rtx -> mise rename

[jdx]

If you're having trouble with the migration, first read this doc: https://mise.jdx.dev/rtx.html

I also welcome you to our discord: https://discord.gg/UBa7pJUN7Z and as always, feel free to post an issue with any problems. I'll continue updating that doc as we discover migration-related issues worth noting.

I really apologize for inconvenience here. There just wasn't a way to do this that wasn't sudden I could think of.

[jdx]

Several bugs were patched today so please make sure you're on an updated version.

Note that we moved the default location install.sh uses from ~/.local/share/rtx/bin/rtx to ~/.local/bin/mise to keep things cleaner and reuse a common convention. This technically happened post-migration so some people may be using ~/.local/share/mise/bin/mise. That shouldn't cause any problems. It doesn't really matter where the CLI is. mise self-update will just use the location it's currently in. I figured I should mention this but it should only impact new installs. Docs reflect the new location already.

[zyrikby]

I've tried to find out but could not do this: what is the reason behind the renaming? Is it due to the new functionality appeared recently?

[sassdavid]

As I understand it mostly because of the name clash with Nvidia, it was difficult to search etc.

See: #1306

[blackhaj]

Not so much a bug as a potential conflict others might run into. Sharing the solution so they don't lose any time on it

After switching from rtx to mise, when I open a new terminal window I get the following error:
(eval):cd:39: too many arguments

I have managed to locate the issue which is an alias I had declared that I am guessing conflicts with an alias or exposed command in mise, namely test:

This is my alias
alias test="cd ~/path/to/somewhere/test-suite"

If I remove this, the issue goes away.

To Reproduce
Steps to reproduce the behavior.

1. Add an alias for test to one of your startup dotfiles e.g. .zshrc alias test="cd ~/path/to/somewhere/test-suite" (add it before the eval line to activate mise)
2. Open a new terminal window and you should see a similar error

[jdx]

yeah what's happening is you're creating a function named test so when I later run test -n "$(declare -f command_not_found_handler)" it doesn't use the builtin test.

I think it's possible to rewrite that code using [ which would be less likely to be overridden.

[jdx]

this isn't related to the mise rename btw, it's the somewhat recent not_found handler

[blackhaj]

Ok great, thanks for the clarity. It's not an issue for me as the test alias was redundant anyway so I have deleted it but I am glad I was on the right track with my debugging.

Thanks for the great tool 🙌

[jdx]

yeah I wouldn't recommend adding an alias over a builtin like that. Any code that you "source", "eval" or paste in your terminal that uses "test" would break. Still, this is something I can easily fix. I think someone else had this problem but wasn't able to diagnose it.

[chadlwilson]

Sadly upgrading via Brew from 2024.0.0 to 2024.1.4 did not migrate installs/plugins/config for me as noted at https://mise.jdx.dev/rtx.html

It seemed to basically start fresh. After correcting my oh-my-zsh config to eval "$(/opt/homebrew/bin/mise activate zsh)" at the new location and running mise once to try and trigger the migration, I ended up manually moving the contents of installs (minus python) and plugins from the .rtx folders to the .mise folders.

Note sure if subsequently addressed or whether any help could be given to figure out why the auto-migration didn't work?

[chadlwilson]

Sadly I didn't capture the first run of mise after the update since it was shell initiated, nor run with higher verbosity, but I suppose I could try and replicate the issue with a rollback if it's not known.

$ brew config
HOMEBREW_VERSION: 4.2.2
ORIGIN: https://github.com/Homebrew/brew
HEAD: 705d2564ddfb9c915ee9425dd9ed408bfac61005
Last commit: 3 days ago
Core tap JSON: 05 Jan 02:52 UTC
Core cask tap JSON: 05 Jan 02:52 UTC
HOMEBREW_PREFIX: /opt/homebrew
HOMEBREW_AUTO_UPDATE_SECS: 50400
HOMEBREW_CASK_OPTS: []
HOMEBREW_MAKE_JOBS: 10
HOMEBREW_NO_ENV_HINTS: set
Homebrew Ruby: 3.1.4 => /opt/homebrew/Library/Homebrew/vendor/portable-ruby/3.1.4/bin/ruby
CPU: 10-core 64-bit arm_firestorm_icestorm
Clang: 15.0.0 build 1500
Git: 2.43.0 => /opt/homebrew/bin/git
Curl: 8.4.0 => /usr/bin/curl
macOS: 14.2.1-arm64
CLT: 15.1.0.0.1.1700200546
Xcode: 15.1
Rosetta 2: false

[jdx]

I'm not really sure, but it might be worth at least a glance at the code (migrate_rtx()) if you might have an idea where the logic is falling over

https://github.com/jdx/mise/blob/main/src/migrate.rs

[chadlwilson]

Not directly related but I infer from the below that its not "safe" to migrate Ruby installs either? https://mise.jdx.dev/rtx.html only mentions Python so I manually moved everything except Python, but I can PR to the docs on that too.

mise/src/migrate.rs

Lines 67 to 69 in 4902791

	if plugin == "python" || plugin == "ruby" {
	continue;
	}

[jdx]

someone mentioned that and I suspect it's probably accurate, but I have not tested it to be certain myself. You're right I forgot to update the docs on that one though.

[chadlwilson]

Mine do not seem to be working, so you're probably right. I'll PR the docs to make them consistent with the impl.

Note the dylib reference to old rtx install location from new bin.

dyld[85289]: Library not loaded: /Users/chad/.local/share/rtx/installs/ruby/3.1.4/lib/libruby.3.1.dylib
  Referenced from: <CA8F653C-3665-3821-BB09-6FD8F7C8D6DE> /Users/chad/.local/share/mise/installs/ruby/3.1.4/bin/ruby
  Reason: tried: '/Users/chad/.local/share/rtx/installs/ruby/3.1.4/lib/libruby.3.1.dylib' (no such file), '/System/Volumes/Preboot/Cryptexes/OS/Users/chad/.local/share/rtx/installs/ruby/3.1.4/lib/libruby.3.1.dylib' (no such file), '/Users/chad/.local/share/rtx/installs/ruby/3.1.4/lib/libruby.3.1.dylib' (no such file), '/usr/local/lib/libruby.3.1.dylib' (no such file), '/usr/lib/libruby.3.1.dylib' (no such file, not in dyld cache)

[EamonKeane]

This one confused me for a bit, with psycopg2 referencing an rtx install ~/.local/share/rtx/installs/postgres/15.5/lib/libpq.5.dylib. It was fixed by doing a pip cache purge to clear ~/Library/Caches/pip