/tech/ - Technology

GPU but the G stands for Gay

Machine learning specific processors have been around for a long while. Tensor processing units. Mostly good for regression, running cnns, rnns.

>>31631
yeah intel went really hard with their neural sticks, google had their tensor cores, apple had their own name for it too, mostly for computer vision, when AI was still cool in the eyes of consumers.

Excellent news kiddo
https://www.tomshardware.com/tech-industry/quantum-computing/new-chinese-optical-quantum-chip-allegedly-1-000x-faster-than-nvidia-gpus-for-processing-ai-workloads-but-yields-are-low
Prices on those babies should now eventually come down in a short time, around two years. You can get what is essentually a devkit for your forthcoming vidya console from orangepi starting at, too much for a mass console … for now

Eventually a lowmem board won't cost too much at which point have fun

>>31727
ps. Softbank just dumped NVDIA
Those who know know

Office hours I guess, any questions?

>>31578
I feel you brother, I go walking and camping alone in the woods and mountains on the regular, and it's great because you're just there, one foot in front of the other, taking a break when you feel like it, getting dirty,…

I feel like I've become increasingly Luddite over the last years, like, I'm not stupid. Loads of technological development is good ofcourse, but we've become so alienated from literally everything, all life is now nothing more then passive media consumption. No experiences, no hardships, no relations, no fun. Just stare and stare at that fucking screen in your hands.

The only reason I'm still working in tech now is so I can get money to buy some abandoned orchard or something, because modern society is absolutely crushing my soul and the only time when I had any semblance of feeling alive was when I was being a fucking hobo.

In SWE it's blatantly obvious. AI pretty much has to remain the primary, and will eventually be the sole, mode of software development because SWEs can't code anymore. They have, through sheer laziness, lost the ability to program a computer.

I have very little respect for SWEs, they always had the easiest job and absorbed the greatest surplesses. So their field is gone already. They simply cannot do it and will never be able to do it again.

>>27264
It's very good at two things: writing slop reports, and writing slop computer software. So rank-and-file reporters, who do not produce creative work, and software engineers, who also do not produce anything worth thinking twice about, are the most impressed and most impacted by the ability to automate their jobs. Unfortunately, these two have an outside impact on the overall evaluation of the technology.

Unfortunately, this won't be able to be corrected in the future. The children who are learning today are being totally annihilated by AI. They are not simply one or two years remedial. They are basically incapable, full stop.

>>31219
It doesn't matter if it gets anything wrong. People think people care if things are wrong. Actaully, they don't. It just looks like it works, in mostly works, whatever. The idea that people want high quality stuff that is carefully made by an attentive and diligent master of their craft is not true. They want MORE things. The MORE the BETTER.

Really. Nobody even cares if anything actually works anymore. As long as it is apparently there and apparently works, just plow through and make more $$$. Slop world is not a meme.

>>31712
>It doesn't matter if it gets anything wrong.
if your job is stupid bullshit, sure. if you're like an accountant or handle finances in any way, which is historically the profession that lead the rise of PMCs (it's in the name!), then your boss will get really mad if your numbers are off.

https://android-developers.googleblog.com/2025/11/android-developer-verification-early.html
>While security is crucial, we’ve also heard from developers and power users who have a higher risk tolerance and want the ability to download unverified apps.
>Based on this feedback and our ongoing conversations with the community, we are building a new advanced flow that allows experienced users to accept the risks of installing software that isn't verified. We are designing this flow specifically to resist coercion, ensuring that users aren't tricked into bypassing these safety checks while under pressure from a scammer. It will also include clear warnings to ensure users fully understand the risks involved, but ultimately, it puts the choice in their hands. We are gathering early feedback on the design of this feature now and will share more details in the coming months.
Looks like they might give you a way to get around it, we will see how it will work in practice.

Has anyone attempted to make a phone operating system on BSD?

>>31713
most people run BSD desktops either because of the server features on FreeBSD or some variation of it being a trve vnix. those aren't huge consideration for a phone though and linux userspace phones presently fill every other niche, so no.

>>31713
Not just attempted, the Sidekick LX 2009 used NetBSD as its kernel: https://mastodon.sdf.org/@netbsd/109338988718834042

>>31715
>@pfr There's videos of the UI on YouTube. It's java-based. Had some cool features, like you could zoom around web pages, since nothing was mobile-optimized yet.
so it's just a forgotten android precursor then.

>>31645
what if videos are ephemeral, and also hosting uses some p2p to alleviate costs, if you wish to keep an archive, you keep an archive yourself, only the latest videos are hosted on the p2p network, this way the trendiest videos are the less costly to host

>>31667
I mean, compared to an html and an audio track a guy talking over a slideshow is wasteful, but among video types it compresses extremely well and you could host it on something like vimeo without problems or even your own site. Client-side rendering only seems realistic in the vtuber case, where you have reusable, high-definition textures, or full-blown machinima if that wasn't dead.

This seems to be built on Pixelfed (which was more of an Instagram style federated platform) so its good that its open source and federated. It seems to be making progress for interoperability at least according to that blog, such as the way it structures a video-centric platform in a way that will work with non-video centric alternatives (ie microblogging and photo-style tech like Mastodon or Pixelfed). Now the big thing they need to do if they want some actual fucking progress is to
>Not allow overzealous moderation by instance owners; leave it up to individual users.
One of the biggest problems I've had with the 'fediverse' is that it becomes a fractious culture war battleground with nodes/servers run by the most obnoxious people who you can imagine, perpetually offended. This means that, for most 'less technical' users - anyone who isn't running their own entire server that is - their ability to communicate with others comes down to what their admin deems worthy of federating with. If I am on Instance A, and you are on Instance B, yet instance A's admin decided that B is full of degenerates or nazis or fascists or communists or pedos or whatever they find problematic, I can't talk to you. This was always one of the weaknesses of not the fediverse itself but the individual projects within. THe "maximum" amount of defederation should be "nothing from your server is showed on my server's public/universe etc..feeds" which would still leave individual users the chance to make their own choices about what to do, where they want to go, and who they wish to connect with. Loops coding this in from the start would be a help, especially considering that it appears their "instance inbox" thing suggests that all the federated messages go through one server side pipe as it were which is more of a hole for both security and overzealous moderation.

>>31665
PeerTube is the best alternative so far and handles this situation by far. It allows both instances to do some degree of the heavy lifting, but if users are watching the same thing there's p2p webtorrent. As it grows the costs go down both with users and servers - its the only way to really handle something like this…along with usage of better codecs for bandwidth.

Post too long. Click here to view the full text.

>Instead of being trapped in walled gardens, your content, your followers, and your identity can move freely across the open web.
I've always felt TikTok is bad for this reason
>>31657
>format cancer
"Zionists need to buy TikTok for peoples own good, we need harm reduction anti-fascism" - cancerous socialists

To the people suggesting offloading video hosting to peers, this is mobile-centric short-form content. I'm guessing you understand neither the technology you're talking about nor how people use Tiktok.

>>31651
>coming up with some kind of parasitic network signaling mechanism clients can exploit to discover each other by (mis)using various public-facing internet services
That avenue of research seems intriguing. I've thought about implementing a messageboard in js on top of a static-page before, which is theoretically possibly with WebRTC, but doesn't allow natively for any discovery mechanism to my knowledge. Connecting people without centralized rendezvous-points and beyond exchanging addresses with your friends, phonebook-style, appears to be a hard problem in the current network.

>>31652
>Can you link/post the paper? Please, I have thought about this sometimes too

I can't find the specific whitepaper I remember reading before, it might be locked behind paywalls now, but the concept is known as a "DNS dead drop" or a "dead drop resolver" and it's often used for covert C&C in botnet and malware networks. DNS is like the ancient legacy backbone that the Internet is forever stuck with and it's a UDP-based protocol with no state or error checking of any kind and there's a lot of nefarious stuff you can do with it. Here is a PDF I found which goes over the general vulnerabilities of DNS as a protocol and some of the ways it can be misused for this sort of thing:

https://blackhat.com/presentations/bh-europe-05/BH_EU_05-Kaminsky.pdf

>>31668
>I might be dumb but I never understand what's the point of messaging apps. How is that better than social media with DMs?

The client-server model works for sending text messages and sharing pictures and videos and stuff like that, but it doesn't work for live video/audio streaming. It is not possible for Discord's servers to handle millions of simultaneous live video/audio streams between all of the Discord users of the world; for video/audio calling, Discord merely functions as a signaling server and it uses WebRTC to create a direct peer-to-peer connections between users for the actual video/audio stream.

>And if you want P2P messaging, why not just self host ?

How does anyone find you? How do they connect to you? Are you behind a NAT? Can you forward ports on your router? Do you have a static IP address or a domain name? Is your device a computer on a network that you control or is it a locked-down smartphone that doesn't let you host anything?
Post too long. Click here to view the full text.

See also: https://en.wikipedia.org/wiki/Domain_generation_algorithm

Keep in mind, security experts struggled to combat this sort of thing long before the days of LLMs; imagine what could be done nowadays, where you could algorithmically generate domains that are completely impossible to distinguish from legitimate domains.

>>31671
>I've thought about implementing a messageboard in js on top of a static-page before, which is theoretically possibly with WebRTC, but doesn't allow natively for any discovery mechanism to my knowledge. Connecting people without centralized rendezvous-points and beyond exchanging addresses with your friends, phonebook-style, appears to be a hard problem in the current network.

You can use a pre decided fake infohash as a shared password and exploit the bittorrent DHT like this board does. uvl4xhrtupoygytfikwttev5pl2n2lyfajbedy5b7lbwtupndd4lh6yd.onion/board/mu
The program will look like normal bittorrent peer finding activity.

>>31673
>Domain_generation_algorithm

You can easily do this too, for example sha256(infohash+month+year).

>>31674
>sha256(infohash+month+year).
Careful, SHA2 is vulnerable to length extension attacks. It's better to use SHA3 or Blake3 for this purpose.

>>29448
In my experience external HDDs have been horrible for reliability and constantly broke. But maybe that's because I was too clumsy and dropped them sometimes and didn't really know you shouldn't move them while they're plugged in

All my important files I have stored in triplicate on three old 3.5" hdd sata drives of different brands I had lying around. I figure if they are stored somewhere out of the elements at least one of them ought to still be readable 20 years from now.

You guys stressing out about pirated content you can download anywhere and you don't even give it the time of your day, meanwhile I lost all my teenage years' photos on my old phone that broke.

I got some WD Red HDD drives years ago and they're been at it just fine, I dunno why they have such a bad rep. You also should probably invest in some sort of NAS device with raid or whatever, 2 TB devices are dirt cheap these days and you can always use the peace of mind tbh

>>29278
the cloud is just someone else's hard drive

>suggestions
NixOS with niri window manager, with KDE window manager in case niri's sattelite-based window positioning borks something (i.e. Krita when using a tablet pen)
Basically install NixOS with KDE then add niri.
Niri is good for general use as it's lightweight.

>>31608
I meant hardware doofus

The EOL is not a sharp cliff.

1. You can still get updates for free with a Microsoft account.
2. Most businesses have already gotten rid of their pre-11 inventory. If you're looking for ebay deals this is where a lot of them come from.
3. A lot of people, especially sellers, install Windows 11 on incompatible hardware, even hardware that's too old to receive updates because the ISA is too outdated.

So you won't see the kind of glut in the market that you might think. That said you've been able to get a cheap T470 for years if that's what you want. It is probably cheaper than it would be if it was officially compatible with Windows 11.

>>31607
windows 10 iot ltsc is still an option. that or linux

I think windows 11 iot ltsc bypasses hardware requirements anyways so it doesnt matter

The newer thinkpads are alright, but no match for the old ones regarding price. Framework has some options if you're part of that sort of crowd.

i still daily an old t480.

>Is a self-driving car really an invention that society needs
Probably, because of points that have already been brought up. Personally I'm wary of them because if they become ubiquitous they might try to take away human-drivable cars. As someone who enjoys driving I'm not interested in that.

>>31606
Massively gay

>>31610
>massively gay
Like you? Anyways it doesn't matter be as gay as you want

Are self-driving buses gay

no. the personal motorized carriage, self driving or not, as a mode of transport and the subsequent design of infrastructure, industry and institutions around it have been a disaster for the human race.

Yes because it makes car ownership more expensive while nonetheless essential
>Self-driving tech is expensive! Cars need to cost more
>Insurance companies are unconvinced self-driving cars are safer! Insurance must cost more
>To improve self-driving systems, really the roads need to be redesigned to become "smart roads, tear them up! Gut public transportation projects to fund it
>The thing that is blocking the potential for self-driving cars is human driven cars, scrap them!
>Any one would tell you getting to the next town takes an hour, thanks to AI predicting there will be a traffic jam, you have been put on a route that will double that!
>You can't possibly service this vehicle, it's too advanced! Neither can your town mechanic, nope you're going to have to send it directly to Tesla I'm afraid
etc, etc

Generally speaking, consumer choice boils down to what brands you can buy, choice in how people use something is not ideal because naturally they will aim to use it for the least cost possible when the point of gaining market share is to lay claim to customers and more importantly their salaries.

Looks like maybe XDG-shell and libwayland-client. [^1][^2][^3]

:[^1] https://bugaevc.gitbooks.io/writing-wayland-clients/content/beyond-the-black-square/drawing.html
:[^2] https://bugaevc.gitbooks.io/writing-wayland-clients/content/beyond-the-black-square/xdg-shell.html
:[^3] https://wayland-book.com/xdg-shell-basics/example-code.html

>>31601
huh, apparently it doesn't have drawing commands like x11, you just render client-side. I guess I will just use cairo then
but man it feels like a regression, as antiquated as it was, I liked that, in theory, you could use pure x11 programs through a network and the remote application would just send you the drawing commands to be rasterized locally

>>31602
It is pretty good, even locally if you're running a browser in a chroot or container like myself.

It's because wayland is garbage, put it in the bin.

>>31586
Better yet do encode image by image, but not in a stupid way:

def render_pages(
    src_pdf: pathlib.Path,
    resolution: int,
) -> List[bytes]:
    pages = []
    with Image() as images:
        images.resolution=resolution
        images.read(filename=src_pdf)
        images.depth=8
        images.colorspace="gray"
        for image in images.sequence:
            compressed = Image(image=image)
            compressed.format = 'jpeg'
            buffer = io.BytesIO()
            compressed.save(file=buffer)
            pages.append(buffer.getvalue())
    return pages

>>31586
I'll admit that I have no idea how to use that code or what it means, but thanks for creating the script to transcribe this! There are a fair amount of errors, but not as many as ABBYY would generate - it seems like the LLM is a lot better at cutting out all the extraneous exponents and apostrophes that ABBYY would pick up from the film grain.

The downside from the LLM is that you sometimes get some true bizarre hallucinations. Like this one in the opening sentence of part IV. Here it is typed out by sight:

In the first article we introduced the reader to Comrade William English Walling, the "new" Duehring, who proposes a "new" Socialism based on "new" methods and principles.

And this is what the LLM spat out:

In the first article we introduced the reader to Conrad Williams*. [*Footnote: Not William, as given in the heading. Editors.] Dorothy, who proposes a "new" Socialism based on "new" methods and principles.

To clarify, there are no footnotes. The AI just totally made that up somehow.

>>31590
>no idea how to use that code or what it means
Well, that's a little too bad. I'm not exactly sure how to run it on Windows either. But it sounds like your existing solution is maybe good enough?

>true bizarre hallucinations
Yah, one of my other runs (maybe the one with the weaker model?) got it repeating "to my knowledge" a hundred times and monotonically decreasing the column width until it was just a word or part of one for one article.

>>31591
>But it sounds like your existing solution is maybe good enough?
I have dozens of hours of experience using it, which helps a lot. ABBYY also has a lot of PDF image editing tools built-in to deskew the text etc., so it looks like I'd be processing through ABBYY anyway before running the images through Tessaract or an LLM. Severely grainy or unevenly exposed images still give it a lot of trouble through with random apostrophes etc. that are a chore to manually remove.

I thank you again for transcribing the Slavit text for me though. It looks good enough that manually correcting it wouldn't be too bad. We'll see about that though, maybe the hallucinations will be so severe I'd have to verify every sentence.

The ultimate solution to this problem will be to get New York University to take the original crinkly newspapers out of storage and scan the broadsheets properly. Which the publication definitely deserves, but it's a huge ask and I'd like to be able finish my William English Walling complete works project (plus polemics aimed at him) before I dive into that.

>>31592
Good luck with your endeavors then.

>William English Walling
This seems like a very interesting character and project.