prosody

There's positives and negatives to each. For government domain seizure, there's due process involved but working around it is harder (the service provider either has to acquire a new proper domain or onion domain, then disseminate it to the audience somehow). For ISP level blocking there's limited due process (at least in the cited case of LaLiga seemingly just issuing a complaint to the ISP), but the audience can easily work around with it with a VPN or sometimes just an alternate DNS server.

in an attempt to hurt relations with freer, traditionally more conservative nations

I don’t think those are the salient characteristics of the US from Canada’s perspective in this development, and because of that I don’t think of your analysis of this as motivated by short term political considerations is correct. Instead, the current US government’s unexpected decisions to turn the thumbscrews on Canada exposed Canada’s economic dependence as a vulnerability, and the Canadian government is at least trying to signal a capability to become less dependent in the future in the hopes that reduces their vulnerability as perceived by the US government. That vulnerability existed before and will outlast the current US government.

What's the motivation for blocking VPN read access for this and other services? Are AI scrapers using commercial VPNs to get around rate limiting?

Advertising a VPN endpoint in country A which in reality is in country B is a security concern for users trying to reduce their visibility to country B’s authorities. You’re right about the more fit to purpose tools, of course, but they’re more of an impediment to normal internet usage.

Kibbutzim in Israel from the 40s-80s tried a fairly radical project of communal child-rearing. It failed when the generation raised there rejected the choice to continue the project.

Viz: https://www.theguardian.com/lifeandstyle/2011/feb/19/kibbutz...

The specific structure and presentation of it could be, and the items individually could be. There are some jurisdictions where a dataset as such rather than just the structure and presentation have legal protections, but those are slightly distinct from copyright AIUI, and as such I imagine you'd want a tailored license for that rather than GPL. They're called database rights.

When I read about the leak of the new Meta internal guidance for content moderation[1], my first thought was that the only things they banned were likely things that they understood to be pre-genocidal speech (eg comparisons of a group to vermin). Rules that seem kind of arbitrary to a modern western audience but which click in place if you look at propaganda that was issued during historical genocides.

[1] https://theintercept.com/2025/01/09/facebook-instagram-meta-...

As I understand it, Cisco licenses the relevant patents and sublicenses them gratis to every user who uses Cisco’s binaries obtained from Cisco directly and not redistributed (thus also the difficulties in the article with working with Cisco’s unmaintained web server). The sublicense doesn’t apply to binaries built independently from source. I suppose this was imposed by the patent holders.

This article isn’t about that. It’s about the externalized costs that LLM companies are pushing onto webmasters because of their aggressive scraping. It’s one thing to believe that LLMs are a good thing, it’s another thing to believe that individuals and cooperative groups that run small internet services ought to be the ones to pay for that good.

I recall now the shutdown of Nekochan.net, which had been the main hub and message board at the time for SGI retrocomputing for several decades, due to concerns about the GDPR. While that might have been an overwrought reaction, and that laws like the GDPR and the Online Safety Act _can_ be followed by a dutiful webmaster, I can't fault any webmasters who choose to throw in the towel when they look at the potential penalties for failing to do so quite right. FAANG and the like can of course make the investments in compliance.

Isn’t that significantly on the Linux kernel not having stable driver ABIs?

The oldest Sanskrit text, the Rigveda, is usually dated to 1500 BC as an oral tradition. It wasn't written down until much later. The oldest surviving unambiguously Sanskrit writing is from 100 BC, using the Brahmi script. It actually isn't the surviving oldest Indic written language either, the Edicts of Ashoka date to 300 BC and are texts written using various scripts and vernacular languages.

In the case of this lawsuit they're suing unknown individuals. The case is Cengage Learning, Inc. v. Does 1-50. Apparently it's a US legal convention to just spitball the number of members of a group of unknown alleged coconspirators to 50.

Would you object if the FBI deputy director encouraged FBI agents to lawfully use their service firearms in the field to demonstrate their use to the FBI? There are real harms involved with increasing surveillance of the public and reducing judicial oversight. Placing the use of warrantless wiretapping as an end in itself rather than a possible means whose use needs to be carefully weighed necessarily means that the weighing will be less careful.

Interestingly, Wikipedia editors are aware that Wikipedia articles are used to find the current URL of sites that are forced to change URLs frequently due to legal or moral issues, and they face the same dilemma registrars and service providers face.[0] So although it seems somewhat more resilient than search engine companies to demands from copyright holders, it's not uncensored, something to keep in mind if you're infrastructurizing it for that purpose.

[0] https://en.wikipedia.org/wiki/Talk:8chan/Archive_2#Inclusion...

This seems like confused causality. The ban bill was put forward and passed because TikTok asked their users to contact their representatives and share their opinion about the ban bill?

Somewhat connected question, what is WPE? It’s something that got abstracted out of WebkitGTK, right? Is it a standalone port itself, or is it more like the framework to build a highly specialized one? I’ve seen at least in past versions that libwpe is an optional dependency of WebkitGTK, so they didn’t like abstract and then rebase it on the new component. Or is libwpe separate from WPE?

Adding to this wishlist, if it were available on Windows then it would be an option for a cross-platform webview widget, but it hasn’t supported it for a while now.

It's deprecated for removal.[0]

[0] https://openjdk.org/jeps/411

To be precise, the US does have life+70 for works published after 1978. The rest of the world went that route and the US begrudgingly followed, but the works published before the switchover date are grandfathered into the old system. The US copyright regime is a real rats' nest of complexity, see https://guides.library.cornell.edu/copyright/publicdomain

I've always liked the visual effect of the East Asian typographic rule that commas and periods at the end of a line hang. Makes the entire block of text look better aligned.[1] Probably matters more for text in those languages since the width of normal text characters doesn't vary as much as Latin text characters do.

e. More to the specifics of its use in CSS, it looks like Safari doesn't support the attribute to aggressively follow that rule,[2] as in the bottom right example of the previously linked image.

[1] https://www.thetype.com/wp-content/uploads/2017/11/AI-hangin...

[2] https://developer.mozilla.org/en-US/docs/Web/CSS/hanging-pun...

In addition to what the sibling comment said, the developer does in fact have a recourse to the legal system, which they wrote that they are preparing to make use of. Attempts to analogize the TOS dispute mechanisms of companies to the legal system frequently fail to note that, at least where money is involved, they exist within the legal system, not in a bubble.

I think it might become a little more useful in some contexts at least once ECH gets wider adoption. In that case afaik only the DNS server and the remote host you’re connecting to would know what FQDN you're connecting to—anyone else would only see the IP of the remote host. If the host serves requests for lots of different FQDNs, that’s less information going around.

Of possible interest to HNers, kinship terminology of different cultures is a subject that has a very elegant taxonomy. It was a major line of inquiry in early modern sociology, which tried to link other attributes of how societies were structured to how they named relations. https://en.wikipedia.org/wiki/Kinship_terminology

The answerer does say that either the perfect past (Latin’s closest to -ed) or the pluperfect past (Latin’s closest to had -ed) would work, they just chose perfect past. Maybe that choice was because the perfect past has a sense of finality that English’s simple past doesn’t, so it isn’t necessary to reach deeper into the sequence of tenses as it is in English.

The term FUD makes sense in low-risk applications where the benefit of the doubt ought to be given. In safety-critical contexts it should be completely reversed, and confidence, certainty and trust should be looked at extremely critically.

As well as being genetically related, Langley says it’s likely that the people who inhabited these islands in the Pleistocene era, 12,000 years ago, had “an image of an inter-island ‘community of practice’ with shared values and worldviews”.

How can you determine that from a poorly attested material culture? The way it’s worded makes me deeply suspicious that it’s backporting the doctrine of Indonesian national unity into the Stone Age.

It’s a little bit interesting to me that XBill has the eponymous figure replacing a diverse ecosystem of OSes, not just Linuxen, with his own. That probably is a sign of the times, when that diversity existed in fact and you might be playing that game on an HP-UX or another prediluvian machine. Maybe not though, maybe the waters were already rising. You might even imagine with some fairness an XTorvalds in which you vainly try to save the BSDs and proprietary Unixen from a Linux monoculture. It makes it easier for me to charitably accept how some systemd detractors describe themselves as for init diversity, even if non-systemd and non-sysvinit inits on Linux seem like a curiosity, rather than assuming that they are just contrarian. It is part of a tradition of valuing a variety of ideas and implementations of ideas, even when it becomes quixotic.

An example from Umineko that might tickle an HN reader's fancy: six people are found murdered in six different locked rooms. In the room with each body is the sole key to the door of one of the other locked rooms. Together they form a circular linked list of closed room murders.

If you didn't already know, the standard OTP schemes used in many systems have desktop and hardware security token implementations. Doesn't help with systems that use non-standard schemes unfortunately.