• 0 Posts
  • 139 Comments
Joined 1 year ago
cake
Cake day: June 9th, 2023

help-circle


  • I have learned linear algebra in a few different contexts now, and each one I learned made it easier. When I first learned it, it was in a pure maths context and I found it tricky. It began to make more sense in university, when I learned it in the context of x-ray crystallography. I think more so than most topics, linear algebra really needs the context of it’s usefulness for it to really make sense, but also, I think I’d have struggled with the x-ray crystallography if I hadn’t already got a grounding in linear algebra from a pure maths angle.





  • I agree, you’ve captured much of why I came away from the article feeling a bit ‘hmmm’.

    Something I read somewhere that I found super interesting is that on Windows, when a process completes, the user often gets a notification or popup alerting them to this, whereas on Linux, it’s more normal for there to not be any confirmation messages when a process is finished. I hadn’t consciously realised this difference until I read this and reflected on how many times I’d have to double check things when I first started using Linux.


  • I was talking to a friend recently who was frustrated because they felt like tech support had been treating them like an idiot. They’re a reasonably techy person and had gone through all the troubleshooting steps in the documentation, but the person on the phone had them do it all again. I tried to explain the perspective of the tech support guy — the fact that people often refuse to restart their PC because it feels like too simple of a step and they feel patronised by the suggestion, to the extent that people lie about whether they’ve done a particular troubleshooting step.

    I told them that it was valid to feel frustrated with how long the call took when it could’ve been much quicker and simpler, but that they should attribute their frustration at people who repeatedly refuse to read the docs, rather than the tech support guy. My friend wasn’t an idiot, but they were tarred with the same brush because of how many people seeking tech support are belligerent idiots.



  • I wonder what would facilitate people to make their own solutions in this way. Like, I have made a few apps or automation things myself, but if I look at my “normie” friends who don’t have the level of tech familiarity that I do, they struggle with whatever out of the box solutions they can find. Poor IT education is a big part of this, and I’ve been wondering a lot about what would need to change for the average “normie” to be empowered to tinker



  • Book bundles make things messy, yeah.

    Something that I’ve really struggled with is that I wish it were easier to make notes of where you’ve gotten a recommendation from. Like let’s say I’m learning Python and there’s a dizzying amount of books available for that. Someone I personally know and respect gives a glowing recommendation of a book, so I download it and process it’s metadata in calibre or similar. Then a Reddit thread has a dump of free books. I download and import those. Then a blogger I really like recommends a few books, so I retrieve any of those I haven’t got yet and import them. Time passes, and I go looking in my library for a book to learn Python with, and I get overwhelmed by choices and I can’t remember which of them were good and for what reasons.

    I have a similar problem with bookmarks, because sure you can add tags, but tags are best when you’ve got a limited vocabulary, to maximise reusability. What if I just want to write a brief note about why I deem this worthy of saving, so that I’m not confused when I’m skimming over bookmarks months or years later.







  • I really love calibre, but agree it can be a bit quirky. Like many ADHD people, I’m my own worst enemy sometimes, when I attempt to do a big import of books and then it’s too much work and my library gets messy. Something I did that helped was I had different tags for different levels of processed. Newly imported books would be auto-tagged “new”. Then I’d batch process them and remove the tag if I fully completed it. Often I’d burn out when I’d discover a book that still had DRM, or was missing OCR (ability to select text within pdf), so I’d add “DRM” or “NoOCR” tags. My workflow works pretty well for me now, because I’ve made it easier to do half a job


  • I enjoyed reading this rant, I found it very relatable, even though it sounds like we have quite different academic backgrounds. Nowadays, I read a decent amount of non fiction for fun, but I still struggle with reader’s block if I don’t feel like I have a reason to read, like a sort of self-given assignment to be studying for. That’s probably school’s fault too