Skip Navigation

Posts
6
Comments
148
Joined
2 yr. ago

I make computers

  • Honestly, I don't mind the lack of email that much since I have always preferred writing on my computer. That said, I do hate the lack of image support. A low-res black-and-white image would be better than none at all...

  • I've been using a Light Phone II as my daily driver since August, and I hope I never go back to my iPhone. The transition was pretty easy since I was already "off the grid", but I do sometimes have to toss my iPhone in my bag for things like Uber or the Starbucks app.

  • What annoys me about companies like StackOverflow, Reddit, Twitter, etc. partnering with AI firms is that they do not actually create any of the content on their platforms. Sure, if you read the terms they technically own the data, but still...

  • There are instances where the user is implied, but there is always a user. As far as Git goes, the user is almost always git.

  • I can provide my two cents regarding Point One only. Throughout my day, I am likely to read news on multiple devices. I use FreshRSS to keep my subscriptions, read status, and favorites in sync; and I treat it like a backend. That is, I prefer native clients compatible with one of the supported APIs (FreshRSS supports several). On Apple devices, NetNewsWire. On Linux, NewsFlash.

  • LLMS are not (currently) involved. The article states that video content, trained to imitate the likeness of a celebrity, is generated to recite human-written information. Or so they say

  • There's no way this could go horribly wrong!!

  • Ollama provides a Python API which may be useful. You could have it generate the story in chunks, having it generate a list of key points which get passed to subsequent prompts. Maybe...

  • I believe that RCS is a specification maintained by the GSM Association. That's not to say Google is not a member (they are) and has a strong influence, but Google doesn't own the standard either

  • Proof that cheap doesn't have to mean wasteful!

  • I read a comment on Reddit a while back that pointed out how much of the open source community has no issue hosting projects on GitHub while also lampooning Snap for having a closed-source backend server. However, since Snap (and GitHub) are open source themselves, nothing is stopping curious and concerned users from auditing the codebase or hosting their own servers. I removed Snap from my Ubuntu installation and use Flatpak instead, but I do not hate Snap. And for what it's worth, I always go for the native DEB when possible...

  • Me too

  • Better yet: use a hardware 2FA token that supports passkeys

  • Is it that hard to fact-check things?? Not to mention, a quick web search uses much less power/resources compared to AI inference...

  • I run Ubuntu and use the Nala frontend for APT to keep a log of my package installs. That way, I can easily remove everything if I no longer need to work with a particular language or set of dependencies. For anything too complicated, I like to drop into a Docker container (which integrates nice with VSCode/Codium)

  • Hot take from a Blogspot site with no theme... Jokes aside, it's a nice interpretation of the xz story

  • Interesting. I didn't realize XCursor predates most image formats XD

  • Thanks for all the good advice. I'll look into these solutions.

  • Yes, I do think you have a point. That's what I've done for several years, but I think it's time...