• 0 Posts
  • 20 Comments
Joined 1 year ago
cake
Cake day: July 8th, 2023

help-circle
  • Case-sensitive is easier to implement; it’s just a string of bytes. Case-insensitive requires a lot of code to get right, since it has to interpret symbols that make sense to humans. So, something over wondered about:

    That’s not hard for ASCII, but what about Unicode? Is the precomposed ç treated the same lexically and by the API as Latin capital letter c + combining cedilla? Does the OS normalize all of one form to the other? Is ß the same as SS? What about alternate glyphs, like half width or full width forms? Is it i18n-sensitive, so that, say, E and É are treated the same in French localization? Are Katakana and Hiragana characters equivalent?

    I dunno, as a long-time Unix and Linux user, I haven’t tried these things, but it seems odd to me to build a set of character equivalences into the filesystem code, unless you’re going to do do all of them. (But then, they’re idiosyncratic and may conflict between languages, like how ö is its letter in the Swedish alphabet.)


  • How about a government-sponsored, non-profit authentication service? That is, it should be impossible to get a loan, open a line of credit, or anything else in somebody’s name, without the lending institution verifying that it’s actually on behalf of the named individual. Eliminate the security-through-obscurity technique of using bits of easily-leaked personal information as a poor substitute for actual authentication.

    I mean, (as a comparative example) I have to go through an OAuth2 consent dialog to connect a third-party app to my email account, yet somebody can saddle me with huge debts based on knowing a 9-digit number that just about everybody knows? It’s the system that’s broken, tightening up the laws on PII is just a band-aid.



  • Just spitballing here, but if I read this correctly, you pulled the Windows drive, installed Mint, and then put the Windows drive back in alongside the Mint drive? If so, that might be the issue.

    UEFI firmware looks for a special EFI partition on the boot drive, and loads the operating system’s own bootloader from there. The Windows drive has one. When you pulled the Windows drive to install Mint on another drive, Mint had to create an EFI partition on its disk to store its bootloader.

    Then, when you put the Windows disk back in, there were two EFI partitions. Perhaps the UEFI firmware was looking for the Windows bootloader in the EFI partition on the Mint disk. It would of course not find it there. In my experience, Windows recovery is utterly useless in fixing EFI boot issues.

    It’s possible to rebuild the Windows EFI bootloader files manually, but since you don’t mind blowing away both OS installs, I’d say just install Mint on the second drive while both of them are installed in the system, so the installer puts the Mint bootloader on the same EFI partition as the Windows one. With the advent of EFI, Windows will still sometimes blow away a Linux bootloader, but Linux installers are very good at installing alongside Windows. If it does get stuffed up, there’s a utility called Boot-Repair, that you can put on a USB disk, that works a lot better than Windows recovery.




  • This just sounds like a bad idea, a solution in search of a problem. Sure, sudo is a setuid binary, but it’s a fairly simple program, and at some point, you have to trust the code. It’s also a very fundamental piece of the system that you want to always work, even (especially!) when other things get borked. The brief description of run0 already has too many potential points of failure.




  • I feel this in my soul, except about Windows. I’ve got a handful of machines at work that refuse to update to Windows 10 22H2. They give an error code during the compatibility check. Googling that error code returns dozens of forum posts with hundreds of users and “Microsoft support agents” chiming in. They give the same list of suggestions—that don’t work—to fix it. Nobody can say what the error code means, or what the compatibility check checks. The official Microsoft fix is to reinstall.

    I don’t want to reinstall. The suite of software these computers run would take several hours to reinstall.

    This is typical of my experience with Windows. (I’m a Unix/Linux guy.) I look up how to do something in Windows, and with the official Microsoft documentation, one of three things inevitably happens:

    1. I follow the steps and click the things, and it still doesn’t work.
    2. I can’t follow the steps because one of the things to click is greyed out for some reason.
    3. I can’t follow the steps because the documentation refers to an older edition, and Microsoft has removed one of the things to click.

    One time, when trying to get Excel to run a mail merge, I ran into all three problems in three attempts.

    The same happens with 3rd party sites. They never say the edition of Windows to which their guide refers, and the feature is deprecated or gone. (Most recently it was about getting a Windows 10 start menu behavior back on 11.)

    Oh, and since Windows is mainstream, a lot of the information is in the form of AI vomit, and covered in ads and dark patterns.


  • This is an infuriating aspect of this case. The courts could have held the clinic responsible for this loss without declaring that all frozen embryos are children by invoking the “prime mover” concept. Other courts have used it in, for example, surrogacy cases. In short, that concept holds that it’s the intent of the parent(s) that matters, as the prime movers in the process of bringing a child into the world, not just the mixing of some genetic material. Those destroyed embryos could have become children, as it was the parents’ intention to do so. And if nobody intends to implant embryos, for whatever reason, without the intent to make a child, they’re merely organic material, neatly sidestepping those questions.

    But, of course, the court wanted to impose its religious orthodoxy rather than issue a sensible ruling. Now we have those thorny questions.









  • This question has really got me thinking about the old days! I thought that it was looking into Debian Linux when trying to repurpose some old IBM PS/2 machines at work, because there were rumors of patchsets for the Linux kernel to support the MicroChannel Architecture bus and ESDI drives. But now I remember that it was actually GeekGadgets, a Unix environment for Amiga based around the ixemul.library. That’s where I first read the GPL, and admired its legal Jiu-Jitsu of using copyright laws to ensure freedom.

    I’ve never been a Windows user on my own machines as a result. I just went from Amiga, to FreeBSD, to Ubuntu.