The Internet Archive explains why you keep getting blocked, and the reason says a lot about the state of the web.
What the article says
- The Wayback Machine has been getting slammed by huge waves of automated traffic.
- The Archive added protections to keep the site running, but they sometimes block real people by mistake.
- One visible fix is a clearer error message explaining why you got blocked instead of a bare error code.
- If you think you were wrongly blocked, you can email the Archive with your operating system, browser and IP address and they'll look into it.
What HN is saying
- Many commenters think scrapers are hitting the Wayback Machine to get around blocks on the original sites they can't access directly.
- Plenty of people say they've been mistakenly blocked from home or work networks, often blaming shared IP ranges that scrapers also use.
- Several see this as part of a bigger shift, with more of the web hiding behind captchas and security walls.
- One sharp pushback argues there's no real evidence AI companies specifically are driving the surge, calling it assumption dressed up as fact.
- Many commenters say they already donate to the Archive or plan to start, calling it one of the few trustworthy pillars left online.
The EU just offered Canada a brand new kind of membership, and Trump is furious about it.
What the article says
- The EU's Ursula von der Leyen wants to make Canada the bloc's first ever associate member, a status that doesn't officially exist yet.
- This follows the collapse of US Canada trade talks and Trump repeatedly threatening to make Canada the 51st state.
- She framed it around shared values on AI, climate and defence. Trump responded by threatening to cut trade with Europe entirely.
- She also proposed a new European Security Council including Canada, the UK and Ukraine, plus stricter rules on kids using social media.
- Details are thin, and most EU countries still haven't even ratified the existing Canada trade deal from a decade ago.
What HN is saying
- Most commenters read this as symbolism aimed at Trump rather than a real path to membership, since the EU's own Canada trade deal isn't fully ratified either.
- Several pushed back on the idea of the EU as one unified bloc, pointing out members disagree constantly and some remain socially conservative on issues like surrogacy and abortion.
- One sharp exchange corrected a viral claim that a negotiation once reduced a Canadian minister to tears, calling the story exaggerated.
- A recurring worry, even from people who like the idea, is that Canada's economy is so tied to the US that closer EU ties won't change much in practice.
- Others welcomed it anyway, hoping it brings European style food culture and less reliance on ultra processed products to Canada.
Firefox is calling its new AI assistant private, but it ships your browsing straight to Mistral's servers.
What the article says
- Mozilla and Mistral are partnering to power Firefox's new Smart Window assistant.
- It reads your open tabs to answer questions, summarize pages, and remember things you clicked away from.
- The models are tuned for different languages and dialects, rolling out first in France and North America.
- Mozilla says conversations aren't stored by default and Mistral has agreed not to retain data.
- You can also point it at your own model instead of Mistral's, though that option is still rough.
What HN is saying
- The loudest complaint is that calling this private is misleading. Your query and browsing context get routed through Mozilla's own servers before reaching Mistral, not straight from your machine.
- Several people push back on the local model alternative, pointing out most laptops don't have the memory or battery life to run a decent model without cloud help.
- A few commenters just want a smarter bookmarking system rather than an AI chat feature bolted onto the browser.
- Someone flags a typo in the announcement itself, which people found fitting given the messaging complaints.
- One reply notes the appeal for non native speakers is real. Machine translation and local search actually help people who don't browse comfortably in English.
Someone shorted Dutch rail signals with jumper cables on budget day, and nobody agrees whether it was Russia or angry farmers.
What the article says
- Objects placed on tracks across the Netherlands, including near Amsterdam, tricked signal systems into thinking trains were present, freezing traffic control.
- Cables and metal pipes were used to short circuit the rails. One train did hit metal near Steenwijk but was going slow enough that nothing happened.
- Routes to Schiphol airport and other major cities were disrupted through the day, though new blockages had stopped by evening.
- The disruption landed on Prinsjesdag, the day the Dutch government announces its budget, when farmer group Farmers Defence Force had also called for protests over nitrogen emission rules.
- That group lit hay bales on roads but denies the rail sabotage, while saying it cannot rule out individual farmers acting alone.
- This follows a separate suspected sabotage derailment in France days earlier involving a much heavier piece of rail.
What HN is saying
- A commenter working in rail signalling explains the attack exploits fail safe design. Shorting the track makes the system assume a train is there, which is easy to trigger but hard to turn into an actual collision.
- Many jump straight to blaming Russia, pointing to a long list of past European sabotage incidents and arguing hybrid war should now be the default assumption.
- Others push back hard, noting the French and Dutch incidents look nothing alike technically, and that local farmer protests over nitrogen rules give a much simpler explanation tied to the exact day.
- One sharp counterpoint is that vague sabotage benefits Russia either way. People assume it was them even if they spent nothing on it.
- A Dutch commenter says locals are already split, with some blaming disgruntled farmers directly and others noting recruited small time saboteurs can be a real and separate thing from farmer politics.
Apple built a way to cryptographically prove a photo is real, and commenters are already plotting how to fool it.
What the article says
- AI makes fake photos trivial to produce, so Apple built a system to prove a photo came straight from an iPhone camera, untouched.
- The camera sensor itself signs the raw image the instant it is captured, before any software gets a chance to alter it.
- Apple's private cloud servers then turn that signed data into a viewable photo, checking every signature and attaching a tamper proof timestamp.
- The signing method is built to resist future quantum computers, and any photo can be revoked later if a sensor is found compromised.
- It only works on the newest iPhone Pro models, and it is designed so nobody can tell which photos came from the same device.
What HN is saying
- Commenters are impressed by the engineering but skeptical it fixes the actual problem people want solved.
- The obvious workaround everyone raised is photographing a screen or a printed fake, though others note the phone's depth sensor makes that harder than it sounds.
- The bigger worry is that this only proves a photo came from an iPhone, not that the scene itself is real, and people will read the badge as full proof anyway.
- Several pointed to Canon and Nikon's older camera authentication schemes getting cracked and broken, questioning whether Apple's version holds up any better over time.
- Others flagged the deeper cost, needing to trust one company completely, and a world where only an iPhone photo counts as believable.