Having read your article, I contend it should be:
P(arentheses)
E(xponents)
M(ultiplication)D(ivision)
A(ddition)S(ubtraction)
and strong juxtaposition should be thrown out the window.
Why? Well, to be clear, I would prefer one of them die so we can get past this argument that pops up every few years so weak or strong doesn’t matter much to me, and I think weak juxtaposition is more easily taught and more easily supported by PEMDAS. I’m not saying it receives direct support, but rather the lack of instruction has us fall back on what we know as an overarching rule (multiplication and division are equal). Strong juxtaposition has an additional ruling to PEMDAS that specifies this specific case, whereas weak juxtaposition doesn’t need an additional ruling (and I would argue anyone who says otherwise isn’t logically extrapolating from the PEMDAS ruleset). I don’t think the sides are as equal as people pose.
To note, yes, PEMDAS is a teaching tool and yes there are obviously other ways of thinking of math. But do those matter? The mathematical system we currently use will work for any usecase it does currently regardless of the juxtaposition we pick, brackets/parentheses (as well as better ordering of operations when writing them down) can pick up any slack. Weak juxtaposition provides better benefits because it has less rules (and is thusly simpler).
But again, I really don’t care. Just let one die. Kill it, if you have to.
It’s like using literally to add emphasis to something that you are saying figuratively. It’s not objectively “wrong” to do it, but the practice is adding uncertainty where there didn’t need to be any, and thus slightly diminishes our ability to communicate clearly.
I think anything after (whichever grade your country introduces fractions in) should exclusively use fractions or multiplication with fractions to express division in order to disambiguate. A division symbol should never be used after fractions are introduced.
This way, it doesn’t really matter which juxtaposition you prefer, because it will never be ambiguous.
Anything before (whichever grade introduces fractions) should simply overuse brackets.
This comment was written in a couple of seconds, so if I missed something obvious, feel free to obliterate me.
Freaking love TUIs, it’s like they took the convenience of a GUI and the efficiency of the CLI and merged them. As a Neovim and Lazygit user myself it’s amazing what I can accomplish in but a few keypresses.
No matter the GUI you use, you’re leaving a lot of useful functionality on the table. By their nature, you only get a small fraction of git’s features. There are many useful commands I use regularly that are impossible to replicate using GUIs.
A good UI (for you personally) should do all the things you regularly do. Git is a complex and messy enough beast that when I have to use the CLI I’m going off the golden path and copy+pasting something arcane.
Tower is pretty nice for mac user too. I paid for it for a few versions back when I was coding full time. Now I just stuck to source tree for occasional freelance and personal projects.
I’m not sure about the exact commands, but you do something like gh auth login to authenticate the CLI and then something like gh ssh setup to change ssh’s config file to authorize using the GH CLI.
I’m not sure if you read the post yet but I also have a short section about alternative notations which are less ambiguous or never ambiguous. RPN has the same issue as most notations that are never ambiguous namely that it’s hard to read - especially for big expressions.
programming.dev
Active