1. 32

  2. 10

    fzy - not so bloated and written in C.

    1. 4

      Thanks for the pointer, it looks lean. Pardon my ignorance, but I’m not sure what “written in C” means as an advantage over fzf here - as a shell-level user of fzf which seems plenty fast enough for me, what does this imply for me?

      That said, icy makes a good point in reply to this, too, relating to the Unix philosophy.

      1. 2

        For one the binary is about 100x smaller..

        1. 9

          Not to be rude, but who cares? fzf is 2.3mb on my system, that’s basically nothing. In fzy’s readme under “Sorting”, fzf gives the same results as fzy, except for file over filter (admittedly yes, fzy is better on that one specific case). fzy claims to be faster but there are no benchmarks. fzf is already near instantaneous. fzf also has extended search syntax for exact match/inverse/prefix/suffix. What you call “bloat” others call “features”.

      2. 3

        +1 for fzy! It’s just enough. fzf does one too many things.

        1. 3

          Yeah there are many like this. skim, selecta are others. I found fzf to provide very responsive UI with extremely large input. Like, even if search takes a few seconds the UI still processes keystrokes. I can’t say the same about every simplified version out there. That said I haven’t tried fzy.

          1. 2

            thanks for sharing this, I’ve effectively replaced fzf!

          2. 7

            The idea of just being able to pipe any list in and then select an item in a quick an beautiful way is a great one. The integration with zsh, vim, tmux and other established programs also makes it more attractive. I use fzf daily.

            1. 2

              I personally prefer percol because it does just what you describe in the first sentence. Well, and because it was already my most used command before others exist. I find some of fzf extra features useful, but it goes a bit too much over the edge on integrations and whatnot.

            2. 4

              fzf is amazing and it’s become the primary means by which i navigate through code, whether it be in the shell, CTags, or my editor. goes great with vim and tmux.

              1. 2

                I’ve found it invaluable to building simple UX. It’s perfect for making bash scripts / aliases really pleasant to use, but don’t have enough complexity to warrant writing a python script.

                1. 1

                  I wonder if any using fzf against some large directories, I’d like to have some built-in caching mechanism to avoid excessive io, fzf over a single text file is much faster than traversing the filesystem especially when it’s on NFS.