• 14 Posts
  • 882 Comments
Joined 5 years ago
cake
Cake day: May 15th, 2019

help-circle
  • I wouldn’t say that. It varies from subreddit to subreddit, community to community, instance to instance. And sometimes they are just staying within goals for the given space, regardless of whether it is Reddit or Lemmy. Your personal experience will often vary with how aligned you are with the viewpoints of mods, if they engage in heavy viewpoint discrimination.













  • 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.