What... is your name? Sir Galleri.
What... is your quest? To seek the Grail.
Do.... I have cancer? Yes. Er, I mean no -- aaarrgh!

If the service is free of charge, Google isn't going to employ human technical support staff to validate people's identity and make a best-guess attempt to restore locked accounts to their rightful owners. That's just too expensive.

If you pay for the service, even if it's only a couple of dollars a month, then you can always be identified as the account holder, just by tracing it back through your bank account (and identifying bank customers is a well known, if not completely solved, problem). Moreover if you're paying then you will have a human "account manager" whose job it is to sort out this stuff.

That's why for important things I prefer to use a paid service. That said, I haven't put the theory to the test - like if I lost access to my current email address and I had to ask for the password to be reset just on the basis of being the person paying for the account. There still might not be a clueful helpdesk, if 99% of accounts are free ones.

I use Outlook and Teams at work but I've found the web versions to be faster. I just open outlook.office.com and teams.microsoft.com in Firefox. Outlook in particular responds faster (for me) than the desktop app -- though it' still pretty sluggish. Teams in the browser is about the same experience as the desktop version (which is currently an Electron app anyway), but at least you only have one browser engine running, saving a few hundred megs of memory.

So obviously, links also open in Firefox. One thing I haven't managed to fix is the obnoxious "safelinks" URI munging. There are Greasemonkey scripts for that but Microsoft has changed the munging slightly over the years so I haven't found (or managed to write) one that works.

From the Bots of New York:

âoeIâ(TM)m freaking out!! Disney just unveiled the lineup for the next twelve phases of the Marvel Cinematic Universe (MCU). Here is everything they announced today: Hulk High School (2024), Bride Of Hulk (2025), Spider-Man: Father Of The Bride (2025), Drunk Hulk (2026), Spider-Man: Girlz With Attitude (2026), Spider-Man: Dance Off (2027), Spider-Man: Magic Mike (2028), Spider-Man: Dance Off: Close Up (2029), Spider-Man: Dance Off: Close Up: Expanded Picnic (2030), Bruce Banner: Bride Of The Father Of The Bride (2031), Mr. Elton John (Agent Of S.H.R.E.D. III) (2032), Spider-Man: Far More Better Than Before (2034), Baby Driver (2036), Spider-Man: Benediction Of Timmy (2037), Venom: Look Who's Talking (2039), Spider-Man: Double Baggy (2039), For The Sake Of Continuity (2042), Deadpool And The Wook (2044), Apokémon Now (2045), A Batman Xmas: Drunk Hulk Blows A 2.0 BAC (2047), Spider-Man: Braveheart Reboot (2049), Spider-Man: Braveheart Reboot: No Spoilers (2050), Thanos: Untitled Origin Story (2050), Origin Story: Untitled Origin Story (2001), Star Wars: Infinity War (2050), Greedo: Untitled Origin Story (2050), Spider-Man: Dance Off: Close Up: Origin Story (2050), Spider-Man: First Dance Off (2050).â

Unlike the ACM article's author, I'm not offended by the new unreachable notation. It won't bite you unless you explicitly decide to use it. The problem is the implicit undefined behaviour allowing the compiler to do crazy things. In most cases I can kind of see why it's that way, but it needs a strong warning from the compiler that it's about to elide half of your code having decided it can never be reached without undefined behaviour being triggered.

We know that ChatGPT can make up stuff that vaguely matches its training set. And you could train it on a corpus of peer-reviewed academic papers and newspapers (not including Wikipedia or other websites in the corpus) and it might then be able to write encylopaedia articles. But as Jimmy Wales noted, you'd have to take its output with a very large pinch of salt, and by that point you might as well write the article by hand.

A more fruitful approach, as well as one that's more of a challenge for current technology, would be to ask a machine learning system to find problems with Wikipedia articles. "The claim in the fifth paragraph cites the study [15] as evidence, but the conclusions of that study do not support the claim in the Wikipedia article. This is because the study says... while the article says... "

Once you've got a system that can nitpick articles like this, even if it's unreliable and sometimes gets false positives, it's still a good starting point for a human to go in and check. And if the model has misunderstood, hopefully it would be possible to use these mistakes to correct the model or as additional training data.

It's not operations per watt, it would be operations-per-second per watt, or in more reasonable units, operations per joule.

FWIW you can change the layout to a compact mode that doesn't waste as much screen space. The search function was famously crap but has improved recently -- you can click to see the message in context.

Where do you obtain chickens that have dark meat? They did when I was a boy but nowadays they grow so fast you get white meat only.

To be fair, having a single snake-like cable connecting all the PCs was a pretty common idea. To us today it seems obviously easier to have individual connections to a central hub or switch, but at the time a common cable was accepted, perhaps to save on the total amount of wiring used. Token Ring also used a single wire, though joined into a loop, and I think there was also Token Bus.

You are right that "transmit, and retry on collision" can never scale. But it is simple and hard to get wrong, as long as everyone is polite enough to wait a little before retrying. I'd say the success of the original Ethernet is an example of "worse is better".

I thought that looking moderately downward was the best position for your neck. Hence the advice to keep the top of your screen at eye level. (I don't quite follow that: my screens are all in portrait orientation and the top of each display is slightly above my eyes, but for most of the screen area I'm looking level or downwards.) I guess there's a sweet spot where you are looking downwards, but not too far down?

Digital security experts told me that bad guys can use software to easily translate your âoeatâ and âoedotâ into a regular old email address.

Well, duh. It's trivial. But the question is not whether they can, it's whether they do. The evidence, as far as I can tell, is that generally they do not. It's hardly worth it for them as anyone with the moderate level of intelligence needed to obfuscate their email address is too intelligent to respond to spam. The exception would be if some large, popular website started displaying users' email address using an automatically applied obfuscation. But I think sites aren't that stupid - they usually have "click to reveal the address", which requires a new request to the server and rate-limits any harvesting.

Ah, but apparently this might be a limitation of Parallels Desktop, and not a fundamental limitation of the x86_64 architecture, since

Windows 365 supports nested virtualization...

How does Microsoft manage to do it while Parallels cannot?

Microsoft does say that there are some limitations when it comes to virtualizing Windows 11 on top of macOS, pointing primarily to features that require nested virtualization to function as not being supported.

And that, I think, is the problem with operating systems like Qubes OS, or indeed Windows 11 itself with its Android and Linux subsystems. The PC cannot be fully virtualized, because the virtual environment cannot itself host VMs.

I have no experience on mainframes but I believe that where virtualization is done properly, it's not like that -- a virtual machine can itself host other VMs, down to an arbitrary depth. Anyone care to correct me?