00:05:06
ofrnxmr:xmr.mx:
Im 90%
00:07:11
ravfx:xmr.mx:
How can you test for the exact % ?
00:12:52
imprevisto:matrix.org:
there are ways
00:14:25
gan:skhron.org:
@ofrnxmr:xmr.mx: Where you'd place yourself on the Kinsey scale? 🧌
00:15:55
gan:skhron.org:
Or the remaining ten percent is your robosexuality?
00:16:03
gan:skhron.org:
I'm pretty sure, there must be something else 🧌
00:17:33
ravfx:xmr.mx:
When your bored and want to test if the slopinator can do assembler in 2026
00:17:35
ravfx:xmr.mx:
https://mrelay.p2pool.observer/m/xmr.mx/fuTdPgAMqbwmaZErNkPyDXpn.png (clipboard.png)
00:30:43
ravfx:xmr.mx:
Still pissing
00:30:43
ravfx:xmr.mx:
it's going to use undocummented shit (unreal mode)
00:30:45
ravfx:xmr.mx:
https://mrelay.p2pool.observer/m/xmr.mx/iVZbdkunrYYoknUOicothzEd.png (clipboard.png)
00:30:55
ravfx:xmr.mx:
If only the model would fit in my vram... it's slow
00:54:11
ravfx:xmr.mx:
That's still dumb kinda
00:54:11
ravfx:xmr.mx:
it know .bss can only contain resb, resw and resd, then why he put other shit in .bss 😂
00:54:11
ravfx:xmr.mx:
They need to invent double thinking model or something!
00:54:19
ravfx:xmr.mx:
https://mrelay.p2pool.observer/m/xmr.mx/JImfxSibrriJFfpHaabnEwVJ.png (clipboard.png)
02:12:33
ravfx:xmr.mx:
So basic simple code do work now.
02:12:34
ravfx:xmr.mx:
(I did the same test without asking it fancy stuff, like cli only, and it just worked, surprisignly).
02:12:34
ravfx:xmr.mx:
If asking to do more compliqued stuff, it get just borked.. Still better than last year test.
02:14:15
ravfx:xmr.mx:
It just got totally mixed between 16 and 32 bits because it choosed to do Unreal (to be able to access the VBE 2.0 frame buffer from "real mode")
02:22:02
ravfx:xmr.mx:
Now it's a infinite amount of "prompt debuging"
02:22:03
ravfx:xmr.mx:
So I will consider "complexer programs" failed
02:31:51
321bob321:
Hair colour > <@ravfx:xmr.mx> How can you test for the exact % ?
02:32:49
ravfx:xmr.mx:
https://github.com/wbcbz7/S3VBEFIX/
02:32:50
ravfx:xmr.mx:
How nice, you can actually take that code and not give any atribution to author
02:33:25
ravfx:xmr.mx:
No idea why people would do that 😂y
02:37:37
ravfx:xmr.mx:
@321bob321: What if no hair?
02:37:58
321bob321:
Wig
02:38:35
ravfx:xmr.mx:
But what if it's shiny skin or something like Vik?
02:45:18
321bob321:
Eye lash colour
02:46:17
ravfx:xmr.mx:
That does not work, and even if it did, you could not extract a exact % from that
02:59:44
321bob321:
Beard colour male or female
03:28:04
imprevisto:matrix.org:
does dans curtains match the drapes
04:16:16
orly_owl:
only if you find a house sexually attractive
04:45:20
ravfx:xmr.mx:
I think I did not reserve enough memory for the context and some got overwritten or something.
11:20:53
basses:matrix.org:
anyone owned a Motorola device? was it good?
11:21:05
basses:matrix.org:
because im 99% sure next gos phone will be it
11:34:22
btcdwed:
i owned one, wasnt satisfied at all
11:34:25
btcdwed:
but some yrs ago
11:34:57
btcdwed:
moto G bla bla 10yrs ago
11:35:01
btcdwed:
100% crap phone
12:09:22
milas900:matrix.org:
Is it now vpn detected easily ?
12:14:07
orly_owl:
99% of phones make the rest of them look bad
12:15:41
btcdwed:
could be, yea
12:15:49
btcdwed:
moto has some nice devices, btu not the one i had :D
12:17:46
milas900:matrix.org:
Is it like there is new method that apps can capture vpn !? What’s wrong with vpns nothing works
12:19:24
321bob321:
https://www.bleepingcomputer.com/news/security/curl-ending-bug-bounty-program-after-flood-of-ai-slop-reports/
12:21:31
Cindy:
this is hackerone's fault
12:21:55
Cindy:
they seem to be too busy ass-kissing clankers to care about the spam ruining libcurl
12:25:45
BlueyHealer:
hackerone has been doing what??
12:35:27
Cindy:
BlueyHealer: AI stuff
12:38:02
BlueyHealer:
Yeah, you told that. I mean whether they incentivize that or just don't punish.
12:41:58
Cindy:
https://www.helpnetsecurity.com/2026/01/20/hackerone-good-faith-ai-research-safe-harbor-framework/
12:42:49
Cindy:
curl pulling out of hackerone happened 2 days after this
12:43:01
Cindy:
so it seems pretty connected
12:43:47
bikrambiswas:matrix.org:
@basses:matrix.org: Yes ..I am using it ..It's stupid
16:01:07
gan:skhron.org:
@milas900:matrix.org: What? the VPN exits always has been easy to detect, as majority of them are on commercial IPs, it's clear by their ASNs
16:52:19
basses:matrix.org:
https://github.com/QwenLM/Qwen3-TTS
16:52:19
basses:matrix.org:
https://www.youtube.com/watch?v=dQ841Pd6YvQ
16:52:27
basses:matrix.org:
FOSS China keeps killing all proprietary services
17:27:38
BlueyHealer:
Is it really FOSS though? Like, what counts as FOSS in case of AI?
17:28:38
BlueyHealer:
That really confuses me.
17:31:16
hooftly:matrix.org:
What is confusing? Yes it is Foss
17:33:59
BlueyHealer:
I mean just what is "FOSS" in case of AI? Models aren't just code people wrote, but the data it's trained on - what exactly has to be published to count as such?
17:35:46
BlueyHealer:
Sorry if the question is stupid but that's just what I still don't understand at all.
17:36:17
gan:skhron.org:
That's classical case of source code vs content, in para-Stallmanite world, the software could be free, but the content could be not, e.g., Quake engine is open-source, but the content itself is not
17:37:15
BlueyHealer:
Yeah, I know that, but doesn't seem as directly applicable to AI.
17:37:21
gan:skhron.org:
The thing about libre content is usually aligned with the free softwareist free cultureist - e.g., https://libregamewiki.org/Main_Page
17:38:07
gan:skhron.org:
BlueyHealer: Depends, the trained data could be libre, see: freedoom for the doom engine for example
17:38:16
gan:skhron.org:
Therefore that'd render the whole project free
17:39:07
BlueyHealer:
So it's about formalities.
17:40:01
BlueyHealer:
Like, what exactly is FOSS in case of Ollama/Qwen/Deepseek?
17:41:01
BlueyHealer:
From what I understand, just the initial training consumes so much more resources compared to just running it - so it's not actually reproducible with "bring your own data"? Or I am misunderstanding something?
17:41:18
gan:skhron.org:
Nothing, FOSS is a corpo and community-friendly term for derailers of the free software movement
17:41:31
gan:skhron.org:
I'd rather prefer to hear open-source being used here
17:41:33
gan:skhron.org:
because that's true
17:42:32
BlueyHealer:
I saw this mentioned in other chats and still don't understand how "FREE and open-source" term is different from just "FREE", but okay.
17:42:40
gan:skhron.org:
BlueyHealer: Indeed, that's why I've said that it could be free, contextually, that's very hard to achieve because of the hardware requirement
17:43:48
BlueyHealer:
So I understand it right that those things are open-source wrappers around black boxes? And do we get more insight into or control over those black boxes compared to fully proprietary ones?
17:48:06
gan:skhron.org:
BlueyHealer: Originally, the term "open source" was simply a way to deal with the linguistic feature of the English language, that is, free as in freedom vs free as in gratis problem, but later on it simply devolved into what it now disregarding, 4 essential freedoms of the individual were exchanged for corpo-related freedoms [... too long, see https://mrelay.p2pool.observer/e/rILiq98KcHhlX3VR ]
17:49:00
gan:skhron.org:
BlueyHealer: Partially true, there's still more control over such data, and it's possible to train the data on your own, but that'll be tedious in multiple ways
17:50:51
BlueyHealer:
I know the difference between "free software" and "open-source", but not sure why "free and open-source" doesn't fit the definition.
17:52:52
gan:skhron.org:
BlueyHealer: The latter could fit the definition, but I think something like non-free ML data is stretching it
17:53:48
BlueyHealer:
I mean in general, not about LLMs.
17:54:00
BlueyHealer:
In that case that was my initial confusion.
17:56:02
BlueyHealer:
Also I really wonder if there is some way to check that "this dataset provably results in this particular model", if there was some that was fully transparent, since it's not like you can just recreate it on your gaming rig.
18:00:33
gan:skhron.org:
BlueyHealer: I generally dislike the term FOSS/FLOSS as a personal opinion, as in old man is yelling at clouds-type of thing to be fair
18:01:05
gan:skhron.org:
That should've been an off-hand comment instead of something argumentative
18:02:25
BlueyHealer:
I just use it interchangeably with "free" because it's clearer, but not interchangeably with "open source" because that's indeed MIT and such. I just still don't get what the difference is there. But whatever, not really wanted to discuss that either.
18:12:51
imprevisto:matrix.org:
I think free means open source (always?)
18:13:09
imprevisto:matrix.org:
because you need the freedom (right) to read it
18:13:49
BlueyHealer:
Free means open source but not the other way around
18:14:01
imprevisto:matrix.org:
right
18:14:07
imprevisto:matrix.org:
that's my understanding
18:14:30
basses:matrix.org:
you will never see them as they are trained on copyrighted material, but them being chinese they might do it but they know western media outlets will give them bad PR etc > <BlueyHealer> I mean just what is "FOSS" in case of AI? Models aren't just code people wrote, but the data it's trained on - what exactly has to be published to count as such?
18:15:37
basses:matrix.org:
so wont every publish it
18:15:41
basses:matrix.org:
ever*
18:15:54
BlueyHealer:
Yeah, and I guess Western models are no different. Thus I am really unsure if we can call this FOSS?..
18:22:25
Cindy:
guys, i wanna ask you a question
18:22:40
Cindy:
how much should open source hobbyist programs prioritize security
18:23:23
gan:skhron.org:
Depends on what kind of security, and if it stands in the freedom's way
18:23:45
gan:skhron.org:
Security should be up to the user in my view, but a common pattern should be accounted for
18:24:37
basses:matrix.org:
Cindy: to max
18:24:57
basses:matrix.org:
or you will be helping people saying FOSS =! secure
18:25:32
BlueyHealer:
Up to the user? I don't think so. Sensible defaults matter A LOT. Even corporate software (like Cisco) turns out to have such a problem!
18:26:41
gan:skhron.org:
@basses:matrix.org: Nothing is secure unless a person vetted the software
18:27:00
gan:skhron.org:
And that's at least possible which is not the case with non-free software
18:27:51
Cindy:
what if you make an emulator, and there was a program that "escaped" your emulator
18:28:04
gan:skhron.org:
BlueyHealer: Depends if those sensible defaults stand in my ways, I don't like the route that Android went
18:28:15
Cindy:
should you add mitigations to prevent those? (costing performance)
18:28:19
Cindy:
or not
18:28:31
BlueyHealer:
gan, nobody can vet 100% of the code they're running though, so it's more about reasonable trust.
18:28:35
Cindy:
(costing some minor amount of performance)
18:28:37
Cindy:
*
18:28:38
basses:matrix.org:
@gan:skhron.org: which even not guarantee they will catch all security issues
18:28:58
BlueyHealer:
gan, you mean like with downloading from arbitrary sources?
18:29:06
basses:matrix.org:
cant you reference already similar foss projects? > <Cindy> what if you make an emulator, and there was a program that "escaped" your emulator
18:29:10
basses:matrix.org:
see how much they care?
18:29:13
BlueyHealer:
Also yeah, Android went overboard witn control
18:29:30
BlueyHealer:
Cindy, YES
18:29:33
Cindy:
i can reference zsnes
18:29:37
Cindy:
and project64 1.6
18:29:45
Cindy:
both of which had emulation escape vulnerabilities
18:30:01
Cindy:
i mean, HAVE
18:30:04
basses:matrix.org:
seems fair, these are ancient, right?
18:30:08
DataHoarder:
if they expect end users to run romhacks, yes
18:30:30
gan:skhron.org:
I'm aware, but there's fundamentally no trust, that's simply an illusion, you're free to read something like "Reflections on Trusting Trust" by Ken Thompson > <BlueyHealer> gan, nobody can vet 100% of the code they're running though, so it's more about reasonable trust.
18:30:54
Cindy:
basses: old, yeah
18:31:47
gan:skhron.org:
Absolutely, and no amount of the safest programming languages being thrown at it would fix the basic issue - PEBCAK > <@basses:matrix.org> which even not guarantee they will catch all security issues
18:31:47
Cindy:
and both emulators have since been discouraged from being used
18:32:01
Cindy:
at least.. i think in most places
18:33:47
Cindy:
i think debian patched zsnes to fix the vulnerability, that's the most amount of update the emulator has seen
18:33:49
gan:skhron.org:
BlueyHealer: Inability to change non-system system applications (like the Camera app) without being rooted, AVB, Play Integrity and so on
18:33:52
Cindy:
in years
18:36:03
Cindy:
gan: yeah, safe languages like rust aren't a magic bullet, but they do stop you from falling into the common pitfalls of writing a program
18:37:01
BlueyHealer:
What is AVB?
18:37:48
Cindy:
android verified boot i think
18:37:52
BlueyHealer:
Also Play Integrity is more of a problem with the multitude of apps demanding it than with Android itself. Never yet dealt with such apps. But yeah, the fact that it exists us scummy.
18:37:57
gan:skhron.org:
Cindy: That's true, and I find them preferable, I generally write in C and I'm learning Rust at the moment
18:38:02
gan:skhron.org:
Cindy: Yup, it is
18:39:02
Cindy:
and some crates do take advantage of the trait system to stop you from making mistakes too
18:39:19
Cindy:
like the rng crate having a special CryptoRng trait only defined for the RNG algorithms that are cryptographically secure
18:39:34
BlueyHealer:
gan, what resources would you recommend on those languages? I want to learn to read such code, if not write myself.
18:41:17
BlueyHealer:
Also I don't know how licensing works. I know the Rust rewrites of some utilities that Ubuntu was doing also meant a MIT license, but don't understand whether this is inherent to Rust or just a choice.
18:41:40
gan:skhron.org:
BlueyHealer: I learn stuff by doing stuff generally, and look at other projects and how they do it, Rust book is already a good entry
18:42:37
BlueyHealer:
I have a book on C/C++ but don't know which one to choose for Rust (I guess I should learn it as I'd be coming across quite a bit of such code).
18:46:19
Cindy:
BlueyHealer: just a choice
18:46:35
Cindy:
there are GPL rust projects
18:47:09
BlueyHealer:
Ah, nice. So just an excuse? Will read on that tomorrow, there have to be discussions on that.
18:47:09
gan:skhron.org:
Cindy: Well, they probably want their stuff to be reused by other corpos, since they kinda hate GPL
18:47:30
Cindy:
i think i remember seeing a list of most used license in cargo
18:47:32
Cindy:
licenses*
18:47:55
Cindy:
https://lawngno.me/blog/2025/03/06/crate-licences.html
18:48:13
Cindy:
MIT is 1st, Apache 2.0 is 2nd, GPLv3 is 3rd
18:48:56
gan:skhron.org:
I use 0BSD and the Unlicense for code, and CC0 1.0 for content personally 🚎
18:50:14
Cindy:
gan: well honestly the choice of license really depends on what you want
18:50:22
Cindy:
if you want everyone to use your library, use MIT or BSD
18:50:38
Cindy:
but if you want your library/program to have copyleft protections, use GPL
18:51:59
Cindy:
https://rust-lang.github.io/api-guidelines/necessities.html#crate-and-its-dependencies-have-a-permissive-license-c-permissive
18:52:55
ravfx:xmr.mx:
copyleft don't protect anything.
18:52:55
ravfx:xmr.mx:
GPL is copyright with permission set in the license
18:53:18
Cindy:
that's every license
18:53:23
ravfx:xmr.mx:
yep
18:53:50
Cindy:
unless you give up your program to the public domain, it's all just "copyright with permission set in the license"
18:54:37
gan:skhron.org:
To me, Capital is the temporary thing - It's impossible to take it to the grave, but the freedom is eternal thus I'm inclined towards GPL, but personally I'm no copyrightist, so, everything that I do is given into public domain using equivalent licenses
18:54:44
ravfx:xmr.mx:
Even public domain, copyright apply, someone else can't come it and say he invented it or made it.
18:55:32
Cindy:
gan: GPL will last for your lifetime + 70 years
18:55:59
Cindy:
or whatever, depending on your jurisdiction's copyright laws
18:56:43
gan:skhron.org:
Furthermore the enforcement of the copyright requires a state, I'm not quite a fan of it, I just like to see corpos being trolled by the GPL-related licenses in the end as well
18:57:00
ravfx:xmr.mx:
It require a lot of money too
18:57:10
Cindy:
it's not eternal
18:58:21
gan:skhron.org:
Cindy: Well, that's a bit pedantic but also true, my comment was more sentimental
18:58:37
Cindy:
also those 70 years require someone to defend for you after you're dead
18:59:13
Cindy:
so for most people it's more like lifetime + 70 years of people probably violating your copyright and letting you scream from your grave about it
18:59:15
gan:skhron.org:
I guess, also my lifetime is the eternality, if we'd think about it
18:59:56
gan:skhron.org:
Cindy: I think, that's why FSF requests copyright assignments to them
19:00:52
gan:skhron.org:
Cindy: Eh, I anyway release the stuff into the public domain 🧌
19:02:04
BlueyHealer:
How often do corpos get exposed for violating GPL?
19:02:47
Cindy:
to expose a corp for violating GPL, you have to reverse engineer their program to find it
19:02:50
basses:matrix.org:
a lot
19:03:07
Cindy:
or find a sign of your program being used
19:04:08
gan:skhron.org:
BlueyHealer: For example, Tesla v. The Software Freedom Conservancy
19:04:11
ravfx:xmr.mx:
more than you think, but it's not that easy
19:04:19
ravfx:xmr.mx:
It require time and money
19:04:39
ravfx:xmr.mx:
and yes, you have to find the violation first
19:04:43
Cindy:
yes
19:04:48
Cindy:
that's most of the problem
19:04:55
Cindy:
finding the violation in the first place
19:05:04
BlueyHealer:
I guessed this isn't uncommon because it's harder to prove than, say, torrenting AI training materials...
19:05:08
Cindy:
y'know, it's pretty hard to extract firmware from a locked down device, especially from a car
19:05:14
BlueyHealer:
Noted down that case
19:05:41
gan:skhron.org:
But also we have Allwinner Technology that has been violating GPL licenses for years
19:05:42
ravfx:xmr.mx:
And sometime it's encrypted too
19:05:55
ravfx:xmr.mx:
@gan:skhron.org: It's the one for ffmpeg?
19:06:00
Cindy:
yes, you have to reverse engineer all of that before you can start finding GPL violations
19:06:13
gan:skhron.org:
They're based in China, so nothing come to them
19:06:17
Cindy:
it's really REALLY hard and resource-consuming
19:06:30
Cindy:
so they only do that stuff when they suspect something's up
19:06:40
ravfx:xmr.mx:
non, that's another one I think.
19:06:40
ravfx:xmr.mx:
ffmpeg was Rockchip
19:07:06
ravfx:xmr.mx:
github even removed there repo because of the violation
19:07:28
gan:skhron.org:
@ravfx:xmr.mx: https://linux-sunxi.org/GPL_Violations
19:07:46
ravfx:xmr.mx:
will check later maybe lol
19:07:48
ravfx:xmr.mx:
https://mrelay.p2pool.observer/m/xmr.mx/oHZrFRwOxjhFBxkPxyflRmrX.png (clipboard.png)
19:07:57
gan:skhron.org:
Lovely
19:09:17
basses:matrix.org:
is it possible to change display animation on anker charger displays?
19:09:52
basses:matrix.org:
I know it is possible, but surprised to see no one attempted to do while searching online
19:10:16
basses:matrix.org:
given how anker chargers are very popular
19:25:13
bikrambiswas:matrix.org:
@ravfx:xmr.mx: How ?
19:25:24
bikrambiswas:matrix.org:
They blocked you or your IP ?
19:46:00
ravfx:xmr.mx:
some vpn i'm using ip yeah, normal thing
19:46:45
ravfx:xmr.mx:
Have to rebuild my private vpn shenanigans (My old setup went to unusable once the western world decided to block russian IP for the lol)
19:47:18
ravfx:xmr.mx:
so now i'm using public vpn shit that always endup being blocked from some ressources
19:50:14
gan:skhron.org:
@ravfx:xmr.mx: I use a Russian IP by the way, on the commercial IP as well
19:50:46
ravfx:xmr.mx:
That's what I used to use until people decided to block from there cloudflare
19:51:26
gan:skhron.org:
The commercial hosters don't have DPI
19:51:50
gan:skhron.org:
I'm not sure if even Yandex DNS obeys the nation-wide DNS hijacks
19:52:00
gan:skhron.org:
that's probably the job of an ISP
19:52:06
ravfx:xmr.mx:
Yeah, I never got blocked because of russian when using there commercial stuff
19:56:05
ravfx:xmr.mx:
@gan:skhron.org How is it from your commerical internet, when you access western/euro site?
19:56:06
ravfx:xmr.mx:
Do people got russia fatigue and stopped blocked them for fun?
19:58:52
DataHoarder:
yandex DNS offers dnssec and DoT/DoH afaik
19:58:54
DataHoarder:
so they'd have to do that too
20:00:08
gan:skhron.org:
@ravfx:xmr.mx: I could access everything freely, including "terroristic" websites like the Anarchist Black Cross or Autonomous Action, however I do think that Russians are getting fatigued, especially the non-technicals
20:01:04
gan:skhron.org:
DataHoarder: Even funnier, considering it's possible to bypass allowlists as well with DNSTT or iodine, if somebody doesn't know the proper hosters for that yet
20:01:21
ravfx:xmr.mx:
@gan:skhron.org: I mean, are the westerner are tired and are fogetting about "must block russian from infra"
20:01:59
gan:skhron.org:
@ravfx:xmr.mx: Haven't experienced that to be honest, but I don't use stuff like fiverr, paypal, and so on
20:02:37
gan:skhron.org:
The Russian state itself wants to cut off the cables to note
20:08:10
bikrambiswas:matrix.org:
I can share with you my seed phrase if you need it..Of my Nym VPN. Literally 2 people can use it . > <@ravfx:xmr.mx> so now i'm using public vpn shit that always endup being blocked from some ressources
20:09:07
bikrambiswas:matrix.org:
You should go for IP of Scandinavian nations
20:10:49
bikrambiswas:matrix.org:
https://mrelay.p2pool.observer/m/matrix.org/rFHBqAwgYMornMmqdgvbhJOA.png (1000163711.png)
20:11:03
bikrambiswas:matrix.org:
If you use IP..try to use these countries
20:56:12
basses:matrix.org:
MiniMax M2.1 ? > <@ravfx:xmr.mx> Any of you vibe coder could recommand other model than GLM 4.7 for local ai?
20:57:08
ravfx:xmr.mx:
@basses:matrix.org: For coding?
20:57:12
basses:matrix.org:
use https://github.com/TrustTunnel/TrustTunnel > <@ravfx:xmr.mx> Have to rebuild my private vpn shenanigans (My old setup went to unusable once the western world decided to block russian IP for the lol)
20:57:48
ravfx:xmr.mx:
I do have local minimax here already
20:58:33
basses:matrix.org:
@ravfx:xmr.mx: minimal tasks, (this was recommendation from someone, I didnt try myself)
20:58:48
ravfx:xmr.mx:
@basses:matrix.org: It have to work with routers too.
20:58:48
ravfx:xmr.mx:
Will look at it
20:59:12
ravfx:xmr.mx:
@basses:matrix.org: YEah, for a llm to talk shit it actually, is quite decent
20:59:30
ravfx:xmr.mx:
Fast enough with 128GB vram (the rest in ram overload)
21:52:15
basses:matrix.org:
https://techcrunch.com/2026/01/23/microsoft-gave-fbi-a-set-of-bitlocker-encryption-keys-to-unlock-suspects-laptops-reports/
21:52:16
basses:matrix.org:
LOL
21:57:30
Cindy:
"But, by default, BitLocker recovery keys are uploaded to Microsoft’s cloud"
21:57:40
Cindy:
wow that makes encryption completely fucking useless
21:58:12
Cindy:
whoops i forgot how to decrypt my hard drive, please daddy microsoft give me the recovery key
21:59:10
basses:matrix.org:
Cindy: this has always been a thing btw