• 1 Post
  • 349 Comments
Joined 5 years ago
cake
Cake day: May 15th, 2019

help-circle


  • Semantic versioning.

    Most of the time. I use calendar versioning (calver) for my internal application releases because I work in IT. When the release happens is more consequential than breaking changes. And because it’s IT, changes that break something somewhere are incredibly frequent, so we would constantly be releasing “major” versions that aren’t really major versions at all.

    OpenDocument.

    Agreed compared to .doc and .docx. And if you’re going to version control it, markdown instead of a binary blob.

    For academic documents in STEM fields, I’d love to see a transition from LaTeX to Typst. Much cleaner, better error handling, and it has a web UI if people don’t want to install a massive runtime on their own computer.





  • Just take the dive into fish. It used to have a lot of problems with incompatibilities, but that’s been less of a problem lately.

    I haven’t found nushell to be that great as a day-to-day shell simply because it integrates poorly with other Linux commands. But when it comes to data manipulation, it is simply amazing. I’m currently (slowly) working on a plugin to query LDAP. The ldapsearch command uses the LDIF format, which is hard to parse reliably. Producing nushell data structures that don’t need fragile parsing would be a boon.


  • pingveno@lemmy.mltoAsklemmy@lemmy.mlWhat isn't illegal but should be?
    link
    fedilink
    English
    arrow-up
    11
    arrow-down
    1
    ·
    3 months ago

    Yup, a late friend of mine was a lobbyist at the state level for a mental health lobbying group. His daughter has schizophrenia and that was his way to give back in his retirement. Without lobbying, it’s hard for politicians to know when there is a problem they need to fix. They have a small staff and they don’t just magically know when there is a problem. The problem is when a politician either can’t sniff out unethical lobbyists or just doesn’t care.













  • I love OSM in so many ways, but unfortunately its address resolver is kinda garbage compared to Google Maps. For example, the string “10392 SE 23, portland, or, usa” correctly resolves to what I meant, “10392 SE 23rd Ave, Milwaukie, OR, USA” in Google Maps. OSM needs the exact city (Milwaukie is a suburb of Portland) and the “rd” on 23 to resolve.