded.
I was using the bin kernel. And the bin kernel is a special snowflake that doesn't share private keys with nobody.
Of course this is never pointed out why this is a problem.
And the linux kernel only recognizes the key it had at build time to sign modules.
So you could either rebuild the kernel with your own private key to sign out of tree modules with (Fuck you NVIDIA) OR you have to enroll a special snowflake key as a MOK key via mokutils like on Fedora. And of course you absolutely have to figure out how to make shims work to get your uefi to acknowledge you want to add a mok key.
Fucking circus.
Anyway if any of you want help with secure boot or Gentoo holy Fuck at this point I know the entire wiki of this damn os.

>>42664975
I switched to GrayJay already.

>>42665361
>proprietary license
automatically worse than (now fixed) barely functional Freetube

>>42660636
I use HDD because it's much cheaper per TB, enterprise HDD sounds expensive I might as well go for SSD at that point. But like I said, the speed is usually not the issue it's only the first opening that takes ages

RISEE MY THREAD

>>42665224
What is the point of enabling secure boot though I just skipped it when I installed Gentoo

>>42664886
idk anything about speakers but this is sovl

>>42655673
I might consider testing or sid ev&oe I want Trixie

>>42668149
Autism.
Also I like the idea of everything being signed with my own key for security reasons.

Lenovo Thinkpad L13 Gen3/i3-10110U/13"
£130

Lenovo T480S/i5-8250U/14"
£135

T480 i5-8350U £159

Lenovo X230/i5-3210M/DVD-RW/12"
£105

Lenovo X390/i5-8265U/13"
£115

Which one T480 T480S and X230 support libreboot, X230 is what rms used to use o algo and the L13g3 is newer + X390 wildcard
For a CS student (the greatest programmer that has ever lived)

>>42668241
x230t to draw ponies, (and for maths lessons in your course), and canoeboot (libreboot with less binary blobs)

>>42666244
Generally using HDD for OS is silly. 500GB SSDs are very cheap nowadays (albeit they're often without DRAM).

>>42668164
>Autism.
That's fair. I used to want secure boot too, but I never bothered, and now since I've got sour grapes it got me wondering about the actual security benefits.
It's mainly supposed to defend against evil maids, right? Yet at the same time UEFI generally allows you to enroll your own keys. Is there anything actually preventing an evil maid from signing a malicious keylogging kernel with her malicious key and switching the keys enrolled in UEFI to allow boot?

>>42668834
>Mainly protects against evil maid
Yup
>Is there anything preventing an evil maid from signing a malicious keylogging kernel
It prompts you out right, and you have to be in setup mode with zero key to completely compromise everything.
In short it's like installing graphene os. You have to allow the third party keys willingly and in a very obvious manner.

>>42668866
>It prompts you out right
Where though? I admit I've never tried it but I fail to imagine how it could do so in a secure way.
>grapheneOS
If you mean the "this device is booting a different OS" message, that happens for anything that's not Google-signed. This message is only useful if you're expecting to boot stock android and are surprsed with graphene instead (doesn't apply here since you're already booting linux and not stock windows), or if you verify the hash shown on every boot to make sure it's unmodified grapheneos. And the only reason it shows the hash in the first place is because google's key is hardcoded in the TPM, whereas the point of secureboot is to be able to change microsoft's key over to your own.

>>42668921
>Where though
You have to deliberately go to your UEFI bios settings, enable setup mode, and import the key for KEK, PK and DB. As for MOK, what I had to do here, you have to talk to the firmware directly and it will prompts you to import the key and enter a password for the key if it is password protected.
>Graphene
Oh no I didn't mean that message, meant graphene has to load its own signing key into the firmware (that's essentially what locking the bootloader does.

>>42669038
Therefore yes you could fool the user into importing garbage, but you can't just do it without the user giving proper consent.
I'd argue it'd be even easier to just exploit existing secure boot keys if they have leaked.
Like MSI's did a bit ago, you could just sign your stuff with their keys and the bios wouldn't bat an eye. Hence why it's important to remove all default keys and reinstate only the ones you need for your OS.

>>42669038
>You have to deliberately go to your UEFI bios settings
No no but my point is the evil maid will do all that.
Say your PC is out of your sight for a while while turned off. You come back to it and boot. Secureboot passes all checks and your kernel loads. You suspect there may or may not have been an evil maid operating while you were away, how do you check the kernel is actually untampered? Without going into your bios and re-checking which keys are enrolled there, because that's impractical to do on every single boot.

>>42669060
Oh, you mean if someone shows up and does all of that.
Well, let's see, for the sake of argument let's say your bios is not password protected so anyone can show up and mess with it.
First: they'd have to boot from a flash drive, add their keys to your bios, which is already hard short of erasing yours to add more, but let's say somehow they add a kernel module with their mom key here.
They'd have to go to your PC, boot a flash drive, somehow access your unencrypted root, add a kennel module to load, rebuild your whole kernel + initramfs, reboot, enroll their MOK ke manually typing a password, reboot to your os to check it works and finally reboot enable secure boot in the bios and shutdown before you show up.

It's pretty unlikely. And it's not something you can do remotely at all. Plus it assumes you don't have other projections: like password limiting your bios, or encrypting your OS.
Overall, to me it's pretty much not gonna happen due to how complicated it is to pull off.
Not to mention you can outright tell the kernel to not approve any random key via SIG_ENFORCE so even if UEFI says it's all good your kernel can and will tell the module to fuck off (exactly the issue I had with the binary dist kernel)

>>42669091
All in all, I'd say there's a way better solution to spy on you and it only costs you a wrench and a couple seconds of time.
Or you know, you can just tell the user: install my rootkit or else you're not playing vidya.
Or, just show up steal the laptop as it's unencrypted and run dreadpirtates style.

>>42669091
>Well, let's see, for the sake of argument let's say your bios is not password protected so anyone can show up and mess with it.
CMOS battery can reset the password on 99% of motherboards, to be fair.
>it's not something you can do remotely at all.
Yeah but again isn't that the point of secure boot? How would somebody remotely overwrite your kernel?

>>42669132
>CMOS can reset the password.
True, and it will reset your secure boot keys as well.
>How would someone remotely overwrite your kernel.
Without secureboot it's trivial. Just patch NTKERNEL.dll or have the user install some patch or something. Secure boot would prevent those patches to load without the user's explicit consent.
It's not a silver bullet by any means, but it makes any of those attacks dependant on you and you alone, which is a pretty good security measure.

>>42669144
Oh I see, so the point is to prevent accidentally installing (or infecting) your kernel, honestly that makes more sense. So my original impression was wrong and so was this anon >>42668866

If you have unwanted software overwriting your kernel files, though, I do think you have a pretty severe problem already, such as unwanted software being able to overwrite and most likely also read your user files too. But maybe it can help against specific attacks, or against carelesness when installing things.

>>42669165
Exactly! That's exactly the point.
>>42668866
But I am that anon, kek.

I guess I'll explain a little better. Because I'm noticing I am forgetting a bit. In my explanation.

What secure boot does is simply checking binary signatures when you boot. Nothing more nothing less.
Then implicitly it does other stuff:

But let's go over how the boot chain works:

UEFI > BOOTLOADER > KERNEL > USERSPACE

First your computer starts the UEFI then that looks for a bootloader. What secureboot does at this stage is verify the bootloader is signed by the keys you set.

Then the bootloader checks the kernel signature if it is setup to do so, 99% of the time, it is and leaves the rains to the kernel to do it's thing.
As an example this is how the apple tv first gen was hacked. The bootloader was checked by the firmware but itself the bootloader did not check for the kennel at all so you could just call a file boot.efi and it'd boot that without questions.

From that point the kernel then can choose to verify the modules it loads via SIG_ENFORCE so no unsigned modules you haven't personally approved load.

This whole chain makes sure everything is who they say they are, this is what I was pointing to when talking about evil maids.

Else, in its most basic state where only the bootloader and kernel are signed, all that happens is just tamper protection as you mentioned without any further security.
All in all, it's like TLS, verifying trust.

I hope this is a little clearer.

>>42668576
I think one person has bought up every X230T and is selling them for the price of TWO X230 with libreboot and coreboot installed, where else can I look, I tried ebay and cex but cex only has 12 and 13 inch X230 for £105 and £150 with the higher spec i5 and bigger screen. Where can I buy "ewaste" so I can get it for like 2digits price instead of getting scammed on ebay

I think I'l just deal with the system freezing and stick with my original plan to get a Framework in mid december and install gentoo then

>>42669233
Perhaps I'll setup secureboot this time too

>>42669471
Welcome to the autism, kek.
Else if you have a supported Thinkpad and you want to be mega autistic, you can try HEADS https://trmm.net/Heads/

>>42669505
I have L15g1 with some sort of issues that the hardware scan in BIOS didn't detect (it happened across different distros, maybe I should test a BSD? but I don't want to spend much time on fixing it and I can use it for the next 2 months after which I can get a framework and maybe a x230T

>>42669233
>this is what I was pointing to when talking about evil maids.
An evil maid is an attacker that has physical access to your PC, and can freely modify your data before you return. So my point regarding evil maids was, if the root of trust is in the key you've registered in the BIOS, the maid can just reset the BIOS, enroll her own key, and insert her self-signed malicious kernel into your boot partition, which then keylogs and exfiltrates your disk encryption key when you next boot. (Presumably the maid will also have cloned your encrypted data and can thus decrypt it all.)

ARM devices often have sealed or otherwise inaccessible TPMs, so overwriting the root of trust can be very difficult compared to a typical x86 mobo. For example on pixel phones, according to grapheneos,
>The verified boot public key flashed to the secure element can only be changed when the device is unlocked. Unlocking the device performs the same wiping of the secure element as a factory reset and prevents data from being recovered even if the SSD was cloned and your passphrase(s) are obtained because the encryption keys can no longer be derived anymore.
It's very rare for an x86 device to have a TPM that's this locked down or has this much control over the storage.

If there's no physical access, then the root of trust is mostly secure even on x86, but then an attacker has to have remote write access to your disk(s) in the first place to actually modify your kernel, and there are really not many attacks where someone can remotely insert a malicious kernel but can't simply exfiltrate your data. The only exception I can think of is if an attacker tricks you into specifically installing a malicious kernel module, so the attack wouldn't work for any other type of software, but if you're tricked then you'll just sign it with your secure boot key anyway.

>>42669667
>The maid can just reset the bios, enroll her own key and insert her self signed malicious kernel into your boot partition
I agree, yes technically this is possible, but unlikely. If your bios doesn't have any password protection. But, you can just swap your kernel and expect that to work.
The attacker would have no idea about what arguments your kernel needs to boot such as: where is your rootfs, what file system it is, what modules do you need to boot? Etc, they would have to have accessed your pc physically before that to already know all of this. And build an identical kernel, not to mention needing to use root utilities to make that possible.

>Arm devices have sealed or unaccessible TPMs
Not really no, I fully agree with what graphene is saying, but it is the same for x86. Putting your device into setup mode is the same as doing a fastboot OEM unlock in practice.
That being said, yes if they clone your SSD and that is not encrypted then they can build this kernel and boot it on your system with your keys and you'll be none the wiser.

I am not sure where you're getting at?
To me if you have physical access to anything then it's GG for the attacker anyway? Could your explain what your grievance is with the UEFI TPM system?

If cloning your SSD is a factor then you can make your system immutable. That way you're fully mimicking android's security system:
Locked down root of trust
Immutable system
The only missing thing is wiping the system upon changing the root of trust

>>42669735
Oh, one important thing, your keys DO NOT need to be on your system at all. Only the public keys, I think this is what you're having issues with, lel.
You should never store private keys on the system itself, of course then it's akin to leaving the keys on the door

>>42669735
I have two issues:
a) You didn't know what an evil maid was, it's a standard attack name and doesn't have anything to do with kernel modules.
b) Secure boot on x86 seems to me like it gives pretty minimal security. What's the threat model where something could overwrite your kernel or load a malicious module, but doesn't simply have root access?

>I agree, yes technically this is possible, but unlikely.
Evil maids are indeed a pretty unlikely attack, which is why most people get by just fine without caring about them. Even without secureboot, you need to prepare a working kernel for the system, and then swap it out in the boot sector of the drive. To do this, you already need to figure out the rootfs, the boot options etc., secureboot or no. Practically speaking you can get them from the initramfs; encrypted boot can protect against this.

>I fully agree with what graphene is saying, but it is the same for x86. Putting your device into setup mode is the same as doing a fastboot OEM unlock in practice.
I've never heard of an x86 computer overwriting encryption headers of attached storage devices when putting them into setup mode.
>Could your explain what your grievance is with the UEFI TPM system?
On ARM, the encryption headers are stored in the TPM itself as far as I understand it, and cannot be realistically extracted by a normal person. On x86, the encryption headers (e.g. LUKS) are stored on the storage disk, so the protection grapheneos uses is impossible.
>To me if you have physical access to anything then it's GG for the attacker anyway?
To a certain degree yes, but it's a matter of threat models. A normal evil maid attack is possible for a particularly tech savvy attacker without major resources; if your roommate is a wizard and can craft a malicious kernel, he can swap it out on your PC, and secureboot won't stop him as discussed above. But for example on a graphene phone, you need access to electron microscopes to open up the TPM and physically access the key data in there, there is simply no way your roommate could pull it off no matter how tech savvy he is; you'd need to be personally hunted by CIA agents with a LOT of funding.

>>42669743
No, that's not relevant at all, you must be misunderstanding me completely if you think I thought you'd need to store private keys anywhere. I'm assuming your private keys are staying completely secure.

>>42669818
>You didn't know what an evil maid was, it's a standard attack name and doesn't have anything to do with kernel modules.
I know what an evil maid attack is, anon. To me it's just not necessarily a physical access attack. The definition I always heard of it was mainly used in relation to tor, where it doesn't necessarily involve physical access.
That being said, your definition is absolutely correct, just stricter than what I thought, my apology..

>Secure boot on x86 seems to me like it gives pretty minimal security. What's the threat model where something could overwrite your kernel or load a malicious module, but doesn't simply have root access?
The main point behind it is to prevent tampering and giving root of trust. It is not a mean to deter attackers. Just like TLS ensures you don't get MITM'd secureboot aims to do the same.
I could direct you to googl.com and your browser wouldn't bat an eye if the cert I used was signed by an appropriate CA, same if secureboot was to load a binary signed with trusted keys.

My main point towards secureboot security is that, if you create private signing keys to sign binaries then enroll public keys inside the UEFI it becomes impossible for an attacker to boot unsigned code on your machine due to this chain of trust between UEFI, the bootloader and kernel, no matter how hard they try, unless they get your private key.
This however, as you mention, has its limits. Mainly the fact secureboot in itself does not wipe the rootfs if trust is lost as it simply does not have that capacity. And to me this lack of wipe is what distinguishes it, in practice. from what we see on android or iphones.

>Evil maids are indeed a pretty unlikely attack, which is why most people get by just fine without caring about them. Even without secureboot, you need to prepare a working kernel for the system, and then swap it out in the boot sector of the drive. To do this, you already need to figure out the rootfs, the boot options etc., secureboot or no. Practically speaking you can get them from the initramfs; encrypted boot can protect against this.
Indeed, secureboot, on its own is not enough, encryption is still necessary.

>I've never heard of an x86 computer overwriting encryption headers of attached storage devices when putting them into setup mode.
this is not doable without the kernel doing this work. UEFI cannot dictate this and it is up to the OS to do so. Basically your kernel should have an option to detect this change in secureboot keys. Or, like what windows does with bitlocker, prompt you to reenter your bitlocker key, not pin, on secureboot changes which is an acceptable compromise.

cont

>>42669818
>On ARM, the encryption headers are stored in the TPM itself as far as I understand it, and cannot be realistically extracted by a normal person. On x86, the encryption headers (e.g. LUKS) are stored on the storage disk, so the protection grapheneos uses is impossible
yes and no, most times yes LUKS headers are on the target device, but it is possible to relocate them to the TPM theoretically, I have no tried this for disclosure, but I don't see why not.

>To a certain degree yes, but it's a matter of threat models. A normal evil maid attack is possible for a particularly tech savvy attacker without major resources; if your roommate is a wizard and can craft a malicious kernel, he can swap it out on your PC, and secureboot won't stop him as discussed above. But for example on a graphene phone, you need access to electron microscopes to open up the TPM and physically access the key data in there, there is simply no way your roommate could pull it off no matter how tech savvy he is; you'd need to be personally hunted by CIA agents with a LOT of funding.
That's where I lose you, anon. I don't understand how your savvy roomie would be capable of doing this with secureboot. As, again, the private key to sign modules is secure.

>No, that's not relevant at all, you must be misunderstanding me completely if you think I thought you'd need to store private keys anywhere. I'm assuming your private keys are staying completely secure.
Cool, I just wanted to make sure, kek. The private key should always stay secure

>>42669961
>yes and no, most times yes LUKS headers are on the target device, but it is possible to relocate them to the TPM theoretically, I have no tried this for disclosure, but I don't see why not.
NVM I am full of shit, kek. I am doing just that with systemd-cryptenroll, disregard my previous post I suck cocks haha.

Does this matter if I have a maid computer (always in my pocket, no one except me will ever have physical access.)

>>42670066
It does unless it's a pony maid.

ofc I am a pony maid, can I go without setting up secure boot? it seems like a lot of effort compared to just making sure I don't give anyone physical access to my device

>>42670095
You don't need secure boot then, you already have pony boot, the best kinda boot

>>42669941
>To me it's just not necessarily a physical access attack. The definition I always heard of it was mainly used in relation to tor, where it doesn't necessarily involve physical access.
I had never heard of it as a non-physical attack, but if it's actually used that way then fair enough, I may be the retard here.

>The main point behind it is to prevent tampering and giving root of trust. It is not a mean to deter attackers.
I don't get what you mean here. Isn't "preventing tampering" literally deterring attackers? If there are no attackers, what's the point of root of trust?
The point of TLS for example is very much to deter attackers. It has mechanisms to try to avoid rogue CAs (high barrier to entry, immediate expulsion on malicious behaviour), which are imperfect but generally good enough that it has worked out in practice so far.

>if you create private signing keys to sign binaries then enroll public keys inside the UEFI it becomes impossible for an attacker to boot unsigned code on your machine due to this chain of trust
I agree but my point is that
>if the attacker is remote
If he wanted to boot code, he'd have to run code first, but if he can run code on your machine why would he care about booting?
>if the attacker has physical access
This goes into the next point:
>That's where I lose you, anon. I don't understand how your savvy roomie would be capable of doing this with secureboot. As, again, the private key to sign modules is secure.
If the roommate has prepared a malicious kernel that would work for without secure boot, the steps to bypass secure boot would be
>reset CMOS if the bios is password-protected (not necessary otherwise)
>generate a keypair
>sign the malicious kernel with his own keypair
>enroll the malicious key into secure boot
>proceed with the normal attack, i.e. replacing the kernel in the boot sector, as he would have if secure boot wasn't there
Or actually I'm retarded, for that matter, he could just disable secure boot in the BIOS rather than changing the keys enrolled.

You would notice next time you upgrade the kernel and signing no longer works, but by that time your rommate has accessed all your bitcoin wallets and set up blackmail with your pony porn.

>>42669989
This is cool though, I hadn't realised this was an option. Is there a way to set this up to wipe the headers if the secureboot master key is changed or secure boot is turned off? That would actually make it pretty damn secure and would make secureboot worthwhile in my view

>>42668576
>canoeboot
trannyware + canoe doesn't support anything past core 2 duo machines

>>42670263
isn't it made by the same person as libreboot though (I believe the same person also makes vimuser.TLD i dont remember the tld) or are you saying I should use coreboot instead or that my boot doesn't matter and I should just get the best laptop/£

>>42670263
https://vimuser.org/
>I am the founder and lead developer of the libreboot and Canoeboot projects

>>42670389
>>42670400
Not him but libreboot is trannyware. It's a basement autistic type of tranny rather than a grooming activist type of tranny though so you do you.

>>42670389
>>42670400
qrd
>Francis Rowe is a tranny
>hosted his castration operation on vimuser.org
>etc etc etc
>libreboot
>rms-certified bootloader
>Rowe made money by selling librebooted systems
>can't sell many because good luck finding laptops from 2008 lmao
>also in debt because of his castration
>so needs cash
>"minimized ME isn't a problem, m'kay?"
>creates osboot, libreboot with blobs
>later just changes libreboot to osboot
>sells modern laptops with his coreboot >>>distro
>Stallman and FSF stop recommending libreboot because blobs
>Francis makes canoeboot just to satisfy Stallman-pilled users
>>42670416
>It's a basement autistic type of tranny rather than a grooming activist type of tranny though so you do you
he supports Drew DeVault, nuff said

>>42670427
I thought Drew "one of the greatest programmers that ever lived" DeVault was a /g/entooman doe

>>42670433
>/g/entooman
he's /u/ and reddit

I will almost certainly be getting a Framework , two months from now

>>42670439
>he's /u/ and reddit
as in the Yuri board?

>>42668241
The T480 is a solid machine. I think the S model might be shitty but I forgot for the T480 specifically, look it up.
L series looks gay, thinkpads can be pretty mediocre outside of the mainline business series.
X230 is IMO too old nowadays for serious programming, you can do better especially for the price. It's not quite unusable yet, especially for a student who won't be working on massive codebases, but I don't recommend it. You'll notice the slowness even on just modern websites.
I don't know anything about X390.

I recommend also checking benchmarks for the processors to see if there are big differences, e.g. I'm not sure how the L13G3 compares to the T480. If it's significant you might consider going for it, better CPU is nice for programming stuff.
Also check RAM, especially on newer models it's soldered so if you get an 8GB model you're stuck with that. On the T480 IIRC it's unsoldered (but look it up to double-check), so you can upgrade it, but if you need to do that factor an extra £30 or so into your price. Also factor SSD upgrades if needed/wanted (less than 1TB is annoying, though you could get by if it comes with a 500GB one and you really wanna penny-pinch).

>>42670442
>as in the Yuri board?
yeah, he had it bookmarked

>>42670427
>he supports Drew DeVault
Fair enough, I haven't kept up on the latest tranny drama
>creates osboot, libreboot with blobs
The fuck, isn't that just coreboot then? The entire point of his project was to deblob coreboot

>>42670433
No his entire personality is simping and white knighting for trannies and "marginalized folx".
>>42670442
>I will almost certainly be getting a Framework
If it's because of the current drama, hold off and wait a bit and see whether they hold steady, there's a very non-zero chance they'll give in to pressure and backtrack on this, bending the knee to the trannies.
In general framework is somewhat mediocre in my opinion, for the price you're very likely better off buying an almost-new thinkpad (T or P series). Only reason to buy one would be if they continue being based.

>>42670453
>The fuck, isn't that just coreboot then? The entire point of his project was to deblob coreboot
He positions osboot/modern libreboot as "distro" of coreboot that's "easier to install"
(tbf he minimizes blobs, but anyone can use me_cleaner to do it themselves on corebooted systems)

>>42670453
>If it's because of the current drama
no it's not because they contain hitler particles it's because I can get a significant discount on it, I was only considering the ThinkPads because of LibreBoot/CanoeBoot and they are rather cheap, but I never thought if there was a use case for librebooting? What is the use case except my bios is libre now?

>>42670504
I don't want to spend a lot of money so the significant discount on the framework makes it the highest value compared to any other device even used thinkpads, is there really no use case for libreboot it's the only reason I wanted T480/X230T apart from the fact I have a thinkpad currently

>>42670234
>Isn't "preventing tampering" literally deterring an attack
The point of secure boot alone is to verify that you're booting untampered binaries. That is all.
I realized I confused it / overreached it a bit by meddling it with measure boot. Which itself aims to prove to you you are indeed running the correct os drivers and possibly rootfs.
My apologies.
>The point of TLS
At its core TLS is just so signing and proving the website you are visiting is who they say they are not so much security, don't get me wrong the encryption is nice to avoid MITM, but it doesn't negate it.
For example if the website you are visiting is not terminating TLS then you're at the mercy of whoever is in between no matter if your browser shows a padlock.
This is exactly what cloudflare does, you connect to 4chan.org, but in fact you're going to cloudflare first and 4chan is forwarded to you.
Avoiding rogue CAs doesn't happen, the only way to do so is if the powers that be agreed on who they put in their cert bundles to the os.

>If the attacker can run code remote
Maybe your computer wipes itself on reboot, or you'd want higher privileges, that could be a factor to ensure the exploit is always present and harder to detect, see logofail for example, the os has no clue a out it as it's affecting the UEFI directly.

>Roommate
If your roommate resets your bios, that will wipe your keys. Then I agree with the strict secure boot definition of checking the binary that is booted this is an issue.

So, i apologize mixed secure boot and measured boot, I am using both.
On short: secure boot just checks identifies that is is. Measured boot enforces it.

>>42670551
>This is exactly what cloudflare does, you connect to 4chan.org, but in fact you're going to cloudflare first and 4chan is forwarded to you
So my paranoid beliefs that I get when I see threads that seem targeted towards me, it is actually possible for a glowpony to go to CloudFlare and actually make a thread that only me and they can see (and archives I guess?).

>>42670551
Now, for measured boot.
You remember that chain of trust mentioned of UEFI, bootloader and kernel?

This is exactly that. Each link needs to verify the other.
So, basically if your roommate makes a kernel and your UEFI boots it, that's cool and all, but your bootloader will not let it slide and won't boot it.
So the only way for your roommate to fuck with your PC would be to:

A) extract the drive and DD your porn
B) boot and mount your roots using their own bootloader + kernel.

And in this very precise situation it's GG for you.
Hence why you do need to utilize encryption and possibly tie the encryption key to your tpm chip that way he only can abuse your files if he is on your machine and didn't reset the tpm keys (which will make secure boot throw a fit in the first place so headache)
And finally, you can add a pin (password) to decrypt that tpm key in the first place.

In short: if you want to be le ebin haxxor known as 4chan

- use secure boot to verify your system
- use measured boot so everyone in the bootchaon verifies they're who they are (possibly make your roots immutable (read-only) with the kernek checking it's signature that way you can even fuck with tahy outside of /home/anon
- encrypt root with tpm LUKS + pin via systemd-cryptenroll or similar

And you're pretty much air tight.
This is exactly what android does security wise for its boot process.
- verified chain of trust
- immutable root

>>42670558
Technically, but I doubt they care to place that phone call.
In any case now you know to not super trust TLS as the be all and all.
It's just a way to prevent unauthorized MITM, checking that you're talking to the right person on the end of the wire and encrypt your traffic between you and who you're connected to.

I encourage you to check the cert of the website you're connected to or use shodan.io to tell you more about who you're talking to when you go to preenhub.gov

>>42668799
I do have SSD for /
Only /home/ is on HDD. One of the reasons being that lot of third party software doesn't ask and just dumps shitton of program data into ~/ and moving it later is huge pain

windows 10 is kil
ltsc chads rise up

>>42670607
It's only a placebo.

>>42670607
inb4 slop like BF6 instantly drops support for W10 to implement better spyware AC

>>42666244
i purchased refurbished enterprise hdds from goharddrive and theyve been holding up pretty well. 2x 6tb WD/HGST drives for about $160. I keep them on 24/7 as part of my raid1 server, but i also make sure to keep a full backup of the server on an external drive as well in case of any malfunctions.

>>42665361
>GrayJay

>>42666244
i purchased refurbished enterprise hdds from goharddrive and theyve been holding up pretty well. 2x 6tb WD/HGST drives for about $160. I keep them on 24/7 as part of my raid1 server, but i also make sure to keep a full backup of the server on an external drive as well in case of any malfunctions.

>>42665361
>GrayJay

>>42670630
even when windows 7 lost support in 2020 it took at least 2-4 years for most major programs to kick win7 out of support (at least from my experience)

>>42670671
modern slop gaems devs make ACs so draconian they'll definitely drop support for systems with no TPM

Also posted this in the "dumbass thread" on /g/ but it seems nobody has an answer there, so hopefully people here can give me some pointers:

>Because I can no longer gen NSFW on Civitai I'm looking for cloud GPU hosts that rent out by the hour/day (or at least not per month) and allow crypto without KYC. I know about Vast.ai but it uses Coinbase or Crypto.com and both require extensive KYC. If there is no cloud GPU host that allows KYC-less crypto I'm okay with tactics to fool the KYC check as well.

>Don't bother with "you're a pedo for wanting privacy" comments, I won't even dignify them with a response. The fact I was fine with genning on a filtered service like Civitai should say enough.

>>42670729
I can't really recommend anything better than just getting into local gen. It can be done cheap if you get a laptop with RTX 4000 series.

>>42670729
local gen + please ask AI questions in AI thread, they have better experience with this shit

>>42670729
Me :)
I'll gen poni pics for ya. no seriously, just don't expect it to be that fast

>>42670994
Not him but can you make rainbow dash and applejack, with rainbow dash doing some sort of speed training workout, and applejack doing training for bucking or if eqg is easier then rd on a treadmill/ race track and aj doing barbell squats.
I want to make something similar to /pts/ but instead of /g/ it's /fit/

>>42671007
I guess, I'll see tomorrow. I'll try not to pollute this thread too hard, but gimme a prompt / prompts on ponepaste

>>42671024
Something I can use for the OP image >>42671131
I have 2 more thread theme's ready but I'm struggling to figure out what I actually want to find so my OP image's aren't planned ahead

>>42670729
>I know about Vast.ai but it uses Coinbase or Crypto.com and both require extensive KYC
You mean it requires an actual Coinbase account, or what? I've bought a 4chan pass with crypto before, and 4chan's Coinbase integration had an option where it would generate an address and you just send coins there from whatever source you like

>>42670662
>2x 6tb WD/HGST drives for about $160
160 each or both? Anyway goharddrive looks really great. I'll certainly pay it a visit when my storage budget replenishes

>>42671812
160 for both. i know serverpartdeals is another option as well

my thinkpad t420 keeps freezing on me, it will disconnect from wifi and the caps lock button will start flashing, i don't think it's a battery problem because it even happens when i use the laptop without the battery and just plugged in
ive read online maybe the ram got knocked loose or something, maybe i just gotta add cpu paste or something? should i check my cpu somehow kek

>>42671518
I read something to that effect on old archived posts on /g/ but I think someone mentioned this doesn't work anymore. Honestly though I have no idea, I have absolutely zero experience with crypto. I thought it'd be relatively simple to buy some with my bank account and then exchange it a few times to different crypto to anonymize it but it seems that's pretty much impossible nowadays because of anti-money laundering laws. I see people talking about meeting up with actual real people to exchange cash for crypto which is absurd.

>>42672605
I mean you CAN
>a) buy crypto with your bank account (and passport, usually)
>b) exchange it to monero (optionally on a fiat-less exchange, if cash-operating exchanges are restricted in your cunt)
>c) move the monero around a bit and also split it into non-exact amounts (never just send back the exact amount you withdrew from an exchange, that makes it super obvious even if nobody can see the underlying transactions)
>d) move it back to the fiat-less exchange, switch it back to bitcoin or whatever, and pay
The thing is doing it truly anonymously doesn't really make much sense unless you also meticulously hide your IP, so everything from step c) onwards would have to be done through tor, and in particular if you buy a 4chan pass then using it would also need to be done over tor or at the very least a vpn else you're not anonymous at all.
And you might possibly also need a story for what you did with the crypto you bought, in case anyone ever comes asking, though for the $30-60 a pass costs it's very unlikely anyone ever will. "I lost it officer" (e.g. forgot wallet passphrase) is as good an excuse as any for the majority of cases.

I haven't checked in a while whether 4chan actually accepts normal crypto transfers, but generally just buying bitcoin on a KYC-ed exchanged, transferring it to your own wallet for a modicum of separation, and then sending that to 4chan should be good enough in most cases.

>>42673635
I'm not doing anything illegal, so I'm not worried about 100% anonymity. I just don't want my pony porn linked to my identity, which would destroy my life regardless of it being legal if normies ever found out (which is very easily possible since places like Coinbase already have leaked KYC data before). So I'm not worrying about the FBI tracing down all my transactions, I'm only worried about the place I'm buying stuff from knowing who the person behind the transaction is. So if I do the following, should I be "safe" (assume I'm using Tor for all exchanges)?

1) Buy BTC with bank account on a site like RoboSats
2) Switch this BTC to XMR
3) Shuffle it around a bit with some tumbler or whatever it's called
4) Use the resulting XMR to purchase what I want (or send it to a different BTC wallet for places that don't accept XMR)

(Still don't understand how that would enable me to pay on a site like Coinbase but at least I'd be able to get a VPN/VPS.)

>>42673944
At this point you should commission an artist, anon.
Better wank material too.
Bonus if they take BTC or XMR if you're scared about the artist telling everyone that you're a degenerate. It's your word against them after all.

>>42674122
Nah, I like semi-realistic coom and only the top 1% of MLP artists ever even came close to what I'm able to make with AI.

>>42673944
>>42674244
just don't coom doe

>>42674422
Okay Visa. Have fun being irrelevant.

>>42674545
Visa? what does that mean

>>42674565
Visa Inc., founded in 1958, is an American multinational payment card services corporation headquartered in San Francisco, California. It facilitates electronic funds transfers throughout the world, most commonly through Visa-branded credit cards, debit cards and prepaid cards.

>>42674611
Recently got rid of my Visa creditcard too (never really needed it except to identify to Gaypal ages ago). Still have a Mastercard debit card from my bank but can't really get around that and I'm only using that for "society approved" purchases anyway.

>>42674611
>>42674658
Oh that makes sense lol because of the recent steam thing

Is it possible to live with no bank account (and no phone number)?

>>42674755
Maybe, but it's probably a huge hassle.

It's better to have two (or more) identities: one for legit, socially accepted stuff, and another for things you prefer others not to know. Privacy tards who do everything as anonymously as possible paint themselves with a huge red cross, even when they probably don't even really have anything they need to hide (i.e. is against the law) they're just being too autistic about it.

>>42674773
I feel like the no phone thing I can get away with just by saying, it hasn't contributed anything positive to my life in the last howevermany years I've had one but if I use something like XMPP or email no one would message me but when I think about it they don't now either so there is literally no negatives to not having a phone only positives (more privacy and presence and less anxiety)

>>42674788
I owned a dumbphone until about last year. Need a smartphone for some stuff like secure login for my country's healthcare and stuff like that. Could still be using 2FA with SMS but that's not secure so in a quest for "privacy" I would be compromising on security where it is needed most (and any leak of information has WAY worse consequences than a company knowing what messages I sent to my mom).

>>42670606
Someone write a fanfic on one of these.

>>42674832
If your country supports standard protocols, you don't need a phone for them. Time-based authenticators use TOTP and e.g. Keepass supports it (and can run on computers).
>inb4 computer isn't a second factor
It is actually, the second factor is "something you have" (vs. "something you know" like your password), and in this case what you have is the keepass database file - this means that if your password leaks or is discovered, people still can't log in unless they also steal the file. It's the same logic as logging into a website on your phone and using the phone for the 2FA code. It's only slightly less secure because you're responsible for keeping the file safe, while on a phone the app typically locks it down quite well since phones are so restricted.
Another less common protocol is U2F which is the one that requires a hardware dongle, it's really quite secure and very cool and also doesn't require a phone.

The only reason to use a phone for 2FA is when the website requires a special app and does not work with any open standards, but that's 100% a website security issue and I refuse to use most of them. Except steam because I'm a good goy I guess.

>>42674755
>no phone number
Basically depends on your social circle. If you only talk to people online you don't need a phone number. If you can convince your friends and family to be autistic you could use a smartphone with XMPP or matrix or something. If not though you might need a number.
>no smartphone
Now this is usually more realistic, though also hard to get perfect; there's usually 1-2 things that are very annoying without a smartphone. For me it was maps, so first I started using a SIM-less smartphone with OrganicMaps, but I live in a city so live traffic data from google is invaluable and it's the only big tech service I actually use on a regular basis.

>>42674869
>If your country supports standard protocols
Yeah, they don't. It's an app or SMS, with no choice for standard TOTPs. Shit sux.

>>42674882
IMO use SMS and sue them if there's a data leak, it's what I'd do at least.

>>42674888
You can't sue the government here, or at least it doesn't do anything useful. They always get away with anything, especially if it "was an accident". It's not America. More importantly, suing doesn't mean shit if my most important private info is on the dark web.

How long until inactive cock.li account's username can be re-registered I made a legendary username (@horsefucker.org) but re rolled the password in my password manager before saving it, so now I can't access it even though I made it and did the mining thing to activate email sending, I tried emailing the password reset from another cock.li account which was likely to have a similar fingerprint and I mentioned stuff in the privacy policy so admins could've checked db and saw it was me but I'm pretty sure the password reset email is never read assuming they even save emails sent to it. I can't believe that the email wasn't taken and I made it and lost it this way.

>>42674869
>steam
Steam and my banking app is the only apps from the app store on my phone the others I can't delete but I haven't played any games in months maybe a year now so I want to sell my account / inventory and then delete the steam app and maybe just get rid of my phone too

>>42675119
>How long until inactive cock.li account's username can be re-registered
What does the website say? I've never heard of them ever becoming available again.

>>42675134
Seriously? I guess I might as well post it:
pinkie@horsefucker.org
I was so suprised it wasn't taken

I think I asked this already a few threads ago in /pts/ but which drawing software do you guys recommend for drawing on Linux, I'm on KDE right now and I am really liking it but in the past I have been a GNOME noob so I have Krita installed now and I will try it soon. Also is there an alternative to Paint.net for simple stuff with a mouse not a drawing with a pen/drawing tablet. I have a drawing tablet and as I have mentioned far too many times I will likely install gentoo or LFS on a framework 12 before 2026

>>42675228
I tried the rest of the mane 6 and a few other mares and they were all taken

>>42675345
the closest alternative for paint.net ive found is pinta, but its not perfect and theres quite a few differences but again its the closest ive found.

>>42675345
Not sure what you mean by "simple stuff with a mouse", but for scribbling on images or adding text for shitposting purposes, I actually often use Inkscape, which is a vector program. Otherwise I use GIMP for everything. I think Krita might be better if you're using a tablet, but otherwise they seem pretty similar—I just stick with GIMP because it's what I'm used to.

>I will likely install gentoo or LFS
Based, do it filly. I've never used LFS as my main OS, but going through the process is great for getting some understanding of how a Linux system and a C/C++ toolchain are put together (assuming you actually read the book and don't just blindly copy-paste the commands)

>>42675345
I use kolourpaint for basic edits

Damn, I'm so pissed by "modern" developpement.
I want to install wan2GP on my server using podman, ok.
Feeling fancy, let's do it "from scratch".
Grabs a debian:latest and play with a temporary container.
And here come the shitshow...
Between the project that need conda that needs to be installed almost automatically, the python lib that are managed by the OS and not pip, the opencv not working...
Fuck, why do they feel like it's compulsory to build uppon layers and layers of shit!?
apt install git gcc nano, and you should be good to go.
Heck, even python, but don't use pile of software just for the kick of it...
The more the time goes, the more developers ship their own computer (as an image) instead of a program...
Thanks for listening to an old man's rant...

>>42675345
For vector drawing there's Inkscape.

>>42676641
Wow this is nice

>>42676925
>simple stuff with a mouse
Something where I want to make an image but not a drawing the specific thing I wanted to make was a desktop background with my timetable
>LFS
Will update mid/late December might just do it in /pts/ instead of making a /g/ thread like when I installed gentoo

>>42677186
>>42678209
I'll search up Inkscape and kolourpaint too

>>42677553
no that's mainly just python

>>42679500
More like bad developpers using python...
Also, C have the library and build chain hell (use cmake, no use conan, wait, was it scons?)
rustranies have the cargo that solves everything... except "obsolete" code from a year ago.
Have a look at "copy party" [https://github.com/9001/copyparty]
You just grab the .py, run it, whatever your python version, and you are good to go.
Only "standards library" from pip.
And any AI project is a fucking mess where every version of everything is fixed...
Seriously, if you develop like this, just give a Dockerfile already, it will be easier for everyone.
Sorry, I'm still grumpy at the state of current software...
It's 2025, these should be solved by now...

>>42680073
>More like bad developpers using python...
No, more like distributions shipping some python libraries, but not all and not always up to date, so some are only available from python's own package manager, but by default it's not well contained and can easily mix up installing things globally rather than per-project or even at least per-user. Running a package manager with sudo should never brick a system yet with python it's the worst thing you could do.
Venvs were then made as a shitty hack, I think they're finally kinda good nowadays but I remember the time when installing venv was a chicken and egg problem (I want venv globally because it's a system tool, but I can't install with sudo because my operating system will immediately spontaneously combust, etc.).

>C library
Awful for development but generally not too bad as an end user: OS packaging of C libraries is actually the standard way to distribute them, so there's no clusterfuck between OS libs and language package libs mismatching. C also supports static linking so if you're developing software using a less popular library that isn't often packaged you can just build it into your binary directly and end users are none the wiser.
>rust
Works mostly fine honestly, except when OSes try to re-create the python problem by packaging system versions, but rustup was actually built to work reliably unlike venv. I've never had a problem with old packages - if the software was written for old packages then cargo will download and build the old version just fine, and a dependency tree usually supports having multiple versions of the same package in it.
Actually I typically use uv for python venvs nowadays and it works quite well and it's written in rust of all things.

>copy party
Yeah if you develop with minimal dependencies then any language can be convenient to run, the average case does not consist of things like these though.

>>42680133
>The system manage the libraries
But why!? Go damn fucking why!? It's not the role of the OS packet manager to keep track of languages libraries!
Also, except for major versions, programs should work without problem with any versions of the libs (or at very least a version >= not == )
venv was an answer to stupid developers to lazy to fix their code to work correctly.
If you program needs a fixed peculiar version of a lib, and don't work with at least more recent, you shat somewhere (I don't take in account loss of functionality)
pip and requerments.txt should be enough for anything and everything. You don't need 4 version of a lib on your system, and you definitely don't need nor want your OS to manage it.

>C
It "kinda" works, until you want to do cross platforms.
Even between linux distro, it's sometime a mess and you need to recompile everything.
Also, windows looking near the exe for shared libraries is very good and for me, lacks in linux.

>rust
I will be honest, I never touched rust. I wanted to one day, and when I saw all the tutorials were lgbtqi+bbqxxl, I just threw the sponge.
I should give it an other go.
But the "obsolete package don't work anymore", I eared it on the internet and from a colleague.

>copy party
But that's the point, you ftp server shouldn't come with a mail reader or play chiptunes.
And, look at it, it supports a lot of protocols, and works on python2 and 3...
I do understand it's an extreme example, and AI is the other side.
But still, somewhere in the middle should exist...

>>42680230
>But why!? Go damn fucking why!?
If you mean python, I don't fucking know. I think that it came in the early era, like Perl, where people still thought the OS had to manage every package ever, since that was the norm with with C. Perl packages are also a fucking mess by the way, thankfully nobody uses perl nowadays. The OS itself also often integrates some python scripts for some reason or another and so requires some system python installation with at least some packages available for what it needs (also same with perl).

C is different because C was, from the start, THE OS language. The OS managed C libraries because the OS was built in C, managed the C compiler and the C standard library, and so it also managed the C packages for you and every other software. It worked (and still works) mostly fine as long as you stay in the same ecosystem. It's when the OS then tried to manage packages for other languages that had independent ecosystems outside of the OS that the shitshow started.

>Also, except for major versions, programs should work without problem with any versions of the libs
>If you program needs a fixed peculiar version of a lib, and don't work with at least more recent, you shat somewhere
Nigger you do realise that doesn't mean you shat somewhere, but the lib dev shat somewhere?
I agree that libs should normally be backwards compatible. But if a lib fucks this up and breaks compatibility, a program written for a previous version can't magically rewrite itself to adapt. Obviously you can update your program but if everyone always updated things to be compatible then there'd be no issue in the first place.
Semver itself is a relatively new concept, people used to version their shit willy-nilly. Even nowadays with semver some people are extremely lazy. Even people who try to follow it sometimes fuck up and break backwards compatibility unintentionally. And if the lib breaks things that's the lib's fault and there's nothing anyone else can do except pin the version and prevent it from updating.

>C "kinda" works, until you want to do cross platforms.
Yeah, that's another issue, but not really related to package management IMO.
>I will be honest, I never touched rust. I wanted to one day, and when I saw all the tutorials were lgbtqi+bbqxxl, I just threw the sponge.
Rust was initially built inside Mozilla which is tranny central. But by now it's been picked up by some tech corpos, and a lot of cryptobros, so it's possible to actually write rust without ever touching anything tranny-related.

>>>/g/106910089
Why they hating on Jeff Geerling?

>>42675345
Why install drawing software? If it's just for simple shitposts, you can edit images on Photopea without an account, which offers just as much as any slightly-older-but-still-usable Photoshop version. If anything, it's better than Gimp at least. I'm sure online drawing programs exist as well.

>>42680975
No idea who that even is, nor do I care, but nu-/g/ is pretty much /pol/ with tech, so it's probably something moronic like "He's a Jew/tranny/MAGAtard/whatever"

>>42682073
nat, but not even.
He may look/act a bit jewish (maybe he is, I don't know) but he just really like ARM and do lots of videos on it.
And granted, most of the videos are "X didn't work on raspberry pi, so I recompiled the linux kernel on my mac, and now it works! Barely."
But they can be as retarded as they want, at least Jeff puts his work as open source and helped quite some bit to bring pci-e to the raspberry pie.

>>42682294
Yeah but /g/ has gotten SO bad that it's filled with conspiratards (and trolls falseflagging as them) who will say shit like "He supports <doesn't matter what>, so he's <some schizo theory why this is somehow very bad>"

It's always sobering to step out of /mlp/ and needing a moment to adjust to the state of the site nowadays.

Don't die plz

>>42681973
i want to learn to draw ponies eventually and the other is for making images like a functional wallpaper layout not an artistic one, just to layout important details and stuff

>>42682073
I'm probably going to only lurk /mlp/ from now on thanks to /pts/ and /ss/

>>42684243
thanks for bumping when I was asleep

Does anyone know what makes Ubuntu and Fedora officially supported changes for Framework laptops. Can those changes be ported to Gentoo/LFS? Also is libreboot supported, + usecase for alternate boots or are they just a meme?

>>42685133
Possibly it's just that they've committed to supporting those distros if you report issues to FW, so if some bugs happen they might forward them to those distros/help triage and prepare fix/etc instead of telling you to do that yourself. It's just an x86 computer, thankfully you don't need to cook a dedicated OS image for each model. I'd just check TLP vs power-profiles-daemon, they recommend not using the former on Ryzen boards for some reason, and iirc that's about it. Firmware updates are delivered via LVFS mostly, and that works on most distros I guess. I should fix EFI capsule updates on my setup, I use efistub without a bootloader so I need to do UEFI upgrades manually from a pendrive, because LVFS doesn't expect that configuration.
I run Arch with KDE on both of mine, the only things I'd complain about and consider Framework related is screen brightness sensor not working in KDE (not implemented, Gnome has that) and some issues with inner screen sometimes turning off or freezing on Zen4 boards. I'm not sure if Ubuntu or Fedora are free from those issues, they both ship recent kernels and this is considered to be some bug in amdgpu driver. It's just that - support from manufacturer when things go wrong, vs support from community of random nerds. I'm okay with the latter, if I was in enterprise IT I'd prefer the former.
Coreboot/libreboot/canoeboot/IWTBIRD are not available, except for that 12 gen Chromebook variant, and some prototype for Zen4 someone did. I suppose openSIL on consumer Zen6 will make that more feasible.
>usecase for alternate boots or are they just a meme?
Well, there's certain degrees of proprietary software one can accept. I appreciate Stallman and the like for being the living examples of how things should be, but personally I reluctantly accept living with some blobs with current hardware.

>>42685184
>screen brightness sensor
Sorry, I meant environment brightness sensor used for automatically adjusting screen brightness. That's a cool feature, especially on macOS where it handles keyboard backlight too. But of course sometimes it's buggy there as well, if you sit next to a window and a bit of sunbutt lasers intermittently land on the sensor.
>pic rel

>>42685133
In the vast majority of cases, official support on laptops means
a) these are the distros that the manufacturer tested, and
b) these distros come with all necessary drivers out of the box.
What this means in practice is that other distros basically weren't tested by the manufacturer so they can't actually explicitly say "we know this works", and that things like drivers and firmware as the other anon said might be less convenient to obtain in another distro. It should almost always be installable unless there's some very exotic hardware and the drivers/firmware are either closed source or not packaged for any other distro and unusually difficult to repackage/build yourself.

>>42685133
>Can those changes be ported to Gentoo/LFS?
Most likely there are no changes. Framework doesn't own Ubuntu or Fedora and most likely would have a hard time convincing Ubuntu/Fedora devs to include Framework-specific patches. This is unlike System76/Pop OS, where System76 does own the distro and includes custom kernel patches to make their hardware work. (Fortunately it's pretty easy to take the kernel from one distro and use it with a different one.)

>>42685184
>>42685195
>>42685209
>>42685215
Thanks everypony

>OpenAG is develop by Yalter
>who also made niri wayland compositor
I do enjoy intersection of Russians, ponies and Half-Life

>OpenAG free implementation of Adrenaline Gamer, HL1DM mod is developed by Yalter
>who also made niri wayland compositor
I do enjoy intersection of Russians, ponies and Half-Life

>>42637044
Use vorapis v3

posting this through my t60p. my t420 has constant "kernel panics", and i spilled some water on it yesterday and now the mouse/left and right clicks wont work, even the trackpad's click function won't work
i plug in a usb mouse, and not even that, works
it can work, but i just have to rely on the keyboard, which obviously isn't ideal, no way i could do things like make music on it now
i think it finally shit the bed, after 3 years.
anyone else had to retire their thinkpad(s)?

>>42686095
My L15g1 will be retired soon sadly, it has problems with waking from sleep it either black screens or freezes. Although the BIOS full system scan said everything just werks I am curious if it might be a software issue but I doubt that becasuse I'm pretty sure it happened on different OSes at least on Gentoo XFCE and NixOS KDE Wayland I sometimes have to restart my PC and on Gentoo it would make some BIOS beeps I posted about on /g/tpg/ but no one responded and the documentation just says to download some app to record the beeps which don't happen anymore so I gave up and am just going to live with the problems until I get a framework 12 + stylus this December

>>42686095
>no way i could do things like make music on it now
https://vimeo.com/22798433

>>42686095
>i just have to rely on the keyboard, which obviously isn't ideal, no way i could do things like make music on it now
You don't even need a whole keyboard for that. There are trackers that will run on a gameboy, which only has like 12 buttons total
https://www.youtube.com/watch?v=fKeDSBr_W7I

>blogpostan
In case somepony else tends to forget, simple solutions for difficult problems can be good enough. I've spent a year of "ponified" K8s at home with no persistent logging because so many tools, platforms even to choose from. I've had Alloy, Loki, Minio and Grafana since March, which was supposed to be simpler and lighter than ELK, but Grafana started imploding every upgrade which quickly surpassed my ambitions to fix it properly, and Alloy was a mess to configure decently and didn't inspire confidence for this reason. I could migrate to their K8s monitoring chart, but I felt like it was getting too complicated for my home needs.
So now I'm on Fluent-bit and Garage (as S3 storage, to test instead of Minio). A little pony^W agent runs on each node, collects pod logs from files, adds basic extras like namespace and labels using tiny RBAC, then simply ships them as files to S3 every 10 minutes. The moment I'll actually need to check older logs, I'll just download JSONs from specific timeframe to my PC and brute-force through with jq and grep/ripgrep to find things in them. For some reason I used to feel like this wouldn't be sufficient, but today I've realized I don't actually look at older logs that often, I don't produce tons of logs, I don't really need indexing. This might give me just enough and require little maintenance in return.
Little ponies, little software solutions. Perhaps even the occasional simplicity of pony lives might become a subtle inspiration for hoomans.

>>42688548
Yep. Lotta people assume they have "big data" or need "big compute" when actually they don't. I really like this paper, which features several fancy big data systems getting owned by 20 lines of code running on the author's laptop: https://www.usenix.org/system/files/conference/hotos15/hotos15-paper-mcsherry.pdf

https://github.com/CHooverShrimp/fx-4chan
how do you think?
>inb4 dicksword

I also created an extra version for the cloudflare worker, but 4chan API blocks cloudflare and thus is useless.

Who want to give ear scratchies?

>>42690582
sovl

>>42690582
I want to punch her face.

>>42690692
why

>>>/g/106930795
lol

>>42691327
I hate everything Microsoft

>>42690582
Me.

>>42689468
Useful tool, has pony in readme, can't complain. Not for me as I don't link 4chins offsite, but I can see the value it provides. Thanks, anon.

>>42688615
>Lotta people assume they have "big data" or need "big compute" when actually they don't.
Uh, I use AI, so yes I do. I mean, I know a lot of you Linux users don't really do any work on your computer besides tinkering with the OS but this doesn't go for most people. Most people let their computer work for them, not the other way around.

I want to get back into Linux. I tried several distros years ago (started with Ubuntu but didn't like it, especially GNOME) but I stopped using them when one major bug wasn't fixed after literal years, namely the bootloader updating the UEFI boot code on EVERY update (thus greatly limiting the lifespan of the flash chip). Windows does this once, on install, and that's the only time it should happen since there is literally zero need to modify it every update (especially since there are no changes whatsoever). This was to me the biggest evidence that Linux will never be mature if such an easily fixable and serious bug remains for years because of kvetching on a bug tracker. Just fucking fix your shit, it's literally a day's work if even that.

Anyway, I'm digressing here, but I think Windows is better written than Linux (the core at least, not the shitty store apps and all that crap) but I just don't like their company's policies, and I like the idea of the source code being available of the OS I use so the people responsible can't sneak in shady anti-user shit. So my question is:

1) Has this braindead bug finally been fixed?
2) What is a good distro that
a) Does not use GNOME
b) Does not use Firefox, Chrome or Chromium by default (LibreWolf is acceptable)
c) Lets me install certain applications to their latest versions without having to wait for a new release of the entire distro (the biggest con of Linux IMO)

I know the last option can be done with application packages like Snap or whatever (I know there's several options and I'm sure one is the best but I don't care enough to worry about it) so it would need to support something like that at the very least for the browser, mail client and office suite (LibreOffice preferably).

>>42694924
>1) Has this braindead bug finally been fixed
idk

>2) What is a good distro that
>a) Does not use GNOME
>b) Does not use Firefox, Chrome or Chromium by default (LibreWolf is acceptable)
>c) Lets me install certain applications to their latest versions without having to wait for a new release of the entire distro (the biggest con of Linux IMO)

NixOS unstable if you want to configure everything in one file (just remove the line that uses firefox by default and add your browser of choice (ungoogled-chromium or palemoon or librewolf?) then run one command to apply all the changes (sudo nixos-rebuild switch)) also with flakes you can use specific versions of any software you want it has a KDE installation live iso,

Gentoo if you want to spend time on your OS (compiling and configuring everything to the smallest detail) requires a lot of time so make sure to have a spare device with everything you need to work in case you mess something up on gentoo

Bump.

>>42694924
What was doing it, since you say bootloader I'm assuming grub? You don't need that at all these days anymore (though a lot of distros still include it purely out of inertia). The kernel is itself a bootable EFI binary (assuming that feature is enabled, which it almost always is)

>>42694924
unironically arch linux, just use archinstall, get limine or systemdboot so you don't have to deal with grub's absolute wretchedness and years of crust.

>>42694924
>the bootloader updating the UEFI boot code on EVERY update (thus greatly limiting the lifespan of the flash chip)
Lol wut? How often are you updating your bootloader? You could update it once a week (which is already unrealistically often) and that flash chip would still outlive you

>2) What is a good distro that
>a) Does not use GNOME
>b) Does not use Firefox, Chrome or Chromium by default (LibreWolf is acceptable)
Why do you care what the distro defaults are? As long as the thing you want is packaged, you can always uninstall the default and install that instead.

>c) Lets me install certain applications to their latest versions without having to wait for a new release of the entire distro (the biggest con of Linux IMO)
Look for a "rolling release" distro

>>42691510
>I hate everything Microsoft
based

I am thinking of going smartphone and SIM free,
while using a spare smart or dumb phone as a "landline"
with a pay as you go sim so I don't even have to pay for anything
Then I can receive calls and messages for free (but not make any from that phone,well not for free at least)

>>42696859
Seems silly

>>42696926
what about it seems silly

>>42690692
But anon... she's so cute

>>42696941
post moar
is there an internet explorer pony

>>42696859
I've done that before. It's 100% viable if your social circle allows and if you don't need your smartphone for travelling (live traffic maps, live public transport updates, uber etc. are all smartphone-exclusive). The latter is what ended up killing it for me

>>42697658
nice, which apps specifically stopped you, I can't think of any must haves for me personally that I can't do on a laptop except run tracking unless anyone knows a way to track runs with a laptop?

>>42697716
>which apps specifically stopped you
Like I said, live traffic maps (specifically google maps, the only google product I voluntarily use) and uber for those rare occasions where you need it.
This is shit that is inherently cloud-based and there is zero alternative that doesn't involve a pozzed smartphone. If you ever need to take a cab, you either have a smartphone with uber and can do it, or you don't and you flat out can't, or have to hunt local cab agencies which can be a nightmare in foreign cities if there are even any around. Or you use public transport which again is vastly harder when you don't have google maps telling you directly where to go and a web browser in your palm for looking up "how to buy metro tickets in X city".
If I didn't travel outside of my city and didn't drive (and need live traffic) I'd probably get rid of my phone again.

Actually no, I'm not entirely correct, I still used to have a 10 year old phone (moto G from 2013), without a sim and mostly kept disconnected from the internet, which I used basically for Steam's auth codes. You can probably install some foss fitness/run tracking app on something like that if you want to. It'll also be usable as a GPS with offline maps (osmand or organicmaps), reader and music listener, though I still used a cheap kindle and a cheap chink $20 music player for those. If it's not connected to the internet it's probably fine unless you're hyper-paranoid about always on location tracking, in which case you can get some of those foss phones that offer hardware modem killswitches (e.g. pinephone pro is garbage as an actual full phone but it is, all things considered, a working pocket-sized linux PC with I think working waydroid for apps).

>>42697843
The thing with phones is they emit signals even when you don't have a sim I think it's called IMEI and the cell tower providers can use it to track the precise movements of that phone which could be linked to your house, workplace/university, relatives houses, place of worship, your GP / hospital / dentist, etc... sensitive stuff.

>>42697374
Yes and firefox pony and Google chrome pony, they've been discussed in the past 2 threads

>>42697994
Yeah, she cute.

>>42697994
whoops I think I asked the same question last time

>>42641240
Man, I remember the old Google Doodles they used to do like this in the '00s. The old sarif logo was so nice

>>42697994
Have more cute Firefoxpone

>>42697873
Yeah, smartphones are basically tracking devices. You have to completely gut any and all useful features they provide for them to be anything close to private.

>>42699927
should I keep a dumbphone with "faraday cage" (tinfoil) or just go with nothing like I plan now?

>>42699706
nice. do browser based crypto miners still exist?

>>42702123
If your dumbphone is pretty old (like mine) it's either already stopped working or is going to soon anyway (because the frequencies are "needed" for 5G) so might as well go phoneless. I never take a mobile phone with me when I go outside (I never leave for very far anyway) and the only time when it was a bit annoying was when my father had a traffic accident (not caused by him) and I couldn't take pictures of the damage or make a call. If you can do without, try to go without, or just keep one around in a container for emergencies.

If you are going to go very far away though, like hiking in the mountains, you should always take one with you, including a solar powered battery.

I hope one day a solar flare will disrupt all electronic devices. It will be great fun to see the zoomers completely panic because they can't use Google Maps to tell them where to go for help.

Since I am going to do mobile phone-less I was wondering if there was a way to track runs (pace and distance) using a laptop (Framework 12 specifically) if not does anyone have watch recommendations, I don't really like watches but then I could use it to track running pace, distance and I guess Heart Rate

>>42702792
I am also wondering, is there even a use case for tracking pace distance and HR, I could just roon

>>42702792
>>42703438
Do you want, like, detailed updates or just general daily stats? You can just record your start and end time, and roon a pre-planned route that you can measure on a map, that gives you pace and distance. You could put this into any spreadsheet, though I'm sure there are also FOSS tracking utilities for this.
But if you want detailed health metrics, continuous pace measurement etc., then a watch or fitness band are actually useful. They generally don't connect to the internet directly but rely on their proprietary app for that, so find one that's compatible with GadgetBridge and doesn't require an initial internet connection for setup.
I heard Garmin watches are pretty decent for that. But also kinda pricy, and they're pushing amoled really hard so it's annoying to specifically find their transflective displays which actually have decent battery life. An alternative would be a random chink fitness band that's approved by gadgetbridge

>>42703538
>Do you want, like, detailed updates or just general daily stats
I guess that's the question I need to think about, I wanted pace and distance because "it seems useful" and well everyone uses a run tracker but at the end of the day I don't need the data what I need is to roon and lift more. I'll keep the solar garmins in mind if I still decide to get one. Thanks anon

>>42703438
>>42703654
forgot to remove my name lol

>>42703654
Yeah you definitely don't NEED it, use common sense and just some manual recording with pen and paper to keep track of your progress
I think the main advantage of trackers is they give you your fitness data, like heart rate and breathing rate, so if you want to be hyper autistic you can minmax your fitness so you always roon right up close to your limit for max improvements but not so much that you do more harm than good. But I'm willing to bet that a large chunk of people using them don't bother and just use them as pretty stats.
The other big advantage of having stats is using them for motivation, number go up. This is very much up to you, for some people this might be the main thing helping them actually keep at it daily, for others it's completely unnecessary, if you're successful at just doing it without any extra dopamine memes then you don't need it

>>42704837
Ok maybe I'll get it as a gift for myself if I hit a goal or something in the next few years I'm a noob at rooning rn anyways

KDE Plasma is very nice

After several hours of messing with GPU drivers and xorg settings, I finally figured out the reason my second display wouldn't work was because my trusty RX 580 doesn't support simultaneous DVI and HDMI output.

>>42706180
>no art of Spike fucking Konqi
Sad!

>>42706923
>dragon bussy
no

>>42706561
Sounds like xorg bullshit, or you aren't using the amdgpu driver. From what i know both xorg and wayland rely on drm/kms and amdgpu exposes all display engines properly.

>>42706965
Standard configuration for that board is 3x DP outputs and 1x HDMI. If anon has one with DVI out as well, I wouldn't be surprised if the manufacturer basically just wired a single output from the GPU to both the DVI and HDMI ports. The two are similar enough (that's why you can get passive HDMI<->DVI adapters)

>>42706965
I've read something pretty similar to >>42707135 when I was trying to figure out how many ports I'll be able to use on my GTX750.
Except it was something about some ports sharing same clock and thus making it impossible to use some ports in some combinations. But I don't remember exactly, I just know that such limitations can exist

>>42707226
>tfw video signal clock mare
>on time, mostly
>constantly describes pictures to the other mare in your monitor, her best friend
>they both get really confused and dizzy when one speaks out of order

>>42706923
Be the change you want to see in the world

>>42707266
Konqi is a male...

I need more pony in my tech

I think of switching to a dumbphone and keeping a smartphone for banking shit at home, turned off most of the time, only on when needed (my bank doesn't have a web version so I'm fucked). Which dumbphone is known to have OPUS support? I need muh music.

>>42710476
You don't need a dumbphone

>>42710476
I guess the Nokia's dumb phones do?
But seriously, why do you want to switch to a dumb phone?
I don't get the point?
Security: you are now using sms which is absolute dogshit, same for calls.
If it's tracking, the cells towers are still logging your IMEI / IMSI so no real point.
If it's surveillance from Google / apple, then yeah fair, but you could get a pixel and flash graphene on there or remove everything google from lineageOS (mainly DNS, NTP and geoip) that'd at least allow you to do everything a dumbphone can do, better, and more securely since you can use e2ee messengers with that.

>>42710476
>I need muh music
Use a music player duh

>>42710476
Speaking of if you want a music player I think shanling still has the M0 around? Else hibys are good too if you want something bigger.

>>42710624
Or get a cheap smart watch like a bip6 that can connect to Bluetooth and play music too.
Plus it's a watch / fitness tracker and doesn't connect to the internet.

>>42710476
unironically this >>42710485
and this >>42703538
I'm going no phone + no sim + eventually getting a solar garmin or similar (some of them support music too which you wanted, or just don't listen to music)
You can use SimpleX chat it works desktop only (unlike Signal) but normie contacts can still use the app on their phone

Speaking of lossless music
How do I convert all music files I have to a more compressed form (mostly FLACs) I have only a few hundred songs saved but it's like 10GB they are different file types not all flacs if not for that it would probably be one command in le terminal

>>42710625
>bip6
Interesting, does this require the Zepp app? It's a lot cheaper than the Garmins although worse battery life it has the 1 key feature I want, Built in GPS (that works without a mobile phone (smart or otherwise)) so I can track roon pace and distance and time, and I guess heart rate too

>>42710736
How many different types can there possibly be? It may not be one command but maybe 4-5 commands should to it.
But also I wonder if ffmpeg can just intelligently transcode from any input to any output. If it can then you'd just specify the encoder options and use one command for everything

>>42710736
>how do I convert all music I have to a more compressed form
ffmpeg -i Yourtrack.flac newtrack.opus you can change the bitrate with -b too if you want.
Or you can write a quick python script / ask ai to use bash + glob to find everything that looks like a music file (.mp3, .flac, .mlp whatever) and create a copy of your tree with all files converted to opus, python works well for that too, essentially automating that ffmpeg command i gave).
Else foobar as a pretty easy to use right click menu so you can point to your whole library and just convert it to opus it needs the free media encoder pack too.
>bip6
Nope you can use it perfectly fine with gadget bridge. Plus you can customize the watch face to whatever you want using their website for that and upload it to the watch without internet.

>>42710760
>>42710773
thank you anons

>>42708421
>implying this will change anything to some anons

is KDE Linux a useable distro or just for testing KDE?

>>42711455
Last time I checked, they meant this to be a distribution for end users rather than developers, but it's still a fresh project. I'd stick to Kubuntu, at least I know what bugs I can expect.
Speaking of desperately trying to plan a way forward for computers of relatives after end of free Windows 10 security updates, ponies would never ask you to log into cloud account to use your computer or keep it secure. If they did, they'd keep it optional. 56k modems are still a rarity, not everypony can afford such state of the art technological marvels yet!
>sneakernet remains most common way of exchanging computer data in Equestria
>ponies meeting up regularly to trade floppies with funny gifs and tiny games

>>42710476
>my bank doesn't have a web version so I'm fucked
get a better bank (I did)

>>42711455
probably just KDE showcase
go for Kubuntu or Fedora KDE

currently debating bidding on another auction lot of pcs from my local uni. did it back in february and made a decent profit, but ive been trying to resell macbooks more recently and its been harder than before. now with windows 10 being EOL, will people still care as much?

>>42713675
Well, precisely because of the end of support of windows10, and the shitshow of microsoft, you may have a shot at selling the laptops with a working mint (or anything normie accessible).

>>42714722
Based. Dibs on making the thinklight her horn.

tech bump to kill anni thread

>>42713675
>>42714722
cute pics

>>42714722
oh thats funny thats the same thinkpad model i had in the picture i posted. shame i dont have it anymore, had to sell it a few months back

>>42714327
i might just give it a shot again since i dont actually know the ratio of nerds to normies who bought from me in the past. I always preloaded them with win10 pro which might not be as desirable now, so ill have to think of something

>>42716090
hackintosh ^:)

>>42716090
Well, why not give the choice?
By default mint (or any normie linux, idk the current mainstream flavor), and if they want, for few more bucks, windows 10 pro whatever?
That being said, you can also use a "debloat" windows11, but idk how it will handle updates.

>>42716296
I primarily sell through facebook marketplace, so I'll have to see if im able to include options on posts. I'd consider eBay but they take a portion of the money and I dont want to deal with shipping.

Does anonfillyx/anonfillytwitter guy show up here? His embed fixer doesn't work when elon's aislop army decides an image needs to be restricted.

>>42718263
looks cool but a waste of zip ties

>>42714722
based and checked or something

Say I wanted to host a tiny website to be used by a subgroup of nor/mlp/eople. I'm looking for ways to do that as anonymously as possible, so naturally I started checking out free hosting services and free domain name providers. Any suggestions? Or should I just learn to use crypto.

>>42719190
Anonymous from who? There's different levels of effort required depending on whether you're hiding from other anons, from Hasbro's lawyers, from law enforcement, or from the NSA. If you just want to make sure anons can't dox you, host it on Neocities or Github Pages, or get a domain from a registrar that provides whois privacy.

>>42719190
You can also host it on github or gitlab pages under the github or gitlab domain, without needing your own domain name.
This is only if it's static of course, for a dynamic website you'd need actual web hosting. In that case like the other anon said it depends on what level of anonymity you want, find a VPS that takes crypto payments, even better if it's accessible over Tor and takes XMR payments. Check out kycnot.me

You can also just host an onion site or eepsite at home, which is free and guaranteed anonymous and you can run any kind of dynamic webserver you want, but people will need to use a browser configured for tor or i2p respectively to connect. Also you need some server you can run 24/7 (could just be you not turning off your PC, or running it on an old laptop, but either way you'd be responsible for keeping it up).

>>42714722
I want a X230T
thanks for reading my blogpost
>44XPV

Technology for meditating on Linux?
I use Medito APK + Waydroid right now.

>>42719886
discipline and the ground

>>42719886
How does that work? I thought mediation is about sitting and thinking about nothing for period of time. How is any computer program gonna help with that?

>Ubuntu replaced GNU coreutils with uutils
>a "drop-in" replacement for coreutils written in Rust™
>uutils doesn't implement date -r
>autoupdate scripts are fucked

>>42720176
Absolutely unhinged move on Ubuntu's part. Rewriting things in Rust is a good idea in the long run, but why the fuck would you swap out essential components like coreutils without EXTREMELY thorough testing?

Main thing I learned from this is that non-LTS Ubuntu releases are basically the equivalent of Debian Sid. "If it breaks, you get to keep both pieces"

>>42720241
>Rewriting things in Rust is a good idea in the long run
the damn thing still has no functioning standard
writing system applications and targeting rustc versions is insane

>>42720248
>the damn thing still has no functioning standard
Neither did C, but they still wrote all of UNIX with it
>writing system applications and targeting rustc versions is insane
How do you figure? As long as you aren't using nightly features, things are normally very stable

>>42720318
>Neither did C, but they still wrote all of UNIX with it
1) C was created specifically for Unix
2) it quickly received a de-facto standard with K&R
Rust's been out about two decades and it's still impossible to use anything but the rustc
>How do you figure? As long as you aren't using nightly features, things are normally very stable
I had a pleasure of failing to compile projects with rust because rustc in Debian was "outdated"

>>42720241
I just expected them to do something like this in first or second release after a LTS, not right before one.

>>42720241
Rewriting things in rust WOULD be a good idea if people didn't just slap MIT onto everything. Someone pointed out a while ago that it's basically a big relicensing effort replacing all the important GPL software with cuck licensed versions and now I don't look at the "modern rewrite :)))" projects the same anymore.

>>42720338
The C compiler is relatively simple, the rust compiler is humongous and they're still not finished with the features they want to add because some of that shit is so complex. The main value of a standard is to allow alternative implementations and it's unlikely we'd see very many alternative rust implementations. Just look at C++, there are exactly two compilers people use, and the only reason it's more than one is - again - relicensing: a few big corpos didn't like the only compiler being GPL so they made a cuck license version. And C++ has had a standard for decades, there are multiple standards revisions even.

>I had a pleasure of failing to compile projects with rust because rustc in Debian was "outdated"
To be fair the main reason for this is that it's just not integrated into the OS. You would never be able to make a distro with like gcc7, since you wouldn't be able to build any C++17+ packages. But rust is not required for the core system so debian treats it just like anything else and keeps it a decade out of date because they have a fetish for that.
The only exception is if the developer explicitly opts-in to experimental, unreleased features ("nightly" rustc) but that's obviously not something you'd do for a system package, just like you wouldn't write system packages in C++23.

>>42720543
>Rewriting things in rust WOULD be a good idea if people didn't just slap MIT onto everything
Actually a good point. I saw a theory that Ubuntu might have switched to uutils as part of an attempt to remove all GPLv3 code from their IoT distro, to avoid the anti-tivoization clause.

>The main value of a standard is to allow alternative implementations and it's unlikely we'd see very many alternative rust implementations.
GCC is making decent progress on theirs. IIRC they're hoping to be able to build libcore from Rust 1.49 sometime early next year.

>Just look at C++, there are exactly two compilers people use
GCC, Clang, MSVC, ICC

>>42720620
>GCC is making decent progress on theirs
As a research project mostly. To be fair rust is also making progress on a specification last I heard of it. Regardless I don't expect anyone to ever make much use of gccrs, any more than people use gccgo or gccj.

>GCC, Clang, MSVC, ICC
MSVC is literally only used by microsoft and windows devs and icc is used by nobody.
I guess you might have an argument that there's a non-zero amount of windows devs so it's really three compilers, not two. Fair enough.

>>42720543
>But rust is not required for the core system so debian treats it just like anything else and keeps it a decade out of date because they have a fetish for that.
anon, debian updates its packages biannually
there's already an updated version of rustc in trixie-backports
the problem with rust are developers that target latest version of the compiler just because they develop on arch and expect others to also run arch
also I'm pretty sure Rust on Linux project uses experimental features kek
>>42720647
>Regardless I don't expect anyone to ever make much use of gccrs, any more than people use gccgo or gccj.
gccgo is good if you want muh optimization
gccj was a thing because JDK was proprietary

Looking to jump off the sinking ship that is Win10- debating whether I should stick to Mint, or dive into something more "advanced" like Arch.
I chucked Mint onto a travel laptop of mine that came with Win11, and it's served me well, but I wouldn't mind having a bit more of a handle over the system as a whole.
I feel like the most daunting part of migrating over would be moving all my files. This windows install is 9 years old, and as a result there's all sorts of shit everywhere.

>>42721119
Get a new drive, install mint on there, mount the ntfs drive as just a data disk
There's also https://github.com/maharmstone/ntfs2btrfs (though take a backup first ofc)

>>42720168
it's a free app with guided meditation I only like the daily meditations it has

i ordered a new thinkpad x220 tablet, a 500 gb ssd, 16 gb ram
the TP comes with a 128 gb (or something) ssd, so i'm going to use that for music
RIP my t420, i hope i can export my virtual machines off it without the left click not working

>>42722784
why you make me jelly

>>42722793
sorry anon
this TP lasted me about 3 years

>>42722840
i want to buy x230t to become a libre priest pony

>>42723436
>libre priest pony
insert you will never be a programmer copypasta

>>42719190
not this guy but bump
usecase:
posting a anonymous blog that cant be linked back to my IP/address/name/face
-with comment section + embedded images (not hosted on my server) where people can leave comments without me seeing their IP or fingerprint (no JS is that even possible?)
hosted on TOR I guess would be where I start

>>42724276
same questions as for that other anon:
>can't be linked back to my...
linked back by whom? if visitors, then whois privacy is enough. if you want to hide from registrars and webhosts too, it's tougher and you need a KYC-free payment method (mostly crypto). if governments then it's even harder
hosting purely on tor is indeed a relatively simple option but again only makes your blog accessible over tor

>-with comment section + embedded images (not hosted on my server) where people can leave comments without me seeing their IP or fingerprint (no JS is that even possible?)
no JS uploading/commenting is easy, though it'll involve refreshing the page (by submitting the comment form), you can't have in-page interactivity without refreshes if you're not using JS
not seeing their fingerprint is easy in that if you don't have JS, and don't do anything to get any sort of fingerprint, you won't have one
not seeing their IP is borderline impossible, it's up to them to use a VPN or tor or something, what you can do is just not log the IP anywhere
or we go back to the tor or i2p option, since your website will only be accessible over the darknet network it forces the user to not reveal their IP and so solve this

tl;dr is just a tor or i2p site is the most foolproof way but it's also inconvenient and 99.9% of people won't bother setting up the connection for it
it's viable if you have some specific community in mind that you want to move to your site and you know they will care and put in the effort to connect. but if you just want to make something useful for random anons and you put it on tor/i2p then probably nobody will ever connect to it

>>42721533
That would certainly be one way to do it, though I figure that migrating would finally be an excuse to actually pore over everything and actually organise/deduplicate/discard things.
Thanks for the suggestion, Anon.

>>42724276
Probably use JS chan hosted or tor / i2p or one of the forks sans js.
Or you can just use one of the altchans. That have tor / i2p mirrors.

aha

>>42725246
I suggest doing it this way first and then once your install is working you can migrate things at your own pace. That way you can keep using your PC in the meantime and don't have to both install the new OS and sort all your files all at once.
The downside is this will let you put off sorting the files until you get around to it never, but the upside is switching your OS won't be blocked by getting around to sorting your files. So that's up to you to see if you'd rather have an excuse to do it all at once asap or rather let yourself do it gradually.

>>42725246
My usual procedure when migrating to a new system is to make a squashfs image of the old drive. Squashfs is a compressed, read-only filesystem, so it takes up less space, but you can still mount it as normal to look around or copy files out any time you realize you need some specific stuff from the old system. You do end up with a copy of all the old OS files, which are usually redundant, but it's compressed so it's not that big a deal.

bumpe !

>>42727642
do iou speake french anon?

TECHNOLOGY!

>>42726054
>until you get around to it never
Yep. Took me until a few months ago to clear out my downloads folder, which had been building up since I built this rig.

>>42726738
Thankfully space won't be an issue, as I'll be going from 6TB to 14TB, more if I feel like reclaiming my current HDDs.
Thanks for telling me about Squashfs though, it seems like a really neat, niche thing to read up on.

>>42729505
>Thanks for telling me about Squashfs though, it seems like a really neat, niche thing to read up on.
The main thing it's used for is live CDs. Lets you pack 2GB of software into a 700MB ISO, and it decompresses on the fly so you don't need to unpack it all into RAM at once. But I like it for backups since you don't need any special tools to read it. Just mount it somewhere and all of your normal CLI tools / GUI file browsers / etc will work as normal.

>>42707237
How should be this category of mares called? So much of our technology is basicaly impossible without them and yet, i dont know how to refer to them

>>42729756
signalmare

>>42714722
holy shitt, WHATTT?!? thinkpad mare? :0
post more pics anon, whats her name?
lets see some art too

>>42729756
Do the knobmares go to 11?

>>42729886
I meant more in general , but signal mares would definitely be a big group

>>42726738
>Squashfs

>>42731930
Wait, it's... a colt!?
I though computers were a mare field???

>>42714722
Use case for hoarding e-waste beside showing off?

Bumping your technologyness

>>42729756
How deep do they go? Are they semiconductor level, or is it some higher level of abstraction? I think it's the latter, like the video signal mare - not a microcontroller, but closer to a piece of software, or rather a concept mare.
Nevertheless, innermares sounds cute and grasps the overall idea, although doesn't reflect that it's a ponification of technology.

would should i listen to/watch while i setup my new thinkpad?

>>42732632
long time investment
today it'll be worth $20 in ten years $200

>>42734313
well, here we go
500gb SSD, 16GB ram, arch linux on the pinkie pie flash drive, and my new x220T in the white box

>>42734794
how it's currently going, the battery is fucked, I already got a new one on the way
for now, it's gotta stay plugged in!
but wow, this screen looks AMAZING.
blackjack is here helping too

Just figured out an absolute doozy of a Linux problem, feels good.

>>42735006
Good luck anon!

>>42735006
IPS? I considered older X series but finding IPS ones is hard. X260 with 1080p IPS cost me $55 in 2022 and it's really swell.

>ios finally catches up to android
>sideloading works good
>RCS is fully implemented
>iMessage mogs whatsapp
>liquid glass is fixed and looks nice
>the phones themselves have become more repair-friendly
>apple silicon A and M series are extremely powerful and efficient
>good security
>cameras have improved greatly
>android slowly being sabotaged by Google and devs just creating ios emulators
>modern androids becoming diet coke iphones
>ios has insane software support; still updating the iphone 6s (and even still offers repairs for it)
I switched to android 2 years ago and I never thought I'd be giving Apple this much credit again.. is 2025 Apple's year? Even when I switched to an Xperia from a Pixel, Google's growing piles of shit are still present even when shying away from their hardware. I genuinely used to think Android was leagues ahead of iOS but now that the two are basically on level playing fields it looks like ios has just taken the lead.

>>42736228
Never spend more than $200 for a phone

>>42736239
nta but you and I speak on the same wavelengths.
>Never spend more than $200 for a phone
I've purchased Galaxy M15 last year for 140€ and I wanted to get myself a second backup phone. I don't know what to get though. I'd go for another M15, but they are unavailable brand new, and used ones cost even more than what I paid lol
Motorola is out of the question due to their lack of quality control (after 9 months, the battery in motorola swelled up and it didn't even last a day on single charge where as the samsung lasts 4-5 days)
Meizu Mblu 21 seems like an interesting option for backup, however I don't even know whether I should spend money on a phone I might not even need to use. I might just end up buying a phone the moment Samsung gets broken, lost or something, but I put a lot of care in it and I plan on using until around 2028 or 2030 if possible.
I'm on Android 15 but I don't plan on updating it further. I wish Lineage OS supported it.

>>42736293
additionally, maybe I should just wait for 2027 when smartphones will have replaceable batteries?

>>42736228
As far as vendor-locked phones go, I do agree that apple is mostly better than android, EXCEPT if you want to use your phone as a computer. Does it actually have a proper file manager now? I know it didn't have any for years. You also can't get a terminal on iOS, whereas termux is fantastic on android.
But if you don't need those things then yeah apple is good.

The major downside of apple is that it doesn't have any options for not being enrolled into the apple ecosystem. Correct me if I'm wrong, but as far as I know it is physically impossible to use one without an Apple ID. If you just want a phone, not an online account with T&Cs and shit, the iphone is basically a paperweight.
It's true that google is getting worse and worse lately but it's still possible to use an android phone without a google account, without google background services, and even with all network connections to any google servers disabled by default. If Graphene's own custom phone works out this will continue to be true despite google's efforts to enshittify the pixel. Meanwhile on apple this is not possible and I'm not aware of any indication it may ever become possible.

>>42736300
Battery life is basically good enough unless you need to use your phone constantly for a dozen hours at a time between charges.
For most people, it's very feasible to use the battery safety feature that limits charging to 80%, and still have enough juice at the end of the day that it doesn't go dangerously low (say not below 20%). This extends battery life greatly and you can basically expect a perfectly usable phone even 5 years into the future. So you can buy a phone now and then in 2030, if phones really have replaceable batteries by then, decide whether you need an upgrade.

Anecdotally I got my first phone in 2014 and used it until about 2023. It had something like 5 hours of screen on time remaining, and it was more than enough for a full day for me. YMMV I guess. I just had to recharge it every night but other than that I had no issues.

>>42736300
i doubt theyll ever bring back user swappable batteries like what was on the older phones. some phones have internal batteries that are getting easier to change out now but nothing like what was before will likely come back.
>>42736239
call me back when it isnt mediatek hell and when it gets more than a year or two of updates
>>42736316
>it is physically impossible to use one without an Apple ID
no? the appleid is just to download apps for the app store as well as icloud features (if you so choose to use them). youve always been able to skip icloud setup on all of apple's hardware and can still use them just find without one. worst case scenario you can just make a burner email only for that.

>>42736326
>no? the appleid is just to download apps for the app store as well as icloud features (if you so choose to use them).
Oh huh. Ok maybe I'm a retard then. Does apple have any equivalent to google play services, some un-disableable shit running in the background?

>>42736316
>Does it actually have a proper file manager now?
forgot to mention this point. it's had a file manager app built it since ios 11 in 2017 but has improved gradually over the years. i dont think its anywhere close to competency like android though.
>You also can't get a terminal on iOS
terminal apps do exist even on the regular app store
>>42736329
>Does apple have any equivalent to google play services
i mean i guess there's icloud services but those wont run at all if youre not signed into it

>>42736346
>file manager
That's neat.
>anywhere close to competency like android
A file manager is a file manager, surely it should be hard to fuck up
>terminal apps
Oh also cool
>icloud services but those wont run at all if youre not signed into it
That's pretty based, so by default it doesn't run shit in the background to track you? That's actually better than I expected. Maybe I was too harsh on iphones.

>>42736372
apple's invested a lot into privacy over the years
https://support.apple.com/guide/security/welcome/web

>>42736326
>i doubt theyll ever bring back user swappable batteries like what was on the older phones. some phones have internal batteries that are getting easier to change out now but nothing like what was before will likely come back.
I believe apple and other big brands might do something stupid that will make it hard to swap the batteries yourself, but iirc the phones sold in EU are meant to have use replaceable batteries by 2027.

>>42736300
Damn it, I was under the impression it was supposed to be in 2026 the entire time. I can only hope it'll start to get implemented sooner
>>42736399
Did they really or is it just more advanced form of privacy washing like they've been doing since forever? I wasn't following this much lately

>>42736399
I basically ignore everything corporations say about privacy these days because they have overused the word so much, "here are our privacy features" is meaningless in 95%+ of cases
Also that link is about security rather than privacy, it's also mostly just corpostuff about iOS being secure. Apple will consider iOS private and secure regardless of how much data is sent to Apple - "it's for your safety and security" after all. I'm not saying Apple is bad but that link is basically meaningless

>>42736409
the eu has always been doing their own thing like that, not sure if theyre gonna follow through. and fwiw theres quite a few phones these days that have batteries easier to remove than what companies were doing in the late 2010s and such. im of course entirely going off of people who know how to work on phones and have the tools for it
>>42736423
>end to end encryption on most cloud services
>dont embed trackers in 3rd party websites
>made cross app tracking opt-in
>private relay and similar services
>moving more things to be on device that used to be done in cloud
just to name a few. the website i linked there is apple giving a more detailed and technical explanation of their privacy and security measures, which i dotn see many other companies doing

>>42736426
this is the right answer really. no such thing as actual privacy these days unless you go out of your way to make convenience nonexistent. its always a game of lesser of two evils and it really just depends what conclusion you choose to draw from your own research. apple does have better security practices but whether or not theyre actually private is up to you

>>42736399
>privacy
>proprietary garbage
>>>/out/

>>42736399
Yeah, like trusting a wolf to protect your flock of sheep

>>42736326
>i doubt theyll ever bring back user swappable batteries like what was on the older phones.
EU mandate. Some gaming laptops already gimp themselves to use 180 W usb-c charger instead of 400 W humoungous brick.

>>42736399
>>42737246
>>42737495
In case you are interested in my rambling comment on this topic, I think it's important not to accidentally mix up security with privacy, optionally with anonymity as well. This makes such discussions less productive.
Does Apple and Google do decent job at security? To my best knowledge, I would argue they do. GrapheneOS autists seem to be even more passionate about this, e.g. they rightfully threw some comments at Apple announcing MTE like they're the hottest shit around. Pixels had this in hardware a couple of years earlier, but stock OS didn't enable that (at all or everywhere), GOS being more dedicated to this cause did enable it early. Caught some bugs in third party and first party software, so I understand why Google might've kept it quiet. Recent bug bounty changes on Apple side do reassure me someone there wants company to stay financially invested in this topic.
Does Apple and Google do decent job at retaining user's right to privacy? I would say it depends on your personal threat model. Sometimes the proprietary solutions might be better on paper, but it'll be difficult to validate their claims. It's not like open source = private, proprietary = not private. Of course I would prefer all software to be open source instead of black boxes, but it's not the world we live in, yet.
Privacy does require security - if someone can elevate privileges on your phone by sending you a funny pony picture with attached exploit for ten year old hole which the manufacturer of your phone didn't bother to patch, the privacy is lost. We do live in a sad world where phones need software put together for each model, and vendors seem financially encouraged to put out more models than they can feasibly support. I would say Apple's approach is better here - release mostly only flagships, release few of them, and poorfags can buy older models which you'll still support for years. Because it's still fewer models to support for 5y (or whatever) than Xiaomi's yearly lineup.
I applauded Google's Pixel team for keeping similar approach and even extending firmware+OS security patches to 7 years, but then some other team fucked with AOSP and made security patches embargoed for 3 months (1m previously). As GOS devs pointed out, this does seem irresponsible and only extends the time for exploiting known vulnerabilities by bad actors. And again, seemingly only GOS treats this seriously and using their contract with some phone vendor they can access those patches and do apply them quickly. Pixel's OS team doesn't, yet.
It's a clusterfuck, really, but as random anon with my current knowledge I am most inclined to recommend whatever hardware GOS devs claim solid, or a Pixel/iPhone on stock OS. Other vendors appear to treat security less seriously, unfortunately.

>>42736228
And to tag into this discussion, it hurts me to see AOSP getting progressively more of a walled garden, but I personally stick with this side of the duopoly. On iOS devices I am forbidden from running the OS I prefer, which goes against my understanding of the word "ownership". Of course for other people this is a less important aspect, so for better or worse it's good that we still have some competition in this market.
On Android side it seems like Google is done with gaining people's (and ponies'?) mindshare by loudly being "the open" platform, and is now trotting towards next step of the enshittification, locking things down. Unfortunately, not enough people care, and as a passerby I'm not equipped with the skills to do something about it, just occasionally comment online to give people ideas.
>>42736326
>no? the appleid is just to download apps for the app store
Which is a valid answer, but I'm sure you realize that this effectively means it is mandatory for most users. Unless you live the early Jobs' approach to third party software on iPhone 2G, where it was all meant to be webpages accessed via Safari.
Now, on Android it's not that much better, but installing custom software (some would call it sideloading, like it's supposed to be a thing for niche nerds only) is possible without any accounts and EULAs to accept. For now. Cue the mess with scammers sending .apk links via SMS to people and gently hoofholding them through the entire process of getting their money stolen, leading to Google getting a bit too monopolistic with their proposed solution of "true Android developers" list, enforced by devices running Play Services baked into OS. Sigh.
>>42736346
>i mean i guess there's icloud services but those wont run at all if youre not signed into it
Pardon me intruding, but I wouldn't speak such claims if this cannot be verified and/or I was not well versed in iOS under the hood. I'm not, and instead I would assume they still run something calling home - remember when people used to do fake activations via blackra1n etc. to have their old 3GS-ish iPhones get features like push notifications, inhouse YT app, etc. reenabled? Excuse if I'm remembering this not correctly, you've just reminded me of this era of tinkering with mobile devices from years gone. Well, similar era as when G4 ponies were published.
>>42736372
>to track you
But anyway, discussing this should be accompanied with a sort of threat model first. Do you need your device to do absolutely no connections to manufacturer? Are you ok with e.g. AGPS updates being pulled from the mothership? When does it stop being an added feature, and become something malicious you feel urged to disable?
Sorry for rambling again, those are such complex topics and it's even more difficult to discuss when you're not sure of the "tech level" of the other person. Or pony - I'd love to tell them about the marvelous and horrifying technology as I perceived it on Earth.

>>42737706
>and become something malicious you feel urged to disable?
To add to that, I've seen so many people online who follow certain chain of thought. I'll use Windows as an example since it's easier to digest, for me at least.
>I use Windows
>I don't trust Windows fully, and I want to do something about it
>I will go disabling parts of Windows using provided toggles
>I now consider this Windows install to be trustworthy
I'm not talking about disabling annoying things like that news&weather widget in bottom bar that for some reason (malicious UX?) is the only element there that requires only an accidental mouse hover to be activated. Ring a bell with AI features in other software nowadays?
I mean things like disabling services for "privacy", "not spying", etc. If such user does not trust the software in the first place - otherwise he wouldn't search for toggles, scripts, patched ISOs from strangers online even - why does doing that suddenly earn the trust (forever, at that)? If a system you don't trust seemingly lets you disable something, why do you trust that it did indeed disable the $badthing? Is it placebo?
Hope you know what I mean. As a techfag sometimes I go into daydreaming about how the fuck has this industry not yet collapsed due to amounts of trust we "must" give to operate our hardware and software daily, quite blindly.

>>42737706
>discussing this should be accompanied with a sort of threat model first.
Connections back to the manufacturer, or in general to places without the user's explicit intent, should either be opt-in or, pragmatically speaking, clearly documented and opt-out, so the user can decide themselves if some connections are worth it for them. For example, downloading GPS satellite almanacs is probably fine. Update checks from the manufacturer is also well advised. But maybe querying a location service for wifi SSID-based location is undesirable. And maybe a given user just happens to not use GPS that much and doesn't need or want to even download almanacs, and maybe you're schizo as fuck and want to only manually check for updates when going to an internet cafe rather than doing it from home. If the OS just says "yeah these things are probably useful for all users, let's have them enabled with no way to disable" then that's the wrong approach, and that's even assuming only necessary connections are used and not extraneous random shit that the manufacturer thought would be cool to have.

GrapheneOS is, again, a good example to follow IMO: https://grapheneos.org/faq#default-connections
Every default connection is clearly documented and they pretty much all have toggles. Want to disable connectivity checks? There's an option in the settings for that (I use it, because I can fucking tell when my wifi isn't working, I don't need my phone to ping some server to tell me that). Want to disable NTP/time updates? There's an option for that. Want to opt out of all location services? There's options for all of them in the settings.

Of course google play is particularly egregious because it's literally spyware running in the background and uploading analytics to google in real time. If Apple genuinely doesn't collect any telemetry on your usage, then I'd be willing to forgive if they don't provide an option to configure the NTP server or disable GPS almanacs.

>>42737706
>>42737760
Also yes to answer your question directly,
>and become something malicious you feel urged to disable?
IMO that's when I find it not useful and the costs are not worth any "benefits" you might get. A good example in web browsers is real-time search suggestions that send every single letter you type to Google, who happily harvests your typing cadence analytics in the name of instant search results; or the "safe browsing" feature that streams your browsing history to Google in real-time in the name of telling you when a link is considered malicious.

As for turning things off - it depends, if the software is not malicious then there may be genuine things I'd like to toggle on and off as actual tradeoffs between privacy and convenience. If a feature is added with good intentions then I don't necessarily immediately have reason to believe that the developer is out to get me. But it clearly has to have a genuine use, the privacy impact must be commensurate with the usefulness and minimised as far as the design allows (not maximised to use the feature as an excuse to gather as much data as possible, as is commonly done for some reason; for example, why not periodically download updates for a local list of known-malicious links for a "safe browsing" feature? the only reason to send your browsing to a server is clearly to maximise data collection). And then I can make a meanigful choice as to what tradeoff I prefer personally.

Another aspect is that sometimes there's legal weight to your toggles. If you go on a GDPR website and say "I do not consent to advertising tracking", legally they're not allowed to track you. Now they may choose to break the law, since enforcement isn't perfect especially if it's a slimier or shadier type of website that would be very hard to prosecute and isn't massively popular either, but at least in theory and at least for big, very visible corporations which are very much in reach of the law, it's a non-negligible assurance.

Now as to whether Apple can be trusted... I dunno, I don't know them well enough to say either way. I do think they're likely much better than google or microsoft at least. But there's also shit like the "feature" they had (have?) on macs where it sends every binary you run to apple to verify it's not a virus. Last I heard about it disabling was possible but required some hacky command reminiscent of arcane registry editing on windows, not a settings toggle (again I'm not sure if it's been improved now). That's kind of heavily leaning towards the antifeature side of things.

>>42737805
Gotcha, and that does align closely with my personal opinions on how things should be. You know, tech should serve the user and respect their wishes. It's not that hard to grasp.
That's another reason why I'm staying on the Android side, because these things are more difficult to hide/easier to fix than with a fully proprietary system like iOS. I'd have to blindly trust OS there doesn't have e.g. buggy toggles, or validate by doing my own traffic analysis with mitmproxy/whatever (assuming OS doesn't act maliciously when under scrutiny, e.g. if httpproxy != null then stopChangelings()), and reassess that every so often, ideally every software upgrade which is unfeasible for a single person. But you get it.
Software being source available (if some can't be FOSS), and working towards reproducible builds with as tiny trust base as Luna's derriere^W^W possible would likely get rid of so many concerns, no? Until then, GOS does seem most transparent about this.
Now, where are my bits for this organic advertising of a mobile operating system?

>>42737881
>You know, tech should serve the user and respect their wishes. It's not that hard to grasp.
Yeah pretty much. You know the web spec calls web browsers "user agents". Everyone knows it's called a user agent string (even if not everyone thinks about why). At some point along the way people just forgot that existed, and now user agents are actually remote agents. Really it should be considered a spec violation.

Oh well. 99% of people are either retarded, or just don't care and value convenience above all. 99% is a big enough portion of the market that mass-market software is free to completely ignore the existence of anyone else. And sometimes small manufacturers might want to use the remaining 1% as their market niche to avoid competing with the giants; but in markets with a strong network effect, like web browsers influencing web standards and web design norms, everyone suffers and there's no escaping.

i was gonna write responses but frankly theres nothing id be saying that hasnt been said before. i frankly take a more doomer take on tech ever since i got way more involved in it a few years ago. I've had my apathetic tech phase and my ultra privacy conscious phase and neither have been that great desu. i really dont think theres anything that can be done to reach a golden mean of usability and privacy/security/whatever buzzword. im tired man, i just wanna use things

>>42738055
Just use your phone less by uninstall the browser or something.

>>42738055
Answering this first since that's an easier post - I've mostly accepted everything in IT is a mess, and try to use and spread via word of mouth the less terrible solutions that some good people create. Instead of focusing on the mountain of misery, I showcase my family and friends things that can be better. I'm not just wallowing in the ever-present misery of being that 1% of more techy users, more and more pushed out of the picture by tech "trends" that 99% of people will be complacent to, not realizing the longer term impacts. I don't like calling majority of people retarded because they dont want to know tech, even though sometimes that is the case. I believe even ignorant ones deserve to be treated well by this field.
We won't fix everything, everything might go to shit one day, but at least we weren't doing nothing. I'm mostly okay with my current tech choices, there are some unresolved matters on my mind but I'm seeing it in more colorful, uh, colors. Some people say they'd like to be ignorant - I wouldn't give up whatever awareness I've got, if that lets me do something about these issues.
>im tired man, i just wanna use things
That's a very well put sentiment, though.
I wish I could just use tech and not immediately forecast how I might be getting screwed by it. To the tone of other anon's posts, I wish the good progress in tech wasn't only given to (You) if you accepted the regressions originating from, well, greed. Until then, we might need to use our remaining 2% of brain not yet consumed by pony, to provide solutions with that greed washed off.
>>42738072
Yeah. In some way, tech minimalism can make (You) feel less overwhelmed.

>>42713675
>havent sold a macbook in a month
>still have 3 or 4
>lowered prices really low
>people keep edging me on buying them just for them to buy another macbook instead
sheesh, i didnt think 2012 macbooks could go from desireable to garbage within literal months. now I have no idea what to do with what's left of these since i keep getting bites but no sales.

>>42738451
Not super following mac hardware, but I'mnot too shocked the market for non-M chip macs is dying

>>42738583
makes sense. given that the m1 series is 5 years old now, you can find an M1 air for about $350 on eBay. still way more than what I'm selling the older models for, but I figure those who can't into apple silicon yet have moved onto 2015 and onward for intel

>>42738055
Normies destroyed everything, as always. Tech was fun when it was a geek thing. It's just a chore now.

In case you're using Bitwarden and using the PBKDF2 algorithm, change the number of KDF iterations from the recommended 600.000 to 900.000 at least or preferably 1.200.000. The recommendation they're using is based on the performance of an RTX 4090. The RTX 5090 exists now and is about 25% faster, and the number they got was because they were counting the reference card (4070 Ti I think, but the document won't load right now to check) as 1.5. If you count it as 1, you needed about 900.000 according to the document. Since this calculation was done with prices from 2022, we can count it as 1 GPU now, so 900.000 x 1.25 = 1.200.000.

>>42739355
Sorry, a bit of a braino: (100 / 76) * 900.000 = 1.200.000 is what I meant.

>>42738055
>>42738154
I'd argue that in a way with how unnecessarily cumbersome a lot of things have become, tech now is a lot easier to use with a learning curve.
>Wanna watch a show
>Oh you have to use Netflix, but also HBO since Netflix might not have it, or.. uh.. maybe it's on Amazon prime now. Or wait no. It's apple.. you have an apple tv, right anon?
Or you could just go
>Fuck it I'm setting up the arr suite with jellyfin + jellyseerr
And now you have a legitimately easier solution to use that legitimately takes minutes instead of hours of trying to figure out how to get your shows, plus it does music as well, so no more Spotify vs apple music vs whatever.
I could go on about books, or any other media, just get an old kindle and jailbreak that, goes from a cumbersome ads filled brick to a "USB drive for bucks that happens to have a screen on it".

File management can be trivialized with copyparty / nextcloud if you need calendar / notes.
It all becomes quite comfy as you realize you don't have to deal with apple / google / whoever's bullshit moving target of an ecosystem.
And I think this is the most important part, these diy solutions may be harder to setup out front, but in the medium to long term they really just blend into the background.
Same for hardware, really, you could buy whatever android slop Samsung gives you, or just take a minute browsing the lineageOS wiki and having a perfectly good phone for less, getting last year's flagship.
And on pixels, graphene is legitimately fantastic too, sure it's secure but it's also a breeze to use and doesn't need you to debloat 10 AI features here, 50 advertisements there.
You just install it, put a poni wallpaper on and you're gtg.
So far my tech journey autism has been more about appreciating the simplicity of it and just mapping out what would cause the least friction medium / long term.

>>42739355
It shouldn't practically matter as long as your password is reasonably secure. Doubling the length of a brute force attack is mostly meaningless, if your password is brute forceable in a reasonable amount of time then double that time is probably still bad for you and an attacker could just pay double the cost. And next year there will be even bigger GPUs.
Besides, shouldn't it be set for the power of your own GPU rather than what's hypothetically on the market? An attacker could have access to an AI rig and then price calculations go out of the window.

>>42738154
>I don't like calling majority of people retarded because they dont want to know tech, even though sometimes that is the case. I believe even ignorant ones deserve to be treated well by this field.
This is a noble sentiment and I agree with it in theory, but my problem is with people that actively refuse to be helped.
>I can't be bothered installing an ad blocker (and especially: "ad blocking is piracy!")
>why do I care if google knows everything about me?
>yeah I heard X tracks you a ton but uhhh I'm used to it I guess... switch to Y? nah I don't wanna use something different
Back when I was in uni, I was working on a group project with genuinely smart and competent people and I was the only one who didn't use fifty social media sites. I got asked "why don't you wanna use facebook? is it just because of that privacy thing?" as if I was a schizo. Again, I want to stress that these were intelligent students at a relatively prestigious first world university, studying computer science, and being generally competent at tech.

>current year
>STILL no functional robot mares
It can't be that hard, right?

does anyone have any ideas for a cool xfce4 rice? how do (You) rice xfce4? Just got all my files moved onto this new hard drive and am ready to rice, but want to switch it up a little
>>42735920
>>42736085
Might be IPS? the tilt screen test works so, idk how else to check it
the listing doesn't mention it being anything but it looks nice

>>42739879
Like the document says, preferably you shouldn't use PBKDF2 in the first place, because it is slow for the defender and fast for the attacker. However, it's FIPS-140 validated so if that's a requirement some may prefer it. Another option Bitwarden has is Argon2id which might be more secure, but I haven't looked into it. This was just a FYI to anyone using it that doesn't want to delve too deep into it. Obviously you need a strong passphrase in the first place - this is just an extra layer of defense on the most important of your passwords: the password guarding all the rest of your passwords/secrets. Performance impact of doubling KDF iterations on my reasonably-high-spec-but-still-quite-old-now desktop is minimal. There might be more of a performance hit on a mobile phone, but I don't use those for this purpose.

>>42741335
maybe any cool xfce themes?
besides chicago95, kek

>>42741290
wasnt there a sweetie bot in the works several years ago? does my memory serve me right?

>>42743570
https://sweetie.bot
yes, and according to their website, we can have our own, soon TM.
https://x.com/SBotProject
they're still relatively active, they take the robot and showcase it's functions and capabilities at conventions
i mainly see them in cons outside the US if i'm not mistaken, idk

goodnight thread

>>42744273
>We are going to make Proto2 open source by the end of the year. We hope to create a open source community for this four-legged robot platform, so a lot of robot developers around the world can communicate and work together on this project.
>On August 19-20, 2017 we will participate in CZequestria convention, which will take place in Prague, Czech Republic. We’ll bring Proto2 and going to make a short presentation about our project on stage. If you will be there come to our table to give a brohoof to the real pony robot!
>Besides that, we are working on ProtoPony3. She has the unique design and a full range of technologies that makes her the most awesome and cute robot in the world!
>We hope to finish her and present at the DerpFest’2017 — the upcoming Russian MLP:FiM convention held in Moscow on November 4-5, 2017. Hope to see you there!
I know that the project is still alive but even like five years ago people were calling it out for milking the patreon. Either they've given up on their english language website and are only posting live updates on tabun or something, or it's basically dead by now.

>>42741335
You could just skip straight to window managers, they do take more tinkering than desktop environments though. But if you're already familiar with Linux they should be a straight upgrade.

>>42745124
I think they showed updates at a chinese con recently but yeah, the project looks to be on life support

>>42745217
what about xfwm? is that relatively easy to install, from what im seeing?
>>42745124
wait...
when were the sweetie bot kits announced?
how long ago...

>>42729756
I'm starting to wonder if they would be somehow related to plones.

>>42746435
how long that take?

>>42746919
>https://docs.xfce.org/xfce/xfwm4/introduction
>Consequently, special features such as making windows borderless, or providing an icon for the application must now be implemented in the application; you can no longer use the window manager to force different behaviour.
That's silly. If you want to stick with X, i3 is a good option. If you want to try wayland, niri is a fun one, that's what I use. Plus, the kirin! (the creator used to have a kirin pfp on his github) Sway, wayland's i3 replacement, has a grudge against leaving xwayland scaling alone for some reason and screws your xwayland programs scaling over, very annoying. Hyprland's developer thinks unused RAM is wasted RAM, so yeah. i3 or niri.

How do I bypass a site's CSP? I just want to use AJAX to load and save comments on a website serving only static content. The site doing the database work can be hosted anywhere, like Google Forms.

Bmp

>>42749104
>xfwm4 adheres strongly to the standards defined on freedesktop.org
oh no it's retarded
fuck, I thought xfce was based

>>42749120
If you want to be able to POST to a different origin, set up CORS headers on the destination server. CSP is a totally different thing though

>>42749104
>Sway
Sway or wlroots? Because if it's just Sway there are fun wlr-based compositors like wayfire or labwc

>>42750189
Aren't all wayland compositors based on wlroots?

>>42752932
KWin and Mutter are independent

>>42752932
Hyprland is independent

>>42746435
Makes me think why even bother include the ability to draw on that when everything is bound to look like shit on that resolution
But the pony is beautiful of course

Any suggestions for a web host that allows NSFW content, doesn't show ads and doesn't block Tor/VPNs from signup or usage of the site? Sites that require KYC-less crypto payments of at most $1 a month are acceptable too.

>>42754154
Free web hosts generally mostly work for static content. My go-to is gitlab pages.
For dynamic content you probably want paid hosting. I dunno any myself but check out kycnot.me. I'm not sure you'll be able to find KYC-less proper hosting for less than $1 though

>>42754154
https://crippled.media/article/free-speech-vps-providers-put-to-the-test

https://lists.debian.org/debian-devel/2025/10/msg00285.html
"Universal Operating System" is getting Rust as a mandatory dependency for its package manager
rustc doesn't even properly support RISC-V, PowerPC and more exotic architectures that some people can only run on Debian

>>42754656
It's almost like there's an ulterior motive.

>>42754656
>rustc doesn't even properly support RISC-V, PowerPC and more exotic architectures
What are you talking about? I use rustc on ppc literally every day

>>42754656
>It's important for the project as whole to be able to move forward and rely on modern tools and technologies and not be held back by trying to shoehorn modern software on retro computing devices.
Ah yes, that's a great argument. Shaming tinkerers for liking to tinker.

>>42754843
>I find this particular wording rather unpleasant and very unusual to what I'm used to from Debian in the past. I have to admit that I'm a bit disappointed that such a confrontational approach has been chosen.
Looks like even Debian devs are getting fed up with this bullshit.

>>42754838
https://doc.rust-lang.org/rustc/platform-support.html
only amd64 and arm64 are "Tier 1" platforms for development

>>42754653
Goddamn that's grim.

>tfw completely forgot I had lidarr setup
>tfw suddenly notices new music in my library
For a second I really thought I was being hacked and the dude was just telling me to listen to that new album.
Fucking hackers, man. Even they aren't above shilling their mixtapes.

btw I use Arch

>>42755964
>look at her
>she's your distrowaifu now
Still the sweet spot of available flexibility and required maintenance for my desktops.
>tfw archpone plush laying on your literal desk top

>>42755964
cause you're a faggot

ive been in a dwm binge... currently settijg up picom

now that 4ch rss is dead what can i use as an alternative? 4ch-xt and dollchan are alright but i need to keep a tab open for them to work and i still miss the filtering i could do with rss for new threads.
inb4
>apps
lol

>>42758570
I was thinking about making RSS generator from catalog for some time. I'll probably get to it now since I won't be able to get it out of my head anyways.
It's nice to see I'm not the only one who misses it

>>42758570
Hmm, I more or less have finished most of the work for this when making a thing dumping out threads as they archive/gets deleted. Just going to setup another Atom-export for the created threads and it should be go

>>42759735
Try it out, it will slowly populate as threads get created.
https://apanx.net/mlp_new_rss3.xml

>>42760097
noice this will save me some work

>>42760097
will it slowly fetch 10 entries? It's full of deleted stuff right now

>>42761169
I didn't add a provision to store creation date for the threads in the database until now as it was only meant to fetch archived threads before. So all threads before 42759942 has a creation timestamp of 0, and the RSS file writer only adds up to 14 days old threads.

so i went on an excursion today to try and find a VCR. i thought it would be easy since nobody uses them anymore, but boy was i foolish. i went to every goodwill in the county as well as 2 of the neighboring counties before finally finding a single one at the 12th store i visited. i knew VCRs were dead but not THIS dead..

>>42761610
Why do you want one?

>>42760097
hey anon, do a proper wildcard certificate and enable full trust https on cloudflare. your VPS is leaking info / doxxing your stuff.
also scarf page when?

>>42760097
also secure that ssh port, anon.

>>42754656
I'm so sick of the rust cult.

After 17 years there's still no reason to use Wayland

>>42762866
There is, or at least I found one, on X my windows tear and the whole OS feels clunky and slow when using multi monitor at 144hz whole Wayland is just snappy.
Though, the tradeoff is having to deal with portals and drag and drop not always working though that's gotten better.
Frankly none of X or Wayland is ideal they both kinda suck.

>>42762866
after switching to Wayland my multimonitor setup isn't buggy anymore
(X is still good and I hope XLibre manages to get into Debian 14 repos)

>>42762866
Maybe it's time to make an Xserverer Xserver to replace them both. Make it hypertextual and full of category theory and, what else people like nowadays? Blockchains? Yeah, sure, add it too somehow.

>>42762387
>Certificate Transparency amre
>sort of an archivist
>notes down every cert her sister - Certificate Authority - forges for ponies
>she's precise to the dot, if you ask her about some pony's orders she will be urged to tell you absolutely everything her sister has worked on for this pony
>she was like that since she was little, at school some fillies and colts called her a gossip filly, but with changeling CAs nowadays she's committed to doing this so that ponies can trust her sister
>Twi is her role model, obviously, but on visits to the castle she likes to discuss pens with Raven
>showing her your long lists and mentioning ink on her coat make her blush

>>42762967
>mentioning
or maybe replace that with "rubbing" or "spilling" to match the lewdness, since that's what this line already implies with spoilers.

>>42762967
Don't forget about CSR filly, she's real fast and wanna learn to be as efficient as her big sister one day

>>42763001
>AES-NI amres
>total nerds
>all they ever cared about is hiding things from changelings using maths enigmares?
>but mare, are they a bunch of raceponies when it comes to crunching PKI (pony key infrastructure) numbers

>mare in the middle
>it's just a friendly mare passing on the piece of information you wanted to send to someone else
>she would never even peek at the contents, it's not up to her code
>she simply takes pride in extending your reach or letting you send things while recipient is offline

>>42763072
>PFS mares
>Don't trust anyone
>Deeply paranoid
>changes their locks every time they enter their houses
>"Hey anon, want me to tell this to your friend?"
>"Sure, I'll tell them that too, but in a different way!"
>Hate metadata mares
>can't change their speech

my stinkcentre just arrived

>>42763150
>cpu i3
>gpu xeon
wtf

>>42763150
Use case?

>>42763162
yeah idk
>>42763508
i replaced my ewaste atom shitbox with this; i'll use it for torrenting and file syncing

>>42763569
>Trixie's dock imprinted onto CPU lid
lewd

>>42730066
her name is thinkpony c:

>>42763995
>magnesium horn and socks
Now that's durability. Also, extremely cute pone despite trotting on the edge of being over-designed. Thank you.
Wouldn't even think about asking about her TrackPoint.

Are there any cybernetic enhancements for increasing milk production?
Asking for a friend

I've been using Mint Cinnamon for a while now, but since I "upgraded" the display to 100hz one, I get white "snow" effect after some time if it's set to 60hz. I blame Cinnamon's desktop compositor for that, so I think of switching to XFCE edition instead. I will do a clean install, however I'll copy and paste my home and config folders, will that be sufficient?

>>42763162
NTA but I thought it might have been a cope GPU soldered to the motherboard like on some server boards, but that CPU has an Intel® HD Graphics 2500 iGPU
Driver issue, maybe?
>>42763641
Unf

>>42754577
>gitlab pages
Isn't that just for code?

>>42764940
It's for static sites

>>42764955
Yeah, for code projects right?

>>42765034
Any website is code.
If you're asking whether gitlab has any rules that you have to have some related coding project that your website is for, no, there's not. Your website itself is code, you put it on gitlab, and set it up to be available as a Pages site, simple as.

Github has a similar thing but with some more restrictions, Gitlab is particularly generous and you can even do shit like commit 1GB of images into your "code project" for your website because you're just that lazy.

>>42765077
Hm, I see. Would feel wrong to put my pony fansite on a site meant for Linux open sauce projects though. And they probably don't allow NSFW anyway.

>>42762154
i need one for something im doing