• 0 Posts
  • 13 Comments
Joined 1 year ago
cake
Cake day: June 16th, 2023

help-circle


  • I don’t think you’re missing anything. I didn’t like the first episode as I found the humor somewhat jarring, and didn’t like Mariner. I kept at it as a show I watch while exercising, though, and it grew on me. While Mariner still annoys me at times, there’s a warmth and enthusiasm in LD that is quite infectious. I think they do a great job at teasing Trek while still loving it, and I am there for it.





  • Agree with many of the other comments here saying that they’d be very wary of such a project based on what these choices say about the project’s maintainers. Something else is that while I have real affection for email and particularly IRC based on past experience, I don’t think these two are without problems. Email is so asynchronous that many folks feel obligated to treat writing messages to a list more formally. This is not totally misguided since everyone subscribed gets this message delivered to them. IRC, on the other hand, is so synchronous that you should reasonably worry if anyone will be there to talk with, and about whether or not there are searchable archives.

    Something (like GitHub) that can be quick but is also perfectly serviceable for asynchronous communication really does have advantages, imho.






  • I went with ggplot2 some time ago, despite not using or knowing R at all. What pushed me in that direction was that I was using other plotting libraries (I don’t recall which at the time), and there was some aspect of spacing between elements or some such that was making a particular plot look ever so slightly ugly in my eyes… and I couldn’t fix it!

    In my frustration, I consciously decided to set aside my version of your “reasonably designed” requirement (I find R consistently frustrating in this regard, though I know some people do all their programming in it and I salute them). I gave ggplot2 a try with a cargo culting approach: search for how to make the kind of plot you want to make, and just tweak that template. I was blown away. I could find recipes for everything I wanted to do, the results were instantly more attractive than what I had before, and I could tweak everything.

    matplotlib is absolutely a reasonable option, but even years later I still have R environments attached to most projects specifically for data visualization, and still produce plots that are delightfully aesthetic. So here’s one voice to say that ggplot2 has real merit, especially if your aim is specifically to produce visualizations rather than explore a programming ecosystem.


  • Are you using DDG in addition to Kagi because of Kagi’s limited number of searches per month, or because DDG does something better?

    I’m a bit conflicted about Kagi because $5/month is a plausible price, but the limited number of searches seems like it would add an extra step of, “Do I want to use my limited search resource on this search?” to every search, which is an unwanted extra bit of friction.