Yuuup. My personal website has been inaccessible to a few friends, they thought my server was down. It turned out they had some blocklist (not related to AI) installed on their PiHole, and for whatever reason my website was on that list. It is, in fact, to this day, because my request to unblock it went completely unanswered. I still don't know why the website is on the list.
Win9x Notepad in particular can only load files up to 64KB in size (edit: and supports only ANSI encoding, no Unicode). There were some actually useful additions to it up until Windows 10 or so - for example being able to handle LF (in addition to CRLF) line endings. But yeah, everything added in Windows 11 is just pure bloat.
To be fair, this grip indicator only (and still) exists when the window has a status bar. It's part of the Windows status bar design, not of the window design. Of course, many more applications used to have status bars than they do now, so that's why you see it less often.
This, a thousand times this. I have gone back to collecting CDs because it's often the only remaining way (short of pircay) to get original masters of many artists. Even lossless download stores like Qobuz don't have them.
The same is true when running Winamp. When I was dabbling with FreeBASIC many years ago, my games performed better when I was listening to music. Same reason!
Note that his was the VST2 era. VST3 was commercial license or GPL 3, which was an improvement, but only slightly, because it excluded open-source software released under the GPL 2, and also MIT/BSD/whatever-licensed software couldn't use it (without effectively turning the whole software into GPL-licensed software).
Entire BASIC interpreters were shipped in the very size-constrained ROMs of almost all home computers of the 80s. There was no luxury of parser generators. It's absolutely simple enough that you can write the lexer and parser yourself.
Objective: 7z (the format) doesn't have the same data recovery options as RAR. As it stands, RAR remains one of the best options for long-term archival of data for casual users thanks to its optional recovery records.
I think "zero-click" usually refers to the interaction with the malicious software or content itself, which in this case you don't have to interact with. I'd say the need to start an interaction with Copilot here could be compared to the need to log into your computer for a zero-click malware to become effective. Alternatively, not starting the Copilot interaction is similar to not opening your browser and thus being invulnerable to a zero-click vulnerability on a website. So calling this a zero-click in Copilot is appropriate, I think.
> At one point a psychic called in saying he’d had a crystal clear vision of a 46-year-old Japanese woman who was angry at J&J for refusing to hire her. She’d bought a jar of olives at the store before dropping off the poisoned Tylenol, he said. The specificity of his details made his story seem plausible until police asked how he came upon his visions and he told them his magic pen wrote the details out for him whenever he picked it up.
These days they could be asking an LLM instead and get a similarly convincing response.
CVEs are always filed against a specific product. If it's in a library, then it's typically the library and not all of its user. All of that information is already in the CVE database: https://nvd.nist.gov/vuln/detail/CVE-2025-31200
What OP wanted to stress is that just the CVE number on its own is not very helpful as a title. It would be helpful to at least mention the type of vulnerability and the affected product according to the CVE database entry.
Sure, and that also didn't vanish in a day. All I'm saying is that there is a ton of inertia, and people will not be switching to something different in an instant.
When "to google" has literally become a verb in several languages, you know it's not going to be that easy to convince literally hundreds of millions of people to use something different. For many unsavvy people, Google is and continues to be their window into the web, and they wouldn't even know they can have it another way.
There's a box labelled "Submit a correction or tip" at the bottom of the original article. Filling that in might have been better time spent than writing a blog post.
For reference, the quote is "I'd say maybe 20%, 30% of the code that is inside of our repos today and some of our projects are probably all written by software"
Microsoft has humongous amounts of source code in their repositories, amassed over decades. LLM-driven code generation is only feasible within the last few years. It would be completely unrealistic that 30% of all of their code is written by LLMs at this point in time. So yes, there is something in his quote that is intentionally misleading. Pick whatever you think it is, but I'm going to say that it's the "by software" part.
As much as I'd like to also dunk on them because of their AI nonsense, this keeps being misquoted again and again. He said that about 20-30% of their code is written by software. If someone like Satya says "by software" and not "by AI", you can be very sure that there is a good reason that he's phrasing it as carefully as this - because that includes a lot of things like auto-generated code, e.g. COM classes generated from IDL files. Of course in the current climate everyone that's not careful enough will just mis-interpret it as "30% written by AI", and that is probably intentional.
Code segments are not writeble by default on Windows, like on any modern OS, but you can make any memory segment in your own process writable using VirtualProtect. That is not unique to Windows as well, on Linux you could achieve the same with mprotect.