[ home / rules / faq ] [ overboard / sfw / alt ] [ leftypol / siberia / hobby / tech / edu / games / anime / music / draw / AKM ] [ meta / roulette ] [ cytube / git ] [ GET / ref / marx / booru / zine ]

/tech/ - Technology

"Technology reveals the active relation of man to nature" - Karl Marx
Password (For file deletion.)

Join our Matrix Chat <=> IRC: #leftypol on Rizon

| Catalog | Home

File: 1621474187293.jpeg (231.6 KB, 1200x801, 1984.jpeg)


I think I'm reaching unhealthy levels of being unable to cope with surveillance capitalism. Some random person caught my face for a few seconds during a whatsapp video chat and my day was fucking ruined, seriously.
67 posts and 5 image replies omitted. Click reply to view.


>computer predicting the future
was a miniseries already


It's funny because China has been making accurate predictions of the future for deacdes and all they needed was dialectical materialism.


and yet they couldn't predict that forcing farmers to make shitty low quality steel instead of food would starve the nation. they failed to predict that their sloppy safety standards would let slip a pandemic from their sloppy chinese labs. and yet many people predicted these things before they happened without dialectricals or whatever nonsense. they failed to predict that building houses on debt nobody can afford to live in might cause some issues. oh dear well maybe one day they will actually predict something and then your comment won't seem completely out to lunch.


They already are predicting things it's just mostly "will X buy this product at this time" so they can serve the ads to you at time you're most likely going to buy it. Whether or not these systems can predict anything on a grander scale is just a bonus to them.


They want to prevent the unpredictable, but that's part of the larger goal of maintaining the status quo. The goal is not to intervene at the level of causes but at the level of effects. Big data is inherently like that, it only spits out correlations and deliberately ignores causation. Causation is too "metaphysical", we only care about raw numbers. Why does somebody want to drive a truck into a crowd? Who cares, what matters is to detect it and manage it. This is why we're also seeing a push for everything "smart" where a set of possibilities is literally pre-programmed, again to prevent the unpredictable. Ideally you won't even be able to commit any act against the system because the "interface" won't have the option, so the system can do away with managing side-effects as well because there won't be any.

This system has no projects, no dreams, no future. It doesn't want to fundamentally change anything. It might create a spectacle of incessant change, but it's all merely an incremental improvement in how to best recuperate and manage.

File: 1633484235839.jpg (291.25 KB, 1920x1011, mastercomputer.jpg)


I believe that hardware is the most essential component of an electronic, and that as a computers task becomes increasingly convulted the type of hardware a computer needs also increases. This electrification of common tools can increase to the point where everything about a society can be maintained and controlled with a single device or a master computer. What massive steps in society must be taken to build such a contraption


decentralization is the way to go, any one central system will be vulnerable to whoever controls it or whoever hacks it or corrupts it. rather then having all the eggs in one basket: we decentralize. that seems to be the trend. the master computer is more software then hardware. it runs decentralized apps, dapps not apps. they call it the world computer: ethereum.


File: 1633536864093.png (186.89 KB, 1636x941, ClipboardImage.png)

decentralization or distribution?


your comment provokes a lot of thought but i have no answers for you. tell me more.

perhaps one issue with full distribution is that it requires all participants to have the computing power and bandwidth to run the chain. where as with nodes you don't need to personally have a computer with that level of power or that level of bandwidth to use it. however data will become cheaper, computers get more powerful. but demands on the hardware also increase.

File: 1608526236979.jpg (572.3 KB, 1800x1200, Asus_ROG_Strix_GL702ZC_9.jpg)


Just want to earn my paycheck to feed myself from something that will work for at least 4+ years and be durable in a chaotic home environment.

I heard stuff about how the new ryzen laptops are crazy good worth of money for the consumer. Don't know much about cpus wars.

Only want to be able to do my work fast on it without lagging under heavy load.

Will be using many many browser tabs, excel sheets, word documents, and pdf softwares. No games or anything. Just work.

What durable laptops would you suggest for these purposes in the $300-500 dollar range. I just want to fully utilize my purchasing power while buying this laptop because i'm kinda loaning the money from somebody else ;-;
20 posts and 4 image replies omitted. Click reply to view.


Framework laptop, if an old thinkpad doesn't cut it for you.


Would any Thinkpad do?


Seconding this. Got a T450s years back. I'm not exactly a power user, but it was a reliable daily driver for school and personal use. IMO these magnesium body thinkpads have some of the best build quality of any laptop out there. I've traveled with it,dropped it so many times, and it's at the age where laptops start falling apart but it's still going strong without any mechanical issues.

T480 is still relatively modern and is the last model with hot-swappable batteries, upgrade-able RAM, and upgrade-able SSD. The last model with socketed CPU is the T440p, but it's an older model and has an infamously poor trackpad (which can be upgraded however).


>>11938 (me)
Wanted to correct myself and say that in addition to T440p, the W540 and W541 also have socketed CPUs. However, they're becoming memepads which has seen their prices go up considerably.

summary: Classic Thinkpads (X200->X230, T400->T430, X60/61, T60/61) are becoming less budget-friendly, especially when you factor in the cost of repair (may be necessary) and maxing out the specs for modern use. If you want something to handle modern workloads and still want maximum upgradeability, then T450/T450s -> T480/T480s are your best bet.

As for buying them, avoid big retailers like Best Buy and Amazon. Ebay is usually the best secondhand market, but you can also find plenty of gems on shopgoodwill and craigslist


I bought a refurbished T450s just this year. It's the last thinkpad model that doesn't need screen scaling (1600x900 @ 14.0in = 131 PPI). Screen scaling is still utter shit on any OS. Build quality is worse than early Lenovo / late IBM, less serviceable as well, but still better than other manufacturers. I replaced the keyboard with a Lite-on one. Chiclet keyboard but better quality and has highest key travel compared to other keyboard manufacturers.

File: 1632339083981.png (227.08 KB, 1008x736, filepicker.png)


Greetings, chinlets and incels. I am a man of the name Dimitrioyivch Khaladrenis. And I have been greatly frustrated with the filepicker situation on linux as of late. Kdialog doesn't force itself on most applications save those which support it, (Only firefox) and the gtk filepicker has no thumbnails. Kdialog also has the problem of being tied to KDE so it's not very lightweight, not a sucklessfag but sometimes one gets the urge to activate ye 'ol i3 but still wants to have a file dialog with thumbnails, if you catch my drift.

Anyways, my solution to the File Picker Question is to basically code one myself. However I am also a NOCODER and don't know where to start or how GtkFileChooser and Kdialog actually work, nor do I know the process of creating a xorg window from scratch, or how necessary Qt or GTK are in that process. The finished product should be able to:
- Have thumbnails
- Let you type in the path of the file you wish to use
- Have shortcuts to the folders where you keep your photos/webms/etc
- Open instead of GtkFileChooser on all programs
- A search function for the current active directory
7 posts omitted. Click reply to view.


Yoshii from texhnolyze?


Been thinking of making it proprietary due to the fact that
>>11722 doesn't seem to think I can do it. So it's going to cost 1 ETH to buy the binary and even then I'm going to put it under a restrictive EULA.


>Why didn't Ubuntu devs fix the file picker? Because even they don't wanna do it.
Sure, but why they don't want to? Not because it's impossible but because it isn't the job of package maintainers to develop software. I don't doubt many hobbyist patches out there are garbage. But what is a file chooser anyway? It's basically a dialog version of a file browser, right? So if so many file browsers on Linux support icon view, why would it be so hard to just port that functionality to file chooser? Make them share the same thumbnail cache and settings for which directory gets displayed in the icon view. The solution is trivial, but it does require a lot of work. Not mentally challenging work but the mundane work of porting, rewriting, refactoring, testing. And for what purpose, what guarantees that their fork will have any effect on upstream or distros? Why don't the Mint devs do it? Because it would require maintaining A LOT of packages which they currently very happily delegate to the Ubuntu maintainers and for good reason. People already have enough on their plate maintaining the existing forks and patches of GNOME fuck ups, considering multiple whole new DEs and programs appeared because of GNOME devs.


The real reason is that Red Hat plans on creating a proprietary filepicker for GNOME, they will call it FILECHOOSIX and rent it out for a $1000 monthly subscription to enterprises. They have played us for fools. The only way to fight back is to abandon GTK or fork a truly libre version without corporate meddling. That's our only choice if we want to save the future of filepickers on the linux desktop.


File: 1632836390195.png (2.17 MB, 875x1196, ClipboardImage.png)

I pray to st. ignucious they would actually do that. Let one thousand forks bloom.

File: 1632644242504.png (178.56 KB, 940x198, header1.png)


>Arcan is a single-user, user-facing, networked overlay operating system.
based or cringe?


smell like a fed honeypot


Extremely based, it goes way beyond current isolation-based solutions (like Qubes) which tackle only one aspect of the problem. But if I understand it correctly you'd have to rewrite huge chunks of the user space to fully benefit from this design.

This is a good read as well, has some more interesting ideas:

LMAO, how?


Super based and the best hope the Linux desktop has post-Xorg, IMO


>But if I understand it correctly you'd have to rewrite huge chunks of the user space to fully benefit from this design.
It has Xorg and Wayland compatibility layers, though maybe not all the "Arcan features" are available with those

File: 1624038678945.png (17.74 KB, 579x382, 34567899876.png)


Hello, tech, I'm looking to develop a leftypol textboard for the gemini protocol. I am curious about where I should get started, or, what language would be optimal for this. I also don't really understand how to get one language to interact with another language. (Gem text and Perl for example) I don't really want to do anything huge or something that will blow up I am just a hobbyist looking to do something fun and interesting. thoughts?
7 posts and 1 image reply omitted. Click reply to view.



Here's another textboard script that tries to remain as close as possible to 2channel but with a database and QoL features.


>That site looks like shi and is just one giant wall of topics and replies. The formatting is horrible.
Gemini markup doesn't give you many options for formatting, just some headings lists and links, and clients can render them all differently.
You're missing the point if your criticism of it is "it looks fugly".


I want something more structured like station.martinrue.org


>he was a windows user all along

oh no no no no no


If anyone is still interested in this. I found a gemini textboard written in rust.
shitty, immature license though

File: 1630964531349.png (43.78 KB, 800x640, 800px-X.Org_Logo.svg.png)


How do you laugh at win users for security when we use shit like pic related?
17 posts and 1 image reply omitted. Click reply to view.



use the cfs scheduler and ionice.


*cfq scheduler


ah, it seems they replaced it with bfq.


You can't make ANYTHING work on modern hardware (as in support all the video, wifi cards and shit) unless you are a corporation
It would be a nice undertaking imho, a piece of code that's just completely illegal but benevolent and usable. It would warm my heart just knowing there are people rebellious enough to do this
And on windows you don't need fine tuning for that


I bareback the interwebz running wind0ws

File: 1631893826186.png (94.85 KB, 465x421, pspherexmpp.png)


There's a new Jabber/XMPP server in town, running on Prosody (https://prosody.im).

prolesphere.one | d7dchu2gnb4xce7y7vwmulonr4k2wh2dmobqyitwwjeiojfotrvw37yd.onion

Registrations are open, and so is room creation. The MUC (Multi-User Chat) server is at hive.prolesphere.one

The server supports OMEMO and OTR encryption, or should. Expect some growing pains in the early stages, there might be kinks to iron out so server may restart spontaneously. You can report any problems here in the thread or by messaging me at [email protected] (not an email).

You can find a list of XMPP clients here: https://xmpp.org/software/clients/

I use Gajim on desktop and Conversations (available from F-Droid) on mobile.
5 posts omitted. Click reply to view.


poezio and profanity according to my local xmpp guru. irssi-xmpp is also an option


Won't this just disappear one day like the board?


That is a fair criticism. I did let the hosting expire. It was a super busy period and it slipped my mind. To that end I am going to keep a better eye on the deadline and fund the crypto wallet + pay for the next quarter way ahead of the deadline so it doesn't happen again. I have also set a reminder on my phone.


Pidgin or gajim for desktop, conversations on mobile


you can always register a new account on another server and re-add everyone in your roster


I'm less interested in the biology of the deep sea(though it does help with understanding the origins of life alot) but rather the physical phenomena that take place at seriously deep depths, what strides should society take as a whole to venture into earths true depth and see what lurks not just below the surface but past the point where water pressure elevates to such levels where one could die simply by being in the right position either from pressure or from earths gravity and what kind of things can man find there


Youd die if you tried breaking through the sea floor to see what lurks below

File: 1624167488802.jpeg (7.54 KB, 293x172, hands.jpeg)


Why does Rust get hate here? I know its a memed language but is there any good reason not to use it other than be contrarian?
32 posts omitted. Click reply to view.


I don't think the language used in the implementation of the compiler has ANY impact on the language it compiles. The language design (usually) is done before even a lexer is written


It does influence how the language is compiled, its internal data structures and its function primitives. But more importantly if, we return to the example of fasm, it would only need to interact with code compiled from C during runtime, yet the lowest level of high level programming would still gravitate towards C-like patterns, because widespread modern hardware is for all intends and purposes "built to run C", meaning imperative code with variables of wordsize or below and continuos memory access, thanks to the bottleneck of read/write operations through a memory bus. http://www.loper-os.org/?p=46
Optimizing for hardware reifies its bias towards this programming pattern, even more so with POSIX compatible software. I programmed in C, so I cannot unsee.


Rust gets hate because a lot of its userbase is basically zoomer train stations, furries, etc hyperliberal types that'll remind you to say "person" instead of man or woman. But it's a better C++ otherwise.


The processor world moved on from C, now compilers do black magic to turn "imperative code with variables of wordsize or below" into SIMD code. If you want to write it by hand, all the architectures have nonstandard extensions to C to enable vectorization, but they are very clumsy and very obviously don't fit into C's model. A language that lets you easily write portable vector code could become really big.


File: 1632346846523.png (18.04 KB, 429x470, simd_cpu.png)

How does SIMD fundamentally change the dominant programming paradigms of C?
As far as I see, it simply amounts to loading data into the SIMD registers, potentially at a size different from wordsize, which could be implemented by loop unrolling of continuous memory access in the form of duff's device.
You may consider how high of a performance gain multimedia processing, the main motivation for adding SIMD extensions to x86, receives from caching, which indicates its proximity to the continuous memory access pattern.
>The high cache hit rates of multimedia applications can be contributed to the following three factors. Most multimedia applications apply block partitioning algorithms to the input data, and work on small blocks of data that easily fit into the cache. Secondly, within these blocks, there is significant data reuse as well as spatial locality. The third reason is that a large number of references generated by multimedia applications are to their internal data structures, which are relatively small and can also easily fit into reasonably-sized caches.

Delete Post [ ]
[ home / rules / faq ] [ overboard / sfw / alt ] [ leftypol / siberia / hobby / tech / edu / games / anime / music / draw / AKM ] [ meta / roulette ] [ cytube / git ] [ GET / ref / marx / booru / zine ]
[ 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17 / 18 / 19 / 20 / 21 / 22 / 23 / 24 / 25 / 26 / 27 / 28 / 29 / 30 / 31 / 32 / 33 / 34 / 35 / 36 ]
| Catalog | Home