1. 26
  1.  

  2. 23

    LtU does have some clever people. The majority of the predictions seem to be really good / close to truth. The most amusing/true I found were:

    Debates that both sides will lose because the debate will be made irrelevant by a third option: Emacs vs Vim

    We got vscode. I know that’s just one specific case and they weren’t really made irrelevant, but vscode really did take the dev world by storm in comparison. (with people using vi or emacs keybindings inside it)

    I think that we know more or less how the hardware will evolve: several (4-16) heterogeneous(not all with the same performance, some with specialized function unit) CPUs (due to the power wall).

    I take it as 4-16 in total, not 4-16 types. Right on with the recent big-little chips.

    Things from academia that will (start to go in the direction of) mainstream by 2020: Functional reactive programming

    Almost all the new web frameworks and a lot of declarative desktop draws strongly fro FRP.

    In 20 years the prevalent market will be the mobile platforms. The primary language used for development will be JavaScript or some variant, with some of the following characteristics:

    Pretty close if we consider that most websites target mobile now. And just perfect regarding “A misguided spin on lexical scope”.

    Therefore, programming will take on a much more “organic” feel: programming by example, programming with help of machine learning

    Only a year off for the copilot prediction.

    1. 11

      I dint think any vi(m) or emacs user is likely the switch to vscode. Vscode is just the latest in the line of sublime/atom/etc

      1. 2

        It’s much more than just the next sublime according to SO developer surveys. Vscode took some users from virtually every other option. Between 2016 - 2021: (multiple options allowed so there’s overlap)

        Vim fell 26% to 24%

        Vscode rose 7.2% to 71%

        Sublime fell 31% to 20%

        1. 9

          Direct links to the data:

          Given that neovim went from not on the list at all to 5%, I don’t think this supports the idea that meaningful numbers of vim users are switching to vscode. Emacs is virtually unchanged, from 5.2% to 5.3%

        2. 1

          I did (for a while), using the VSCode neovim extension (which literally runs neovim as the backend for editor commands while still getting VSCode’s GUI + language extensions). Ended up switching back emacs when i got a new job because CIDER is still the best Clojure tool imho.

        3. 6

          Debates that both sides will lose because the debate will be made irrelevant by a third option: Emacs vs Vim

          We got vscode. I know that’s just one specific case and they weren’t really made irrelevant, but vscode really did take the dev world by storm in comparison. (with people using vi or emacs keybindings inside it)

          As someone who doesn’t use any of these 3 editors, I don’t think this quite matches up to the prediction. VSCode is a good editor, and it surprised people and rapidly gained dominance for many, but it did not “render emacs vs vim debate irrelevant”. That fight is still raging, whether it’s for the editors themselves or keybindings inside other editors.” If the prediction was something like “a new editor will gain dominance over emacs vs vim” maybe this would count but let’s be honest – most popular editors already did this. Neither Emacs nor Vim have dominance anywhere; the prediction here is merely about the Holy War between the editors. VSCode did not solve this; No editor ever will lol.

          1. 8

            I’d argue that the distributions of emacs-as-vim and to a lesser extent vim-as-emacs (spacevim) as well as emulation in VS code or jetbrains IDEs is what made the debate w.r.t. UX irrelevant.

            The “holy war” debate between vim and emacs was made irrelevant by a bigger holy war (one which both vim and emacs take the same side of) which is the fight for free software and “non-corporate” software.. from that pov it was VS code that ended the debate.

          2. 2

            I can’t imagine a GUI code editor ever making a difference in the Vim/Emacs world. Vim and Emacs are fundamentally different from something like VScode because they are inside a terminal. Some people either prefer to or have to use a terminal for a variety of reasons and I don’t think any level of technological advancement would change most of those reasons (probably ¯_(ツ)_/¯ ).

            1. 3

              While I almost exclusively use Emacs in a terminal it is, in fact, a first-class citizen of GUIs (albeit with rather weird habits).

              1. 1

                Yeah, one of the nice things about emacs vs vim is that there’s a GUI (so stuff supports it) and there’s only one GUI (so you don’t have to worry about feature differences).

              2. 2

                If you look at NeoVim, there are lots of attempts of creating a GUI for it, e.g.:

                https://www.onivim.io/

                This suggests that there are a lot of people who want to use NeoVim as a GUI tool, not just terminal. As far as I’m concerted however, I value the terminal side of NeoVim, since I can run it on a Windows machine through SSH and start compilation without having to use RDP nor Windows on my local machine.

                1. 1

                  Definitely! People want to put Vim and Emacs (mostly Vim) inside any of their existing GUIs or create GUIs just for those tools.

                  But, as you said, the Terminal has special uses that aren’t replaceable and so GUIs like VScode will never directly compete with terminal editors.

                  1. 1

                    Onivim is actually based on vim, not neovim. Don’t remember why.

                    1. 1

                      Thanks, I didn’t realize that. From what I’m searching now, OniVim was based on NeoVim, but OniVim2 switched to Vim due to build issues of NeoVim in author’s environment.

                      https://github.com/onivim/libvim#why-is-libvim-based-on-vim-and-not-neovim