pull down to refresh
0 sats \ 0 replies \ @Norbert 2h \ parent \ on: ExposedByDefault - What Your Browser Reveals About You tech
Here I meant "the more hampered".
This is a rather impossible battle to win as long as we insist on having such a feature-packed web. The more we try to protect ourselves against fingerprinting, the less hampered our experience becomes.
I use LibreWolf (a privacy-centric Firefox fork) which normalizes my timezone to UTC, makes my window size less unique, disables certain HTML5 features and a ton of other small tweaks – but it only helps to a small extent, and it's not easy to reason about. I'm just hoping what I do is measurably better than nothing.
There are alternatives to the web, especially Gemini, which have essentially zero fingerprintable surface. But these are mere geek novelties and will never have broad appeal.
Looks like existing Moon cards did stop working. Someone got an update from Pay With Moon's customer support:
The tech team is aware of intermittent card issues experienced by Moon users. They're working on a fix.
Please note that new card creation is temporarily paused.
So they're not going to say it outright, but it's pretty obvious they lost their card provider. On the same day as Jan3 lost theirs. It would be interesting to know if this happened because they had the same provider, which got enough and canceled them both on the same day, of if there's a wider crackdown going on. I'm guessing the former, since I haven't heard of more cases.
Pay With Moon just removed their prepaid Visa cards from their site too, possibly yesterday just shortly after tweeting about Black Friday (https://x.com/Mrcode100/status/1994459257987637676), so new cards can't be created. Anyone have existing cards and can confirm whether they still work?
Good, we need geopolitically diverse mining.
Although they probably just contribute hashpower to "Antpool & Friends" :-(
Yeah, it's not in the article, but in the judgment, which is worth a read: https://www.canlii.org/en/bc/bcpc/doc/2025/2025bcpc192/2025bcpc192.html
Threatened with having his genitals cut off, the father gave the men access to his and his wife’s cryptocurrency accounts. Over the course of the evening, they made multiple withdrawals from both accounts totalling US$1.6 million (roughly $2.2 million Canadian), effectively draining the accounts.
Security is all about friction and inconvenience. If you have convenient access to USD 1.6M, there are scenarios like this that you're not secure against.
But if they had been secure against it, wouldn't the attackers have cut his genitals off? Of course there's a chance, but the calculus doesn't really check out. It's a significant and unusual escalation for no realistic gain.
The link stopped working. Here's a working link: https://winnipegsun.com/news/b-c-family-tortured-for-13-hours-by-thieves-impersonating-mailmen-in-2m-crypto-heist
It's a sarcastic remark about stale narratives. In the bcash runup a lot of people concerned themselves with "who would get the bitcoin name" and "who would get the subreddit" like it's some sort of divorce proceeding.
Let's be careful with terminology; there wouldn't be more than one "fork". Your question is better stated as "which side of the fork".
Nobody would "inherit" the name "bitcoin". The name would stay where it is. There aren't any kingmakers who declare who "gets the name". Perhaps in the 2017 era it was reasonable to speculate whether exchanges would swing their scepters and declare such things, but we learned that's just not how things work.
It's been almost ten years of this. I'm tired. Let's just get this fork over with, get our little airdrops and move on.
I still wonder how they have push access to the bitcoin repo in the UASF project on GitHub. That repo was last used for the BIP-148 UASF in 2017, and now it's suddenly sprung to life again. There are no publicly listed owners of the project, but perhaps Luke was one of them?
Doesn't really matter, just a point of curiosity and a perhaps unfair suspicion that they're using that project to establish legitimacy.
I think it's quite likely they break the law in many jurisdictions. And if not, I think it's very likely that mint operation will be outlawed in the near future. It's not like financial legislation tend towards permissiveness.
Cashu might be wise to adopt the assumption, right or wrong, that mint operation will be cracked down on. If it chooses to exist at the mercy of lawmakers, it is doomed. That's why I think the future of Cashu, should it survive, is pseudonymous mints, possibly running as Tor onion services. This path will be riddled with rug pulls, but over time, trustworthy mints will emerge, not merely by the Lindy effect, but also through relentless user vigilance and by the application of proof-of-liabilities schemes.
100 sats \ 21 replies \ @Norbert 26 Oct \ parent \ on: Are Cashu mint operators breaking the law? bitcoin
Thinking mints break the law doesn't mean you think they should shut down or that you agree with the law.
So NixOS has a board. The board created a constitutional committee, and gave it a mandate to create a constitution. The constitution creates a steering committee, and that committee created a moderator team and mandated it with moderating the forum. The moderator team demanded that the steering committee resign because they felt like they meddled too much in the affairs of the moderator team. When the steering committee refused to resign, the whole moderator team did instead.
I mean ok but … this is an open source project maintaining a Linux distro! Why do they need governance the scale of a small nation?? It's like it's designed to generate drama. These kinds of structures only attract people who enjoy navigating them.
Paywalled for me, the article (somehow without the graphic) can be read here: https://archive.is/XD1IW
So disappointing that CoinDesk would make an embarrassing mistake like that. I've written the reporter and I hope he will correct it even if the article is a bit old. This debate certainly doesn't need misinformation.
What a relief. This is not the first vote, and it will not be the last. They will vote again, and again, until something passes. Hopefully they can't garner more support for the current proposal. The best we can hope for is that what finally passes will be severely watered down.
The authoritarian trend in Europe is very scary.
Not many will move from using inscriptions (Taproot input scripts) to OP_RETURN (an output type). But some that are storing arbitrary data that aren't inscriptions (JPEGs etc), but are using fake public keys creating unprovably-unspendable outputs that permanently bloat the UTXO set, will hopefully switch to OP_RETURN.