3 ms·
That isn’t even about black magic wizardry. That tutorial covers adding files, committing, switching branches, and merging: in other words, the absolute basics.
by stouset 5mo ago
That isn’t even about black magic wizardry. That tutorial covers adding files, committing, switching branches, and merging: in other words, the absolute basics. It was, and maybe still is, one of the best beginner resources on how to make sense of the git CLI for people who are confused.
And it required at least twenty diagrams.
What do I want from you? Maybe a little less high-and-mighty a tone when dismissing a tool for being so complicated someone decided to use diagrams to explain it, when the very thing being replaced is so complicated that one of the best beginner resources needed to use diagrams to explain it.
- applfanboysbgon 5mo agoYou certainly do not need 20 diagrams to learn the basics of Git. If I had to consume the article you linked to start using Git, I would not have been convinced to. The article itself states that it assumes readers already know how to use it, so if that's what beginners are being pointed to, I think that's a crime against pedagogy.
- stouset 5mo agoYou certainly don’t need 20 diagrams to understand why you’d want to use jj. This author happened to think it would be useful. Apparently pictures of data structures are crimes. Meanwhile, that article is inarguably the thing that made git “click” for an enormous number of people regardless of your disapproval of how some people prefer to learn. Maybe be less judgmental.
- applfanboysbgon 5mo ago> You certainly don’t need 20 diagrams to understand why you’d want to use jj. You're talking about Git and jj. Surely, then, you must be a programmer, or at the very least have a passing understanding of programming. Perhaps you would like a refresher on control flow? "If it takes 20 diagrams to make your case for why jj is better... Git is fine" if (jjArticle.Diagrams.Count >= 20) Git = Fine; else if (jjArticle.Case == Compelling) Git = NotFine; else Git = Fine; I made an if statement. If the condition of my if statement isn't true, then a different branch is followed. The article is extremely presumptuous, asserting that the software the reader currently uses is not good enough. I find the article strongly fails to make that case, and a part of why is because anything that takes 20 diagrams to explain is more convoluted than I need, ergo, Git is fine. That doesn't mean jj isn't good. Maybe jj is even better than Git. But the premise of the article isn't convincing the reader that jj is better than Git, but rather than Git is not good enough and that the reader should explicitly go out of their way to stop using it in favour of jj, which is a much stronger case to have to argue. You suggested I'm being judgmental but if anything it is the article that is judgmental.