delta VS lucky-commit

Thanks for the difftastic & zoxide tips.

However, I've been using this git pager/difftool: https://github.com/dandavison/delta

While it's not structural like difft, it does produce more readable output for me (at least when scrolling fast through git log -p /scanning quickly

View on GitHub

I'm currently waiting on the integration between Delta and Difftastic:

https://github.com/dandavison/delta/issues/535

Difftastic now has JSON output, whic should make it much easier to build this.

I'm surely in the minority here. I've been using Emacs for almost a decade now, but I just can't get into the Magit workflow. I've tried several times, but always end up going back to Git on the command line. I have dozens of aliases, shell integrations, a nice diff viewer[1], etc., and interacting with Git has become muscle memory. I can commit, cherry-pick, rebase, bisect, fix conflicts, etc., in a fraction of the time it would take me to navigate Magit's UI. I'm sure with enough practice, a Magit user could do this more quickly and efficiently, but honestly, with some custom-built porcelain, Git's UI is not so bad. Though this could very well be Stockholm syndrome after using it for such a long time...

For whatever reason, Magit's opinionated workflows never clicked with me. A part of it is the concern that it will do something weird to my repo that I'll then have to waste more time undoing manually. I usually don't trust sugary wrappers around tools. And another is the fact I don't use Emacs on all machines, and setting up Git on a remote system is just a matter of copying over my config and some shell integrations.

Also, on a more personal note, I find the cultish fanboyism whenever Magit is brought up slightly offputting. Does anyone have anything bad to say about it? No software can realistically be this infallible. :)

[1]: https://github.com/dandavison/delta

For looking at diffs I still prefer the command line though, and use delta to view diffs between commits or branches.

See also Lucky Commit [0], which uses various types of whitespace characters instead of a hash inside the commit, which makes it look more magical.

I wonder about performance, though. Why is the author's method slower than the package I linked?

[0]: https://github.com/not-an-aardvark/lucky-commit

It can also be reproduced. https://github.com/not-an-aardvark/lucky-commit

There is also lucky-commit, written in Rust, that has the same idea as git-power, but with GPU acceleration and the ability to choose a custom prefix. Looks like their Rust implementation (with or without GPU) is significantly faster than this C++ implementation. Crab language wins again, so it seems.