/int/ – No shittings during wörktime
„There is no place like home“

Currently at Radio Ernstiwan:

Hail Odin! by Christenklatscher666


File (max. 4)
Return to
  • Allowed file extensions (max. size 25 MB or specified)
    Images:  BMP, GIF, JPG, PNG, PSD   Videos:  FLV, MP4, WEBM  
    Archives:  7Z, RAR, ZIP   Audio:  FLAC, MP3, OGG, OPUS  
    Documents:  DJVU (50 MB), EPUB, MOBI, PDF (50 MB)  
  • Please read the Rules before posting.
  • Make sure you are familiar with the Guide to Anonymous Posting.

No. 67702
229 kB, 625 × 937
Hardware, software, tech news, programming, operating systems, retro computing, we've got it all.

last one >>56447
No. 67704
After reading this article https://www.alicemaz.com/writing/program.html & talking to a friend who works as a freelance software dev I'm thinking about getting back into programming myself. For context I think I have an okay theoretical foundation and some practice in coding from university (as well as some formal credentials) but I never really completed what one'd consider a proper programming project or did it professionally. I kinda underestimated how much money can be made and it also seems there's quite some flexibility and many part-time positions even if you're not flat-out freelancing.
So I'm thinking about either brushing up my Python or learning JavaScript and then looking for some entry level part-time position. Can Ernst recommend any tips/resources to get started?
No. 67712
36 kB, 636 × 478
I started here:


Didn't progress to coding pro, but got a grasp of the basics by working through those modules.
After that, I worked through these, which were good for demonstrating how the small bits tie together into functional projects:

Build 15 JavaScript Projects - Vanilla JavaScript Course
No. 67716
My Roccat Savu mouse seems to be about to die.
It's just something over five years old, but no more than seven or eight.
I had already openend it to see if I could clean the sensors and contacts etc., but it didn't help.
Now I am looking for a new mouse that has
>a good mousewheel with click
>two thumb buttons
Anything else doesn't matter, as long as it is sturdy. And please don't talk about what mice you have been using for over ten years, I want something that is still sold.
No. 67728
> It's just something over five years old
I have to replace my mouse at least once a year, no matter if expensive or cheap. That's why I stopped buying expensive stuff.

I might be clicking too hard.
No. 67777
210 kB, 1318 × 990
> Roccat Savu
I was about to shittalk it since I had remembered it not lasting very long, but checking the dates when I bought it, it actually lasted me about ~4-5 years until the the unintentional double-clicking got too annoying. I'm now using Logitech G203 which works fine for me (and fits your description afaict), since I don't really game a lot anymore I hope it'll last me a bit longer.
It has a kinda annoying complementary software ("Logitech G HUB") but looking it up now apparently they also released a tool called "onboard memory manager" so you can adjust settings without having to run that shitty software in the background all the time.
No. 67778
Thanks, I will check this out.
I don't even know why I bought the Savu; I think a friend recommended it to me.
I am not into gaym0r stuff and am glad when I don't have any rgb bullshit on my gear, but at least it could be turned off.
No. 67781

Are you guys serious? I'm still using my Logitech G9, which I bought when it was new and Wikipedia [1] informs me that this was around 2007. So that's 14-15 years now that I'm using this one mouse.

[1] https://en.wikipedia.org/wiki/List_of_Logitech_products#Gaming_mice
No. 67783 Kontra
Hätte ich es nicht noch EXPLIZIT erwähnt...
No. 67790
>I might be clicking too hard.
Much probably! Wow, once a year! What do you do with it?
No. 67791
> What do you do with it?
Browsing the internet. Seriously, I have no idea. Maybe it's because I'm left-handed and mouse makers save money by using weaker plastic on the right mouse button?
No. 67799
20 kB, 500 × 409
Check if there is wear in the part of the mouse you click on. Most keyboards have wear from usage (brutal example in pic). Regarding mice, it depends on the surface of the mouse, though. As an example, my 2nd hand Asus mouse with el cheapo plastic, looks fine. My Mars Gaming mouse, with a layer that gives it more friction... Well, it's worn.
No. 67803
From the thumbnail I thought there were traces of poop on the keys.
No. 67806 Kontra
This was a quintessentially German post.
No. 67807
I knew someone would comment on that. Did it make you feel smug?
No. 67808 Kontra
No. 67809
Well, happy to help.
No. 67931
Can somebody recommend a good and especially robust hard drive for backups?
What I want to do is just having something I can connect to my computer, backup everything I want to backup, and put it somewhere safe. I don't want any cloud or online bullshit, just a drive I can save my stuff on.
Should be several TB and not need an external power source. It also doesn't need to be an external drive, I can always build a case or just plug it into the mainboard directly.
No. 67938


Also at least two is needed if serious.
No. 71970
88 kB, 908 × 450
Hello computer expert persons, i have a question.

I just updated the client for my VPN (NordVPN because i got a 3 year subscription for free) and now i can't connect to certain websites.

90% of the VPNs use is to avoid IP related timeouts from filesharing sites when downloading porn.
It worked without a problem before but now i can for example connect to google but not wikipedia (or any filesharing site i tried).

Why could that be?
No. 71971 Kontra
1,6 MB, 1480 × 1080
>downloading p*rn
That aside, idk what's the reason for your problem, but why not use torrents (e.g. p*rnolab) if you have VPN anyways, surely that's faster and more convenient than DDL sites
No. 71974 Kontra
>why not use torrents (e.g. p*rnolab) if you have VPN anyways, surely that's faster and more convenient than DDL sites

I have accounts for one or two private trackers but especially for niche stuff there's far more interesting stuff to find in other places.
I didn't torrent since the problem occured but i could imagine that that stopped working as well. I might check it out later.
No. 71975 Kontra
>not having a well curated stash in case there is no internet or they take down that video you like
No. 71983 Kontra
>niche stuff
mein gott...
No. 71990
Some of the exit gateways of your VPN service may have been used to deface sites or spread spam, landing those IPs on blacklists that are shared by various hosts. Possibly wikipedia and your DL sites make use of the same blacklists.

Have you tried switching to different exit gateways? Most VPN providers support that.
No. 71992
I've just seen this short video showing how heat-sinks are manufactured and thought it surprising enough to share with fellow ernsts:

No. 72017
668 kB, 1378 × 2000
Thank you all for trying to help me or wondering why i download porn (4k high with high bitrate is hard to find on streaming sites. Also the very weird stuff),

It seems NordVPN has a new "feature" which i wasn't aware of. I found a new network in my settings which my connection was routet trough. This network is called "NordLynx" and is supposed to make the VPN faster. For me it just blocked a lot of sites. Switching to OpenVPN in the settings solved the problem.
I don't like things like that happening at all.
No. 72795
Finally took time to enable dns-over-https on local proxy server. EC experience is smoother now.
I might want to look into the fake-ip defined in rfc3089.
No. 72818
64 kB, 450 × 540
all this complicated stuff you are mentioning in order to get into EC or...
No. 72821 Kontra
It's not complicated. I was just too lazy to configure it. EC isn't blocked here, but it helps that I forward some of my outbound dns queries to a DoH proxy based on their GeoIP.
No. 74451
57 kB, 640 × 360
Does anyone use a personal wiki? There are so many different wiki platforms it feels like an impossible task to pick one. I'm looking for something that will be hosted on a server and has a fully featured in-browser editor and supports images and videos, and would be nice if you're able attach arbitrary files to pages.

At the moment I'm thinking of using MediaWiki that's used with Wikipedia itself as it probably has had the most development and use, and does feature a nice in-browser editor and supports at least images and videos. Though it definitely feels like an overkill in terms of the scale it's designed for so maybe there is a simpler alternative.

Software like Notion and the like are just somehow so unwieldy with their stupid goddamn apps that always have some syncing errors and their own quirks regarding everything. Usually they support only text and some basic image formats, which isn't enough. I just want to offload some of my work and hobby stuff somewhere that's well searchable, available and editable on every goddamn device with an internet connection without any additional software or hoops to jump through.
No. 74495
7 kB, 225 × 225
You know what I just noticed.
On my phone, I have notification sounds turned up higher than in-call audio, so whenever I get a notification during a call, Android turns up the global audio volume, and when the notification sound is finished playing, turns it down again.

Linux doesn't have proper audio mixing even under Android.
No. 74636
110 kB, 640 × 397
98 kB, 602 × 640
With the advent of AI text generators that are so good you can't tell them from humans, people are having renewed debates on sentience, sapience, consciousness, theory of mind, etc.
I wonder if the people who were writing about these things back in the days even suspected how soon this issue would become practically relevant.

But, I'd also like to ask the opposite question. Instead of asking "is GPT-3 acting like a human would", why not ask "do people act like GPT-3 would".
I think most of us have experienced, and also engaged in, speech that was constructed in a way that displayed knowledge of the language of a particular field or topic, but do not synthesize ideas as they speak, but merely construct believable and parsable statements.
I mean, I had a job interview a while back. I've also written some bullshit essays for school. I've also engaged in countless internet arguments where I had only some vague idea of a topic and knew some terminology, and some opinions I read elsewhere, and made posts and responses using them.
There's also people who do "believable statement generation" as a job.

I guess my point is that people talk about "consciousness" as if it was a constant property of the human mind, but humans aren't always internally conscious, even when awake. It's more of a neurological mode or state that we sometimes find ourselves in, one among many other modes and states that our brains go through during their operation.
It's weird that this obvious observation was so difficult for me to make.

so maybe humans shouldn't be the ultimate yardstick of "consciousness" by which we measure other things.
No. 74641 Kontra
First image is me IRL
No. 74662 Kontra
> Linux doesn't have proper audio mixing
Then just stop using plain Alsa. Pulseaudio has been working for a long time despite being written by he who shall not be named. Pipewire solved the remaining issues. At least for sane hardware.


No. 74677
13 kB, 474 × 474
>play linked video
>audio and video desynchronized
lmao video production on linux
No. 74680
Yea I know right. Linux people are stupid.
No. 74714
I just wanted to drink from a new bottle of Dr Pepper, which kinda exploded upon opening despite not being moved for days.

Why am I posting it in the computer thread?

Because some drops landed on my keyboard and mouse. I immediately turned of the PC (and pulled the plugs) before cleaning everything.

So far it still seems to work, I'm using it to write this. Will the drops I missed kill my hardware later?
No. 74715

Can Ernst tell me something about it? Is it useful for absolute beginners? I wonder if it might be useful to learn coding over a longer span (lets say 1-2 years). Just to get a grip on how coding works. I don't want to become a programmer but at least get a feel for coding. Maybe it's a fruitless idea to learn stuff on the side but I don't want to restrict my knowledge to humanities, which already takes a lot of time though.
No. 74716
50 kB, 1280 × 266
Consciousness (presence of inner observer) is not equivalent of sapience (ability of reasoning) despite they can be connected somehow. I'm sure that animals are consciousness though silly. And advanced AI would have totally different inner experience from human.

Generating believable text does not require reasoning, but artificial neural networks seem to be able to do even it last years. Already posted in other tech thread:

I see this site for the first time, but in general such websites with exercises and automatic checking system are great for your goals. Can advise https://www.hackerrank.com/ too.
No. 74717 Kontra
Came across excercism via this page

Don't know how I came about that page though.
No. 74882 Kontra
How to start programming.

Step zero find a problem you want to solve. Draw fractals, download insane amount of porn, scrape image boards, whatever.

Step one don't look at any "teach your self to program" they are shit as step one goes and mostly deals with made up topics that isn't really interesting.

Stop two pick a popular language. Go, Python, Ruby maybe Javascript or what ever kids today use. Make sure it has addons/packages, or what the language calls it, so you don't have to write an HTTP client as your first task for example. The language you want is partly based on the problem you want to solve. You want to do maths, Python with Jupyther might be what you want for example.

Step three start coding. Now could be a good idea to start looking at the tutorials as a reference or maybe they are teaching exactly what you are after. But now you have limited the scope and can choose better.

Stop four, realize shit takes time and is boring. Overcoming this is the hard part.
No. 75359
514 kB, 1051 × 638
151 kB, 1375 × 720
276 kB, 2048 × 1152
Some may remember the fire in the OVH datacentre in Alsace last year (because lots of websites had outages). Well, the report about what happened is out and it was a fauly cooling system that was leaking water into a UPS (backup power supply):

No. 75360 Kontra
Wasn't this the data center with the Encrochat data and didn't it burn down right when investigations took place so people suspected it to be an intentional fire?
No. 75655
Oh boy, this is one delicious attack:


>Intel’s security advisory states that all Intel processors are affected.
>AMD’s security advisory states that several of their desktop, mobile and server processors are affected.
>To our knowledge, Intel and AMD do not plan to deploy any microcode patches to mitigate Hertzbleed.

What does it do?

Well... it's an attack that extracts information about the numbers that are currently computed by the CPU (ie. everything) based on the frequency changes (ie. how many Hz any thread is running at).

But unlike several older attacks, it's not looking at the power frequency in the wall-socket, which only a local attacker can do...

No, it's somehow inferring the CPUs power frequency from the interference caused by the electromagnetic waves sent by the CPU on the ethernet output. So this is a remote attack, apparently possible even through the internet, but I don't quite understand how any traces of those interferences survive the first (or any other) switch or router...

Oh and the juiciest bit: This circumvents time-constant post-quantum cryptography thought to be safe from side channel attacks. Let me re-phrase: This "breaks" not only the cryptography we are all using today but also the cryptography that is meant to be secure from attacks by hypothetical quantum computers in the future.

tl;dr: An attacker can read what numbers your CPU is working on all the way from the internet for all software on all operating systems on most hardware. This thing is not theoretical, it's been proven practically and has now been disclosed. No patches in sight (from the hardware vendors).

I can't really judge how feasable it is to carry out this attack right now. The limitations aren't clear to me after skimming the paper.
No. 75656 Kontra
I barely understand, what are the resources and efforts necessary to do this? High or low?
No. 75657 Kontra
The concept of "workload must run for long enough to trigger frequency scaling" might be easy enough to eliminate. Frequency scaling can probably be toggled on or off completely, with some tradeoffs of course. Seems like mostly another problem for cloud computing. I doubt that JavaScript for example could be used as a vector since the browsers have reduced timer accuracy since the previous round of this sort of shit.
No. 75670
Possibly the best explanation of the stack on the internet. This should be shown to every student right after the chapter on functions.

I'm starting to think that learning the basics of EE and a bit of simple assembly is actually quite important to learning how to program, even if you want to stay on the software side.
All of those seemingly obtuse conventions and rules that you first encounter when learning to program, that you are simply presented with on a "that's just how things are" basis, actually make perfect sense when you understand WHY they are there.
I have trouble remembering things unless I have a mental model for them, and just memorizing the caveats and moving on never worked for me.

Another great video by the same guy.
No. 75682
> This "breaks" not only the cryptography we are all using today but also the cryptography that is meant to be secure from attacks by hypothetical quantum computers in the future.
No. This is side channel attack, it doesn't break the crypto, it circumvents it.

Frequency can be turned off. Increased power consumption is a downside among others.

> I doubt that JavaScript...
Don't. Assume it can. Specter and meltdown could be exploited using a browser.

Well time to boot the Raspberry Pi.
No. 75683 Kontra
>Specter and meltdown could be exploited using a browser.
Yes, exactly why I pointed out that browsers made sure that JS timers used in those attacks aren't accurate enough anymore. Could be wrong, but that's what I recall reading. Though personally I use NoScript and block all sites from running scripts anyways. Still, have to enable scripts for some sites that won't work without them, but that's mostly youtube, online stores and banking.
No. 75685 Kontra
>have to enable scripts for some sites that won't work without them
Isn't that where JShelter comes to the rescue? I haven't read this thread and don't know computers, but one of the things JShelter does is it rounds times, which may be related to whatever you geeks are discussing in this thread.
No. 75693
>No. This is side channel attack, it doesn't break the crypto, it circumvents it.

This is why I put "breaks" in quotes (and called it a circumvention in the previous sentence).

The point of this attack is that it doesn't measure the power consumption or the frequency based on any information available on the system itself, but instead analyses the side-effects that frequency changes of the CPU have on the network interface.

I would doubt that the existing restrictions in js engines of browsers mitigate this attack.
No. 75925
57 kB, 524 × 561
Finally set up that personal Wiki and after using it for a bit I'm incredibly pleased. Let's see what the final verdict is after some months.
No. 75926
Sauna Mämmi
No. 75927 Kontra
kittos Suomi
No. 75928 Kontra
kiitos Suomi
No. 75936 Kontra
1,8 MB, 811 × 1147
kiitos Suomi
No. 75941
408 kB, 1080 × 1194
What conclusions can we draw from this?
No. 75942
1,0 MB, 768 × 924
That you should feed it with a whole sentence next time.
No. 75947 Kontra
1,1 MB, 768 × 923
Mysterious soul
No. 75955 Kontra
1,2 MB, 768 × 923
1,5 MB, 768 × 923
1,3 MB, 768 × 923
1,4 MB, 768 × 923
Truly, the intellectual and spiritual avant-garde of the Internet.
No. 75967
633 kB, 768 × 924
Literally me.
No. 75968
722 kB, 768 × 924
No. 75969
109 kB, 709 × 770
No, that's literally you.
No. 75970
1,1 MB, 768 × 924
After posting >>75968 I had to try something. Sadly, or maybe thank doge, the program is not self-aware enough to know its results are presented in a three by three grid.
No. 75971
59 kB, 650 × 699
846 kB, 768 × 924
Picture 1: What I thought of when I typed
Picture 2: The dalle result
No. 75977 Kontra
We can tell from these that most post here are done by closet homsexual russian bots.
No. 75980
How can a bot be homosexual for a building?
No. 75981
575 kB, 518 × 673
No. 75982
601 kB, 565 × 734
That bot has heavy Bacon influences yay
Unrelated but I had to say it: Google sentinent AI is wankery IMHO

My attempt pic: well it's kinda biutiful

Gonglusions from photos: well Internet is still very weaboo. And Bacon. Well Bacon is very human
No. 75983 Kontra
472 kB, 1080 × 1237
No. 75984
Belarus is same Gaudí as Catalonia, so the trencadís is not Gaudí.

Bot does know we have a star sometimes, although I doubt it knows its meaning. Perhaps the cretin Google AI knows.

t. super AI expert
No. 75985
778 kB, 768 × 924
I tried to describe the picture in the OP to DALL-E. Both "keys" and "organs" having multiple meanings doesn't help matters.
No. 75986 Kontra
Definitive proof that AI is nowhere close to achieving even human level intelligence, let alone surpassing it.

btw an*me watchers aren't human
No. 75992
Dall-E is not from google but from OpenAI.
No. 75993
>btw an*me watchers aren't human
btw asterisk users are total dumbasses
No. 76033 Kontra
60 kB, 620 × 395
> rare Belarusian ball
No. 76041
212 kB, 1411 × 1418
I know, thanks, please excuse my ramblings. It's that I knew about that AI from Google that told wankery things and I was comparing it with the one from this OpenAI you mention. I like more the OpenAI one. The Google one reeks of sci-fi nerd worship IMHO.
No. 76046
B*tter a d*mbass th*n an an*me w*tcher
Wh*t's y*ur v*ndetta ag*inst ast*risks anyw*y?
H*ard y*u c*mplain ab*ut th*m m*ltiple t*mes alr*ady.

The above text was generated using the following regex:


I invite everyone to start using it to asteriskize their posts. It's fun for the whole family!
Here's an example:
M*ry h*d a l*ttle l*mb,
Its fl*ece w*s wh*te as sn*w,
And ev*ry wh*re th*t M*ry w*nt
The l*mb w*s s*re to go ;
He f*llowed h*r to sch*ol one d*y—
Th*t w*s ag*inst the r*le,
It m*de the ch*ldren l*ugh and pl*y,
To s*e a l*mb at sch*ol.

W*nderful, isn't it?
No. 76047 Kontra
Go **** yourself.
No. 76063 Kontra
Not sure if taking the piss or being serious
No. 76088
56 kB, 859 × 498

It's honestly exciting to see what this thing can do. Take the conversation in the attached image: To me it's indestinguishable from what a run-of-the-mill bulshitter would write. OpenAI is now succesfully emulating reddit posters. Well, at least some of the time - this is just a selected snipped and not representative of how well the AI chats 99% of the time.
No. 76090
34 kB, 1024 × 576
I like this AI's style.
No. 76099
It has longer attention span than 90% of internet users.
No. 76130
Wait, so who is the AI here, the question asker or the answerer?
No. 76133
I can't wait to see how this develops further. It's amazing to think about the potential applications for something like this.

It has a very humanlike quality to it.

It is able to understand and react to complex questions.
This makes it the perfect tool for businesses and organizations who want to get the most out of their online customer interactions.

There is no AI here. This is just a regular conversation.
No. 76136
The text marked in green is from the AI.
No. 76138 Kontra
8 kB, 245 × 279
Ernst, I was mocking that poster, it was not a serious question.
Pic related, it's you.
No. 76146
More like "I made a dumb-dumb and now I have to blame someone else for my mistake"
No. 76148 Kontra
That is a marvelous projection. I wonder which of you I've debated in the past if it's not a monolog anyway.

t. another German
No. 76153
How can I know you are not an AI?
No. 76155 Kontra
198 kB, 1387 × 1468
An AI wouldn't post this picture.
No. 76170
Looks AI generated, so why would an AI not post it?
No. 76179 Kontra
>Looks AI generated

Maybe you are the AI and a lazy one on top! Perhaps I chose to upload that picture because it made this impression (and because I found it only lately and downloaded it, as so many others that find a way in my posts).

it's a human made picture if you would have done your research
No. 76183
55 kB, 580 × 346
Now this AI is totally neat.

Open source for victory.
No. 76191 Kontra
Man I am still not convinced. On the one hand, you clearly don't understand irony or any kind of banter (although you're operating under a germanball, so this might not be as strong an indicator), but on the other hand I sense a bit of irritation at me presuming you're not human, although, frankly, that's exactly what I would expect of an AI.
No. 76239 Kontra
Yes, he's AI (assburger irritated).
No. 76251 Kontra
I think you are mistaken in thinking that I was seriously trying to prove to you if I am an AI or not. You not being able to read me makes you an autist in my eyes. I guess we end up in the situation that is the German Mirror or otherwise known as assburger stalemate Germans accusing each other of being an assburger
No. 76263 Kontra
Thanks for proving me right, Ernst-AI.
No. 76749

Another topic I was one day hoping to research deeply and then write an essay about or something, that it turns out has already been discussed by someone way smarter than me a long time ago, and now presented and explained by someone way smarter than me.

The key insight here is that models, protocols, architectures, organizations, etc. etc., are not things that help us understand things, but quite the opposite: they are tools that help us work with something we don't or can't fully understand.

It's quite insidious how models, especially good models, by making things easier for us, trick us into thinking that we have improved our ability, when the truth is that we have dumbed the problem down enough to be within our ability. That's not to say that models are bad, just that it's important not to confuse models with reality.

All of this is quite obvious in hindsight, but it didn't click for me fully until I watched this video.