>>107540449
>China will continue to take the nothing that America gives them and like it.
Didn't Trump just sign a deal that undoes the last like 5 years of tech embargoes the US was trying to enforce on china?
Your president is blatantly open to bribery and China has deep pockets, they can probably take whatever they want.

>>107534418
Imagine chips monopoly, probably worst than now.

File: 1749050281482217.jpg (1.73 MB, 1280x6846)

1.73 MB JPG

>>107536174
I am eternally thankful for Deus Ex redpilling me on the Reds all the way back in 2000

File: delusional.jpg (33 KB, 524x391)

33 KB JPG

>>107538940
>Tariffs are a thing of the past

>>107534917
>Check prices on amazon to see if this anon is right
>Cheapest 32gb set of RAM that doesn't take months to arrive costs 400€

File: 1762465892968559.jpg (97 KB, 1079x1057)

97 KB JPG

>>107544980
I've been looking through /g/'s good sites archive and I'm not yet sure, I'm leaning towards building a multisearch that connects every search engine and just spamming the same search across all of them. Anyway, the reason I'm bringing up the good sites thing is because I found this and it seems close to what I want: https://www.faganfinder.com/

I can't be the only one who remembers it as faggotfinder kek

>Also, the most recent good sites pasta I found: https://rentry.org/good-sites

>The /g/ archive, on this topic:
https://igwiki.lyci.de/wiki/Search_engines

>The pasta that keeps getting posted, just to get it out of the way:
>Why not just stick with Google?
Search engines are a biased window into the web. We want to filter out SEO and discover actual interesting webpages.
>SHIT tier alternatives
duckduckgo.com

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

>>107544980
There are none left. They all suck. Yandex used to be good, but the Russians completely butchered it.

>>107545122
I forgot to mention, but this wiki page helps in understanding what index/backend each search engine uses: https://en.wikipedia.org/wiki/List_of_search_engines

>>107544058
Just use firefox at this point

>>107544105
Because firefox is full of telemetry trackers you dumb fucking faggot. Please go back to using chrome, all of you dont know shit about online privacy.

>>107545036
I use Brave.

>>107545066
How's the weather in Mumbai?

>>107545099
Dusty

>>107544853
>>107544817
Nvm I realised you were talking about Firefox's chrome not the Chrome browser

>>107543056
>>107543302
they dont have any plans for that yet. theyre focusing on making a working browser first. later on they will differentiate, which is fine because its not for end-users at all yet, its a big project after all

>>107544747
and those apps are just a webview, running the website

>>107543056
It's a noble effort by somebody to keep himself productive as a way of staying away from drugs. As far as the premises of it being actually usable, maybe one day it will be if he continues to work on it, but it's not going to see widespread adoption, not even remotely coming close to something like Firefox.

>>107544853
answering your question, vivaldi modified the source for tabs yes. I'm guessing edge had to too for its vertical tabs. it's why I said to build on vivaldi

>>107544821
>For that to happen all timestamps in the .tar have to be the same.
If it wasn't clear what I meant by that: I'm saying they have to have the same set of timestamps on all the files in the tarball, same user who created them, same tar and zstd version. E.g., for "aalib-1.4rc5-18-x86_64.pkg.tar.zst". A related topic is TorrentZip / trrntzip. That topic is: deterministically creating .tar or .tar.zst from a set of files. TorrentZip creates deterministic .zip files given one or more files as input; doesn't matter what OS or version of whatever you're using.

As >>107544794 said
>mirrors get their packages from syncing from another tier 1 or 2 mirror, etc
So only like one individual/team creates the .pkg.tar.zst - not multiple different computers creating the .tar.zst which usually means different .tar.zst as everyone has different file timestamps unless trying to make it the same.

>>107544839
>Most software
Huh, didn't know that.

>>107544846
>Check the checksum or verify the packages signature, that's what you actually care about.
True. That's what I was getting at by "determinism".

>>107544881
>For example, say there was a problem with a package. It would be completely valid to replace that package with a new problem-free version and to do it without bumping the package version (they wouldn't do that, they would of course bump the version but as long as the signature and checksums match then you know that it has come from Arch infrastructure)
I find this to be confusing. Oh, a problem package would be one which doesn't match to the checksum or .sig. No need to change the version as it's non-bit-identical one replaced with the correct bit-identical one.

>they wouldn't do that, they would of course bump the version
Also no really understanding this part.

>>107544903
>only like one individual/team creates the .pkg.tar.zst - not multiple different computers creating the .tar.zst which usually means different .tar.zst as everyone has different file timestamps
Necessarily the case, otherwise package.1.2.3.pkg.tar.zst.sig wouldn't verify in all mirrors.

>>107544976
I mean a problem as in a bug with application that requires a re-build of the package to happen. They could in theory do that without bumping the version number so your locally downloaded version is no longer the same as what's on the mirrors.

They would never do this, but hypothetically they could. The important thing is the checksums and signature matching, not what's in the files. It's the checksum and signature that tells you that the package you have downloaded in some way, shape or form originated from Archlinux infrastructure.

>>107544998
Now I understand.

>The important thing is the checksums and signature matching, not what's in the files.
A checksum like CRC32 or a hash (also referred to as a checksum I think) like SHA256 is based on the exact series of binary that the file is made of, so that part does matter for "what's in the files". Any change to the data at all = different hash. So if a .tar has the same files as another .tar but different timestamp(s) on the files in the .tar = different hash.

You probably already knew this.

Being honest with you, I don't currently know how to manually verify "abseil-cpp-20250512.1-1-x86_64.pkg.tar.zst.sig" to the corresponding .pkg.tar.zst file. Does it contain a SHA2 hash and/or a GPG thing which tells me I got it from a good source?

>>107545063
https://wiki.archlinux.org/title/Pacman/Package_signing

Pacman uses GPG so I think you'd do something like this if you wanted to manually verify it:

gpg --verify "abseil-cpp-20250512.1-1-x86_64.pkg.tar.zst.sig"

GPG then reads the signature file and will tell you if it matches.
For example:

$ wget https://archive.archlinux.org/packages/a/abseil-cpp/abseil-cpp-20250512.1-1-x86_64.pkg.tar.zst{,.sig}
$ gpg --verify abseil-cpp-20250512.1-1-x86_64.pkg.tar.zst.sig 
gpg: assuming signed data in 'abseil-cpp-20250512.1-1-x86_64.pkg.tar.zst'
gpg: Signature made Tue 17 Jun 2025 16:30:48 UTC
gpg:                using RSA key F00B96D15228013FFC9C9D0393B11DAA4C197E3D
gpg: Can't check signature: No public key

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

Do people really want extensions with LLM-hallucinated code?

File: 4818611651.jpg (85 KB, 1280x720)

85 KB JPG

>>107544934
>he doesn't bloatMAXX
>thinks he's big
NGMI

>>107544946
Post physique.

File: 1683387890814785.jpg (1.71 MB, 4160x3120)

1.71 MB JPG

>>107544948
My site is BVILT with SOVL not troongoop, you post your generic slop if you can afford to

you guys make me miss the days when /wdg/ was dead

>>107545081
>NOOOOO STOP MENTIONING HOW TROONWIND IS INFECTING WEBDEV AND LLMS GOYIM WE HAVE TO CONTROL THE NARRATIVE

>>107544676
Slow as shit
Telemetry shit
AI shit
Cloud shit

>>107544723
Yeah, still best desktop environment there is in comparison to really anything commercial or non-commercial.

>>107544676
Windows 10 IoT Enterprise LTSC
just werks

>>107545021
you are cucking yourself using 10 ltsc if youre on higher end hardware

>>107545032
Nah. He can simply install all the drivers he needs.

File: 1751829974419303.png (8 KB, 531x385)

8 KB PNG

>>107544649
What do you mean? It still is.

File: laser in hdd.png (717 KB, 2075x1344)

717 KB PNG

I didn't know this

File: 2GB-MO-disk.jpg (123 KB, 876x954)

123 KB JPG

>>107545035
Still rare tech but you see it in some high end high capacity datacenter HDDs. Also, MO drives worked basically the same way.

>>107542795
Try Meshcore or Meshtastic LoRa. It's actually nice for text messages on minimal power.

But it's not a replacement for 10 to 100Gbit per fiber on your ISP.

Bluetooth.
I wish there was an app like what Nintendo did with PictoChat, like a series of chat rooms but offline only and only for those in close proximity.

>>107544974
You can use Briar over BT but I don't see why that would ever have larger appeal as BT transceivers really are very short-ranged in practice. In almost all situations you're basically more or less in speaking distance anyhow.

Meshcore makes a lot more sense.

>>107542795
Direct dial modem. Startech still sells all the hardware you need to create your own old school BBS or UUCP network. Even your own personal (quasilegal) dialup ISP if you get creative.

>>107543516
meshtastic

>>107544269
Was meant for >>107544201

>>107542891
You know what, uefi is niggerware. I'm done with that shit. Should've never went with it.

>>107544252
>https://www.youtube.com/playlist?list=PL980gcR1LE3L_RdprUI2GkbyZPY998lLT
nice pick, I really like sphaerophoria. his WM playlist is p good as well [https://www.youtube.com/watch?v=TPw73HPEZA0&list=PL980gcR1LE3LIo4XuxUi7x616jUdf5o7H]

File: 2025-12-13-223849_1366x76(...).png (10 KB, 1366x768)

10 KB PNG

Langdev, functions edition

File: file.png (9 KB, 289x186)

9 KB PNG

>>107544730
I'm at the stage where I should compiled basic expression to bytecode. But the code generator needs to make decision based on types (need to insert zero-extending and sign-extending instructions for example) and I'm procrastinating on some aspects of the design of the type system.

Wether to have union types like in C if the sum types are going to allow you to choose the exact data represention.
Structs vs tuples, what syntax should they have. Wether (value, value) constructs a tuple or is a comma expression.
Wehter to have anonymous sum types like in Hare.
Should arrays, structs, union and/or sum types be copied when assigned to, or passed by values when making a function call by default? Knowing that you will have the possibility to take and pass references.

RIP murica >>107543802

>go to work
>hey anon can you make the thing no one understands or comprehends work?
>y-yes sir
>spend 8 hours a day banging my head against the wall
>dread going to work every day, spirals of feeling worthless and hopeless, dreams of going into farming, suicidal ideation
>sudden breakthough
>1 week of glorious progress where I get to write code and go into the flow state and love my life
>hey anon good work, can you make this other thing no one understands or comprehends work?

>>107544613
You can just get a job at a place that doesn't have a photocopier and gen z workers, you know.

>>107525617
I'll just answer your question. Max 2 years.

>>107544613
That's good shit though. You take on things no one else can solve and defying your own expectations. That's how you become a specialist which lets you demand way more money.

>>107532510
Depends on the D-Pad. If it’s a pile of shit like the Decks then fuck this gay shit

>>107544350
>>107544303
yeah i did a usb c mod on mine as well, note that it does prevent you from connecting it to a ps4 again, learned this the hard way. (you might be able to connect it with bluetooth but you wont be able to turn it on through the wired connection) Otherwise it does work great

>>107544595
>note that it does prevent you from connecting it to a ps4 again
You can do a 1:1 USB 2.0 based USB-C conversion, using the data pins too. The micro soldering drove me up the wall since I had shit tools at the time too, just moved.

>>107525250
The steam controller 2026

>>107525996
Yes a cable. Why would you want to add any input latency? Are you dumb?

>>107544940
you could probably get away with a normaly closed switch and wire it to the reset contacts on the motherboard. so if it gets released it connects the two contacts.

>>107544992
>>107544960
I was thinking more like a zigbee hall sensor and a relay that I automate with home assistant but yeah, those would be better.

>>107542287
Fuck, I just migrated my password vault to my Pi, now I'm scared

File: yr2oyu9m7u8e1.gif (653 KB, 368x200)

653 KB GIF

>>107545020
Sure we're just spitballing here
My idea:

>>107545052
Good stuff man, I liked the ideas. But that setup also probably looks more immediately related. Like you wouldn't want the sensor at the door go to the rack. But if everything is in software, that logic might not survive a reboot.

>>107544404
Not the feature. The other part. It's patched

Why is ltsc recommended over enterprise? If I use shutup10 and then disable automatic updates and telemetry via gpedit would that make the later completely private?

>>107544529
Because of the extra long EoL for monthly updates and how it doesn't come with any third or first party UWPs besides system ones. Though normal Enterprise 10 now has ESU which allows updates until October 2028 now, so LTSC 10 is less of a straight use case now.

File: {3C0B165F-C063-4968-BF97-(...).png (46 KB, 478x669)

46 KB PNG

>>107544550
>besides system ones
Wait which system ones?

>>107545028
Shit like Settings