frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Bitchat – A decentralized messaging app that works over Bluetooth mesh networks

https://github.com/jackjackbits/bitchat
396•ananddtyagi•10h ago•172 comments

I extracted the safety filters from Apple Intelligence models

https://github.com/BlueFalconHD/apple_generative_model_safety_decrypted
415•BlueFalconHD•14h ago•282 comments

What every programmer should know about how CPUs work [video]

https://www.youtube.com/watch?v=-HNpim5x-IE
21•bschne•3d ago•1 comments

Intel's Lion Cove P-Core and Gaming Workloads

https://chipsandcheese.com/p/intels-lion-cove-p-core-and-gaming
193•zdw•11h ago•33 comments

Show HN: I wrote a "web OS" based on the Apple Lisa's UI, with 1-bit graphics

https://alpha.lisagui.com/
351•ayaros•15h ago•108 comments

Neanderthals operated prehistoric “fat factory” on German lakeshore

https://archaeologymag.com/2025/07/neanderthals-operated-fat-factory-125000-years-ago/
85•hilux•3d ago•30 comments

Show HN: Piano Trainer – Learn piano scales, chords and more using MIDI

https://github.com/ZaneH/piano-trainer
64•FinalDestiny•2d ago•19 comments

Building the Rust Compiler with GCC

https://fractalfir.github.io/generated_html/cg_gcc_bootstrap.html
173•todsacerdoti•12h ago•29 comments

A non-anthropomorphized view of LLMs

http://addxorrol.blogspot.com/2025/07/a-non-anthropomorphized-view-of-llms.html
162•zdw•11h ago•186 comments

Why English doesn't use accents

https://www.deadlanguagesociety.com/p/why-english-doesnt-use-accents
141•sandbach•13h ago•194 comments

Fictional K-pop bands zoom to top of US music charts

https://www.bbc.com/news/articles/clyl1zyv1y2o
25•ranit•3d ago•21 comments

The first time I was almost fired from Apple

https://www.engineersneedart.com/blog/almostfired/almostfired.html
217•chmaynard•3d ago•88 comments

Crypto 101 – Introductory course on cryptography (2017)

https://www.crypto101.io/
180•pona-a•13h ago•14 comments

XAI data center gets air permit to run 15 turbines, but imaging shows 24 on site

https://arstechnica.com/tech-policy/2025/07/xai-gets-an-air-permit-to-power-its-supercomputer-but-pollution-fears-remain/
15•rbanffy•34m ago•7 comments

LLMs should not replace therapists

https://arxiv.org/abs/2504.18412
104•layer8•12h ago•128 comments

Deno 2.4

https://deno.com/blog/v2.4
21•hackandthink•1h ago•5 comments

Opencode: AI coding agent, built for the terminal

https://github.com/sst/opencode
221•indigodaddy•16h ago•60 comments

Async Queue – One of my favorite programming interview questions

https://davidgomes.com/async-queue-interview-ai/
139•davidgomes•17h ago•127 comments

Poland's clean energy usage overtakes coal for first time

https://www.ft.com/content/ae920241-597e-49d9-a4b9-bfdfa9deabb6
27•stared•1h ago•3 comments

High Performance Image Sensor Processing Using FPGAs [pdf]

https://oda.uni-obuda.hu/bitstream/handle/20.500.14044/10350/Gabor_S_Becker_ertekezes.pdf
41•teleforce•7h ago•1 comments

Thesis: Interesting work is less amenable to the use of AI

https://remark.ing/rob/rob/Thesis-interesting-work-ie
67•koch•13h ago•34 comments

Nobody has a personality anymore: we are products with labels

https://www.freyaindia.co.uk/p/nobody-has-a-personality-anymore
413•drankl•12h ago•323 comments

Portability of Tar Features

https://mgorny.pl/articles/portability-of-tar-features.html
14•Bogdanp•3d ago•2 comments

Get the location of the ISS using DNS

https://shkspr.mobi/blog/2025/07/get-the-location-of-the-iss-using-dns/
289•8organicbits•21h ago•81 comments

Uncommon Uses of Python in Commonly Used Libraries (2022)

https://eugeneyan.com/writing/uncommon-python/
31•sebg•3d ago•4 comments

Swedish Campground (2004)

https://www.folklore.org/Swedish_Campground.html
88•CharlesW•10h ago•24 comments

Backlog.md – Markdown‑native Task Manager and Kanban visualizer for any Git repo

https://github.com/MrLesk/Backlog.md
147•mrlesk•14h ago•34 comments

Functions Are Vectors (2023)

https://thenumb.at/Functions-are-Vectors/
194•azeemba•19h ago•100 comments

There's a COMPUTER inside my DS flashcart [video]

https://www.youtube.com/watch?v=uq0pJmd7GAA
57•surprisetalk•10h ago•12 comments

Show HN: A Language Server Implementation for SystemD Unit Files

https://github.com/JFryy/systemd-lsp
41•arandomhuman•9h ago•17 comments
Open in hackernews

Web3 Onboarding Was a Flop – and Thank Goodness

https://tomhadley.link/blog/web3-onboarding-flop
16•solumos•7h ago

Comments

geekodour•7h ago
Absolutely not a crypto person but I've been following this team called noice.so, so far I think they've been doing the right thing in the space, pretty new and think they are operating at the right level of abstraction.
solumos•6h ago
This is actually a good example of what I'm railing against.

Giving your users airdrops is almost always a perverse incentive, and trying to figure out how to onboard to Farcaster as "Step 1" in your product journey is extremely limiting.

rasengan•6h ago
There are things that people want and, yet, can’t know they want, until much later in their lives; often times, too late.

Not an endorsement for “web3,” which seems to be a sort of centralization of the decentralized technologies at this point in time, unfortunately.

solumos•6h ago
There's an argument that the only truly decentralized money is Bitcoin. I don't necessarily disagree with that.

Centralization has its perks though. A lot of things are priced in dollars, and it's somewhat convenient to not have to worry about rampant speculation and boom/bust cycles affecting the medium of value transfer.

There may be a future where rather than a claim on fiat, there are tokens that provide a claim on some underlying stable-priced commodity or similar asset that isn't tied to a specific government. I think that would be a step in the right direction.

But yeah — "Web3" was a buzzword, and thinking that a peer-to-peer programmable payment network was the next version of the internet was a bit misguided.

kragen•6h ago
Never mind that. Give Freenode back to the people you stole it from. Every day you hold onto it you are worsening the lives of tens of thousands of people.
__MatrixMan__•6h ago
This article is here to tell us that stablecoins give us what we wanted all along: a faster, more convenient, and more secure value transfer experience.

Those things are nice I guess, but if you had asked me in 2021 what people wanted out of Web3, I'd not have listed those things. We wanted an economy based in types of value which better align with our values than fiat does.

Stablecoins are just a fresh coat of paint on the same problem as before: Money that is issued by a select few in furtherance of schemes that will make that few richer, with no reason to expect that those schemes will benefit the masses--masses who are expected to value that money just because we're told to.

nine_k•6h ago
People prefer widely accepted tokens of value, which are the few strongest currencies (USD, EUR, RMB, JPY, CHF, not much more). BTC has made enough inroads to be considered semi-widely accepted, and be useful as is in certain circumstances, but that's about it.

Blockchain is a useful technology, but, it appears, mostly not for carrying fancy coins.

XorNot•6h ago
What exactly is blockchain useful for?
__MatrixMan__•6h ago
Being a single source of truth which changes only according to certain rules, and not at the whims of some administrator. In short: accounting.

It solves a lot of headaches compared to having each stakeholder maintain their own database and then do pairwise syncs in hopes of convergence.

dehrmann•5h ago
Isn't an issue with Bitcoin that it's actually controlled by the miners, so they could collude to change some of the issuance rules?
__MatrixMan__•5h ago
I'd rephrase slightly:

Bitcoin's problem is that, since it's controlled by the miners, they're unlikely to change the issuance rules. Adequate security can be had at a far lower cost, but that's a change that would damage their investment in mining hardware, so the miners will never do it.

I hope one day we start valuing abstractions based on the desirability of their externalities. After such a shift, bitcoin will look pretty silly--who would want use waste-electricity-coin when they could instead use capture-carbon-coin or go-to-mars-coin? But it would seem that we're not there yet.

dehrmann•5h ago
At what point do large AI players have enough GPUs to take over the blockchain?
__MatrixMan__•5h ago
That would be 51%. I'm not a big fan of bitcoin, so I haven't tried to forecast it in terms of years.

Asic miners outperform GPU's though, so it would have to be an awful large pile of GPU's.

wbl•5h ago
If it solved these headaches why didn't it catch on?
__MatrixMan__•5h ago
I used to work for a company which later got bought by Fiserv (a large payments provider). I had these headaches re: transaction data. Wrote a lot of SQL to reconcile the pairwise sync problem. I tried to convince them that this is a problem for a blockchain, but I grew impatient and left.

Just a few weeks ago Fiserv launched a stablecoin on Solana: https://investors.fiserv.com/newsroom/detail/2848/fiserv-lau... I doubt they listened to me, probably just came to the conclusion independently.

So I'd say that it has caught on, just in a way that doesn't involve the users' preferences as much as I'd like. Because it's not like Fiserv is going to let their end users transact with any of the other assets on Solana. They'll wrap it in an interface that just calls it "$" and the user will be none the wiser.

__MatrixMan__•4h ago
The anthropological evidence suggests that people prefer densely connected networks of interpersonal debt to any kind of tokenization.

Those widely accepted tokens of value are the result of somebody violently imposing that widespread acceptance, usually because they won a war and now they want to continue bossing the losers around without having to remain combat-ready all the time. The only people who benefit from using widely accepted tokens of value are those who have a hand in issuing them. To the rest of us they represent a disadvantage.

For a long time this was a stable configuration because densely connected networks of interpersonal debt become computationally burdensome to work with at the scale of the empires created by those wars. But if our money can be programmable, then maybe a new way can be found which better respects the needs of all of its users, rather than just the few at the top. I don't know what such a thing would look like, but it would look nothing like a stablecoin.

Incipient•2h ago
>Those widely accepted tokens of value are the result of somebody violently imposing that widespread acceptance,

I get the poke at the US there, but I have to refute this bit. The EU is probably the main example of acceptance through ubiquitousness. Rmb through aggressive economic policy, the yen through (traditionally) stability.

The usd would arguably be through technological superiority at the advent of global financial markets, and their companies offering global services to facilitate trade (via usd).

We'll probably see the rmb replace that gradually/rapidly, again not through violence.

tootie•6h ago
Stablecoins also largely evade regulation and are widely exploited for criminal use and sanctions evasion.
gregmac•6h ago
Am I the only one struggling to decipher this?

I thought web3 was supposed to be some kind of decentralized compute, where rather than run on your own hardware or IaaS/PaaS you could make use of compute resources that vary wildly day-to-day in availability, performance, and cost, because they were somehow also mining rigs or something? But it's "decentralized" because there's not one entity running the thing.

There is not a mention of that in the article.

Is it actually supposed to just be microtranscations paid with cryptocurrency? Where's the "decentralized" part of that?

Anyway, instead the best I can see this article seems to be talking about how it turns out people aren't using blockchain for buying things, and makes the (apparently) shocking conclusion "the one thing people always wanted: money that just works."

aurareturn•6h ago

  I thought web3 was supposed to be some kind of decentralized compute, where rather than run on your own hardware or IaaS/PaaS you could make use of compute resources that vary wildly day-to-day in availability, performance, and cost, because they were somehow also mining rigs or something? But it's "decentralized" because there's not one entity running the thing.
Web3 is a rebrand of crypto scams. The 20018 crash exposed the crypto industry to how wide spread its scams were. So the industry rebranded to "web3". Nothing changed.

The scam message was that so many people got rich investing in web2.0 companies like Facebook. You can get rich too by investing in web3 (shitcoins & NFTs).

jowea•6h ago
Web3 was quite a cryptoassets sector buzzword, but the decentralized part of that is "cryptocurrency".

If we take the buzzword seriously, yes there was more to it than microtransactions with cryptocurrency, but without personal wallets on the blockchain capable of those transactions the rest of the ideas like NFT tickets, Decentralized Finance etc are impossible.

mxwsn•6h ago
Stablecoins transferred $27 trillion in 2024 - more than Visa and Mastercard combined. This is right in the article.

Stablecoins operate using decentralized ledgers on e.g. Ethereum which use decentralized compute. This isn't mentioned explicitly because the target audience knows this already.

aurareturn•6h ago

  Stablecoins transferred $27 trillion in 2024 - more than Visa and Mastercard combined. This is right in the article.
Vast majority is from one exchange wallet to another. Left hand to right hand.
oersted•5h ago
Aren’t stablecoins also backed by a central authority that guarantees it will always exchange the coins for a fixed amount of cash? That’s what makes them stable right? At least the major ones like Tether.

And by now we have seen many cases of stablecoins predictably crashing when trust in that backing authority dissolves. Most famously UST/Luna but it’s a long list.

I suppose they are useful for covert transfers, and the actual transfer mechanism is decentralized. But they are strictly worse than normal currencies for storing wealth, since the backing authority is a private company with virtually no oversight. And the utility for transactions would vanish if you were not confident that you can exchange it back and forth with cash immediately before and after the transfer.

TheDong•5h ago
That comparison isn't really equal.

Visa / Mastercard have such large fees that they're mainly used for commercial payments like a coffee or couch.

If most of the Stablecoin transactions were for buying a coffee, I think it'd be fair, but the vast majority of stablecoin transactions are for shuffling money around, i.e. to buy and speculate on bitcoin, or to move money to an exchange to liquidate some crypto into cash.

I think the current use of stablecoin transfers is closer to a wire transfer.

SWIFT apparently deals with about $1.25 quadrillion/year, so ~50x the claimed amount for stablecoins in the article... though there's more than just SWIFT out there too.

idk, I don't really have a point, I'm both amazed stablecoins are such a big number, but also feel like the comparison the article's making with VISA is misleading for how they're currently used.

solumos•6h ago
Not exactly. Internet Computer or Akash Network would be more closely aligned with "decentralized compute" — smart contracts in the EVM world aren't really meant to decentralize "general" compute, but they do provide primitives for programmable transactions. Those transactions can include a "token" as well as abstractions on top of tokens (e.g. a vault)
zerkten•5h ago
Web 3.0 was a synonym for the Semantic Web (https://en.wikipedia.org/wiki/Semantic_Web), but the term was taken over by crypto folks as it dropped off after the mid-2000s.
aurareturn•5h ago
Side note: LLMs made semantic web completely irrelevant.
pests•5h ago
To give you a real answer instead of just bashing crypto:

There is no real compute happening on the actual blockchain. The more instructions you have to execute, the higher the gas, the larger the fee the user will have to pay. The blockchain is used more as a database which the Web3 app can then query (for free) and use as a source of truth. It's not just currencies and microtransactions - game stats, property ownership, a users notes or todo list, anything you want. But you are definitely not doing any kind of major compute.

Web3 now refers to web apps that interact with a blockchain instead of a server API. They are mostly built with modern web technologies (React, Vue, etc.) Many are opensource and lots of communities participate in developing standards and in participating bodies. Uniswap, for example, is a website and underlying protocol for changing one coin into another, is now on v3 (UniswapV3) and many other projects and companies (SushiSwap, PancakeSwap, etc) take this frontend and the blockchain contracts to implement their own compatible offerings.

These API call's go through your wallet (usually a browser extension or app) which injects the JS needed into the website or handles it itself. The wallet will usually perform an RPC to a remote server to perform the actual blockchain communication. Most chains have default free RPC servers but you can also pay for a more premium/less latency/higher uptime RPC services. You can also run your own RPC infrastructure and handle that yourself.

User identities are tied to wallets. You do not have to sign up or join any Web3 app, you just "Connect" your wallet which is a free action which basically consists of signing a pre-generated message vouching you are joining which the Web3 app can verify.

Many standards have developed like Uniswap mentioned earlier. Interfaces for NFT's, delegated spending (to condense transactions and save gas fees), notifications, chat. Because all contracts are on a chain and every frontend is just a webapp, and many services ensure transparency by posting the source code for their contracts, it enables seamless integration of other APIs.

For example, another layer has been abstracted on top of UniswapV3. Since there is hundreds of swaps all supporting UniswapV3, you then had the development of apps called "routers" which would check all the swap sites and find you the cheapest rate. These routers can even swap between multiple token chains and assets to arrive at your end coins.

One more way this is more decentralized. If any of these web frontend's were to go away, the contract still exists and you can manually call it's functions to achieve your goal like withdrawing money, reassigning an NFT, or evolving your pet.

Furthermore, since all Web3 apps are just really calling contract's behind the scenes, it's easy to just send these actions yourself in a program. You can turn 20 clicks in some UI into a script that does the same interfacing with the blockchain directly.

One last point, but all this data is being stored on chain and is exportable and backed up forever by millions. I have personally had trouble with losing devices and having poor backup discipline. So, I have stored my personal notes and todo list in a custom (encrypted) contract I wrote a few years ago. I can lose access to every device or password I know and can still access it anywhere in the world because I have memorized my 12 word wallet mnemonic and every node has a copy.

Animats•6h ago
Stablecoins can, when interest rates are reasonably above zero, support themselves with the interest on the money. That could potentially work. The industry tradition, however, is the promoters stealing the money.[1] Usually by investing it in something risky, with the intent of keeping the excess profits but dumping any losses on the stablecoin holder.

Trump has a stablecoin, "USD1". It's partly backed by Trump's memecoin, "TRUMP" Really.[2] What could possibly go wrong?[3]

[1] https://www.web3isgoinggreat.com/

[2] https://www.msn.com/en-us/money/companies/usd1-the-cryptocur...

[3] https://coinmarketcap.com/currencies/official-trump/

solumos•6h ago
I would be very surprised if any of USDC, PYUSD or USDG followed in that tradition given their attestations.

But you are certainly correct that there is no shortage of bad actors in this industry.

since7•6h ago
The article brings some valid points. There's a lot to talk about in Web3 UX (as well as AI UX) and this is where the hatchet is buried.

Having said that, to me this article seems written with the help of ChatGPT. That is not bad per se but I feel it should be mentioned (eg. Venkatesh Rao's Sloptraptions)

snvzz•1h ago
Next is Web5.

Disregard HTML, CSS, JS and other nonsense; Embrace RISC-V machine code.