In my experience in the U.S., I can manage my accounts at various banks and brokerages just as well from the website as from a mobile app. I’m ok with them restricting how their apps can be used as long as the browser remains an option. The requirement to use an app for all account management (or for authenticating to a website) seems to be mostly a European phenomenon.
Rather than sunk cost fallacy, I would call it survivorship bias: the projects that someone has consistently worked on over time are a small fraction of the total projects out there, and those that have survived are probably useful to someone (at the very least, their developers).
Yes, but (if I understand correctly) it would allow you to use your same X window manager (transparently under Xwayland) and run the odd Wayland application when needed.
Debian has quite a lot of code production. APT and dpkg are pretty standard open source projects, and Debian also maintains substantial patches for a lot of its packages. Not to mention any of the infrastructure.
I have managed to print out a paper ticket (from some dark corner of the website) and scan it on the barcode scanner at an MLB stadium. The employees were surprised it worked.
Leaving it on all the time would probably dim the backlight eventually more than turning it off, but I’m sure the monitor could handle it. Mostly a waste of energy.
I agree with you that this is a noticeable effect, but I have a hard time figuring out why it is, especially on Macs with the sealed system volume (which should make the operating system files bit-for-bit identical between a freshly set up Mac and one you've been using for years). Certainly it could come from the accumulation of cruft in /Library and ~/Library, but you can clean that up too.
Just watched that after I got home from work. I was pleasantly surprised with how in-depth it was. Thanks for sharing it! It does seem like the worst issues he was seeing had to do with parallel IO performance, which is exacerbated by the agent development use case, so for sequential file creation/deletion, APFS is bad but not horrible.
Is APFS that much slower than ZFS or btrfs (more comparable Linux file systems) or just ext4? Btrfs is noticeably slower than ext4, but at least all the Mac and Linux file systems are way faster than file operations on Windows :)
The design is supposed to be better than a true VPN, because neither Apple nor the exit node (Akamai, Cloudflare, Fastly) are supposed to know both who you are and what you’re doing. Of course, Apple pays them for this service, so they could exchange info.
When you lower the barrier to taking the shortcut, you get more people to take it who wouldn’t otherwise have done so. If cheating becomes incredibly easy, lots of students who would have otherwise been honest take the easy way out. Moral conviction gets harder to keep as doing the right thing becomes less rational.
What I like about uBO Lite in Safari is that I can use it without allowing it access to any of my data (it works in a purely declarative mode if you want). Even if it’s unlikely that uBO would be sold to a sketchy company, at least this provides protection against any supply chain attacks.
It’s an interpretation of the Fourth Amendment, which is loosely supported by Supreme Court precent in Carpenter and Jones. See my previous comment: https://news.ycombinator.com/item?id=46359826
Edit: To be more clear, the issue courts have considered is warrantless searches of ALPR databases (and whether that evidence should be excluded), not whether it’s constitutional to put up cameras.
This makes some good intuitive sense, but to me the sycophancy feels like it is an emergent property of turning a next word predictor into a conversational chatbot whether or not it’s intentionally trained that way. Your prompt and its earlier responses is all it has in its context window, so of course it lends undue importance to everything you say. Does that seem like a contributing factor to you?
I live in the U.S., have a U.S. phone number, and get maybe one spam call every six months. (Spam texts about once a month, with the exception of before elections, when it’s more.) I wonder what causes the difference. Do I have a newer phone number that isn’t on as many lists? Or is my carrier filtering calls before they get to my phone?
I briefly activated a second SIM on an MVNO that gave me a new number, and that one previously belonged to an old woman who got spam calls roughly every ten minutes. Clearly there is an issue here, but it doesn’t affect everyone equally.
I bought a desktop with “strange and undesirable” hardware (an Intel discrete GPU) from Facebook Marketplace at a substantial discount to any new machine with the same performance. Perhaps consider that route if you’re just looking for a stopgap?
Does your local NPR station (or even PBS) have local journalists? Obviously their primary formats aren’t written, but my local NPR has a fantastic website also.
5G is probably not a requirement for your phone service (although I can’t say for sure). The critical bit on networks in the U.S. and many other countries without 2G or 3G is VoLTE for phone calls and texts, which it does support. Some carriers have device whitelists that are only loosely related to technical features, though.
It’s fairly common for gifted people to struggle to communicate when they’ve made a multi-step inference in their heads. Whether this technology actually allows them to express themselves at the speed of (excited, fully focused) thought or not, I can see how the possibility of achieving that is exciting.
Microsoft’s Copilot products are a very similar situation where the enterprise and consumer (and GitHub) features only make sense if you think about the org chart. Both companies need stronger top down product thinking.
I’ve read this article every time it’s gotten posted here and it’s always gone a little over my head. I was able to follow how he used 8 x86_64 registers for the VM’s stack slots and how the VM instructions were implemented. How the padding and alignments of each version of the instructions was calculated is impressive and I can imagine how much of a chore it would be to figure out with a normal assembler.
Using SBCL as a macro-assembler is extremely cool, and then allowing CL code to call into the VM is where it really blows my mind.
Obviously it’s been over a decade since this article was written. For someone less familiar with SBCL internals (or CL in general), would something like AsmJit or Iced be a good way to achieve similar things?