There's TODOs all over the documentation! There's no background task tools for scripting, and in interactive use background tasks are barely supported - an issue about background tasks has people going roughly "nobody needs to do tasks in parallel, that was only important when people were working on mainframes". The shell hooks have (undocumented) weird restrictions. Lazy iteration of lists is only supported by using functions with callbacks. Stable = development appears stopped.
This is half baked and dead. For my new computer I really really wanted a lightweight new shell with orthogonal syntax thought out from the ground up and not glued together over 4 decades, and this seemed like the closest option! But this isn't it.
However if you’re looking for an alternative then there’s:
- Murex (disclaimer: I’m one of the maintainers) which does support background processes and has extensive documentation. https://murex.rocks
- Nushell: I’m not personally a fan of its design choices but it has a large following of people who do really enjoy it so it might also appeal to yourself too.
As for Elvish, I do encourage others to give it a go themselves. It’s really well thought out and what might be a deal breaker for some people isn’t for others.
Nushell also had very minimal background task support, so I rejected that. They explicitly say use some other program for background tasks in their docs.
I actually looked at Murex after seeing it in previous threads, but I bounced for some reason... I just took another look though skipping the tutorial and I see you have `bg` and `fg` support! But does `bg` return the `fid`? Can you use those in scripts, or are they hobbled the same way bg/fg are in bash?
It's been a good 4-5 months since I went down this rabbit hole, but IIRC the basic things I wanted to do and got blocked in multiple shells were:
- System-wide interactive-use config file, I use Nixos and manage my system config using that
- Background task support - I need to start an ssh tcp proxy, pipe a command over it, then kill ssh once the command is done (all in a script).
- Post-command hook, to send a notification when a long command finishes
- Async iteration of command output, i.e. streaming events with swaymsg subscribe and running a command when certain events occur
- Value/call arity safety - i.e. a clear distinction between a single value and multiple values that doesn't rely on stringification hacks. I.e. in `command $x` `command` should always have one argument, regardless of the contents of `x`, and making that plural should be explicit.
And then other standard evaluation criteria, like I looked at xonsh but it seemed like a massive hack, despite handling a lot of the above.
There's two kinds of process IDs in Murex: FID (function IDs) and PID (process IDs)
Forking is expensive in POSIX and has a number of drawbacks such as the inability to share scoped variables without resorting to environmental variables. So a FID is basically a PID but managed inside the scope of Murex's runtime. You can manage FIDs in much the same way as you can manage PIDs, albeit using Murex builtins rather than coreutils (though PID management tools in Bash are technically builtins rather than coreutils too).
What this means in practice is you can have entire blocks of code pushed into the background, eg
» GLOBAL.name = "rendaw"
» bg { sleep 5; echo "Hello $name" }; echo "not bg"
not bg
Hello rendaw
» jobs
JobID FunctionID State Background Process Parameters
%1 2109 Executing true exec sleep 5
fid-kill 2109
» ps aux | grep sleep
hnlmorg 72749 0.0 0.0 410743712 1728 s012 S+ 4:24p.m. 0:00.00 /usr/bin/grep --color=auto sleep
hnlmorg 72665 0.0 0.0 410593056 432 s012 S+ 4:23p.m. 0:00.00 /bin/sleep 5
While the above seems very complicated, the advantage is that `bg` and `fg` become much more script friendly.
> - System-wide config file, I use Nixos and manage my system config using that
This isn't Murex's default behaviour but you could easily alter that with environmental variables: https://murex.rocks/user-guide/profile.html#overriding-the-d...
The latest version of Murex (v7.0.x), which is due to be released in the next few days, makes this even easier with a $MUREX_CONFIG_DIR var that can be used instead of multiple specific ones.
> - Background task support - I need to start an ssh tcp proxy, pipe a command over it, then kill ssh once the command is done (all in a script).
Murex has another layer of support for piping in addition to those defined in POSIX, which are basically channels in the programming language sense. In murex they're called "Murex Named Pipes" but the only reason for that is that they can be used as a glue for traditional POSIX pipes too. This is one area where the documentation could use a little TLC: https://dev.murex.rocks/commands/pipe.html
> - Post-command hook, to send a notification when a long command finishes
There are two different events you can hook into here:
- onPrompt: https://dev.murex.rocks/events/onprompt.html
This is similar to Bash et al prompt hooks
- onCommandCompletion: https://dev.murex.rocks/events/oncommandcompletion.html
This hooks into any command name that's executed. It runs the comment in a new TTY and buffers the command's output. So, for example, if you want a command like `git` to automatically perform a task if `git push` fails with a specific error message, then you can do that with onCommandCompletion.
> - Async iteration of command output, i.e. streaming events with swaymsg subscribe and running a command when certain events occur
I'd need to understand this problem a little more. The channels / Murex Named Pipes above might work here. As might onCommandCompletion.
> - Value/call arity safety - i.e. a clear distinction between a single value and multiple values that doesn't rely on stringification hacks. I.e. in `command $x` `command` should always have one argument, regardless of the contents of `x`, and making that plural should be explicit.
This one is easy: scalars are always $ prefixed whereas arrays are @ prefixed. So take the following example:
array = %[ a b c ]
» echo $array
["a","b","c"] # a single parameter representation of the array
» echo @array
a b c # the array expanded as values
This is quite a lengthy post but hope it helps answer a few questions
you are right about lack of support for job control, it's annoying. but my understanding is that the problem seems to be a difficulty in implementing job control with go. when people say nobody needs parallel tasks that doesn't make sense because you can run jobs in the background. you just have to do it explicitly before starting, and you can't switch back and forth. yes, that's a problem, and for me it is one of the most annoying missing features. but it comes up seldom enough that it doesn't disrupt daily use for me. which is to show that the things i need for daily use are all there.
Just to add some further qualification, I was fully prepared to learn something from the ground up, throw away all my preconceptions, and give some weirdness a try - including no string interpolation. I wanted to 100% replace bash, both as a shell and for scripting everywhere. I was exactly Elvish's target user.
Nine Reasons to Use OSH - https://oils.pub/osh.html - it runs existing shell scripts, ...
What is YSH? - https://oils.pub/ysh.html - It's the ultimate glue language, like shell + Python + JSON + YAML, seamlessly put together
string interpolation is useful where concatenating strings requires an operator, but i don't see the benefit otherwise.
for more complex examples i can use printf, or someone could write a function that does string interpolation. since there is no need to fork, that should not be that expensive
Ultimately I've found that for my interactive shell I just want something widely supported and with easy syntax for `if` and `for` loops for short multi-line commands. For anything longer than that I just reach for real Python using either the `sh` or `plumbum` package.
I don't need the extra features very often, so I just run things in full Python where I'm already comfortable.
I've tried oils/ysh, elvish, xonsh, nushell, and while they are _fine_ I don't want to learn a different language that's not quite Python and not quite shell.
> While Elvish provides a lot of useful functionalities as builtin commands, it can’t do everything. This is where external commands come in, which are separate programs installed on your machine. Many useful programs come in the form of external commands, and there is no limit on what they can do. Here are just a few examples: Git provides the git command to manage code repositories
At first I thought, wait, is this a shell or not, do I have to write code or something to get access to normal external commands? But no, this is more like going to a car dealership and having the salesman say "Hey thanks for coming by, a car is a mechanical device consisting of metal parts and rubber parts for the purpose of taking you where you need to go! Now that we're on the same page about that, money is a thing made of paper for the purposes of .."
Docs are hard, once or twice is fine, but lots of parts are like this and I gave up reading. Not sure if it's AI generated, but if the project is doing that then it should stop, and if it's not doing that it should consider starting to
They have a separate set of docs for people who do have some experience with other shells (https://elv.sh/learn/); you may find the quick tour more suitable for your speed: https://elv.sh/learn/tour.html
> Line continuation in Elvish uses ^ instead of \
> Bash: echo .[ch] vs Elvish: echo .?[set:ch]
One more example, guess what this does: `echo &sep=',' foo bar`. Is it bash, elvish? Some combination of the two with markdown? Legal in all three? Elvish certainly cleans up conditionals and stuff, but you probably shouldn't introduce new things with exactly the same name unless you've created a genuine superset/dialect where the older version still works without rewrite. Namespace it as elvish.echo or use your module system. Shadows aren't friendly, this is equivalent to the guy that monkey-patches sys.stderr=sys.stdout to work around their one-off problem
The thing I like about Nushell is it does away with some of the things that I found hard with bash, and made data formats a first class citizen (something I enjoyed about powershell).
I think if you like Lisp elvish would be ideal but for me the lack (seeming, I've not done a deep dive on the docs) of built-in data parsing is a no.
I've been using fish for the last year or more, and I like some of the "batteries included", particularly the predicting of the command you want to run. But fish is too much like bash in syntax, meaning that I just think of it like bash until I have to type "(foo)" instead of "$(foo)", or "end" instead of "fi". The zsh plugins for doing command predicting and fancy prompt seems to get me all the fish benefits with none of the rough spots. And, frankly, the changes fish does doesn't seem to have any benefit (what is the benefit of "end" over "fi").
Even xonsh (I'm a huge Python fan) doesn't really have enough pull for me to stick in it. Oils, nu, elvish, they all have some benefits for scripting, but I can't see myself switching to them for interactive use.
It's kind of feeling like zsh is "good enough" with no real downsides. Maybe this is mostly that I've been using sh/ksh/bash/zsh for 40 years, some of these other shells might be easier to switch to if you lack the muscle memory?
Note that fish does also support bash's "$(foo)" syntax and has for a few years now.
(Ftr, I've been using zsh for maybe 5-8 years, managed to avoid oh-my-zsh, and only use 'zsh-autosuggestions' and 'zsh-syntax-highlighting' plugins. I've customised a theme to suit me, but barely know anything about zsh to be honest...)
Anyone have any experience of both?
The Nushell and Elvish scripting languages are similar in many ways. I personally find the "shell" experience better in Nushell than Elvish.
Nushell
- Bigger community and more contributors
- Bigger feature set than Elvish
- Built in Rust (Yay :-)! )
Elvish
- Mostly developed by one person
- Built in golang
- Amazing documentation and general attention to detail
- Less features than Nushell
- Feels more stable, polished, complete than Nushell. Your script written today more likely to work unaltered in Elvish a year down the line. However this is an impression. Nushell must have settled down since I last looked at it.
For "one off" scripts I prefer Elvish.
I would recommend both projects. They are excellent. Elvish feels less ambitious which is precisely why I like it to write scripts. It does fewer things and I think does them better.
Nushell feels like what a future scripting language and shell might be. It feels more futuristic than Elvish. But as mentioned earlier both languages have a lot of similarities.
> Built in golang
Does that matter?
If you intend to be a contributor, of course the chosen language matters, but only a very small proportion of users will be contributors.
Rust tends to be marginally faster and compile to smaller binaries.
Go projects tend to hit maturity faster and develop quicker.
Its a relevant factor to quickly stereotype certain characteristics of development, but its not anywhere close to important.
The language choice has certain implications and I would say Rust & Go have fairly similar implications: it's going to be pretty fast and robust, and it'll have a static binary that makes it easy to install. Implications for other languages:
C: probably going to have to compile this from source using some janky autotools bullshit. It'll be fast but segfault if you look at it funny.
Python: probably very slow and fragile, a nightmare to install (less bad since UV exists I guess), and there's a good chance it's My First Project and consequently not well designed.
I have not really had problems with installing C (on the rare occasions I have compiled anything of any complexity) nor Python applications. Xonsh is supposed to be pretty good and written in Python, and most existing shells (bash, zsh, csh etc.) are written in C.
Amusing aside, I use fish and until I decided to fact check before adding it to the list of shells written in C, I did not realise it was written in Rust.
There's a few obvious features missing in fish like backgrounding an alias or an equivalent to set -e, other than that I have no complaints.
The first thing I do on any machine is install fish.
But for writing scripts I would reach for Elvish/Nushell. More powerful.
I don't care much for the Elvish shell experience, rather I like the Elvish scripting language. The documentation is top notch and the language evolves slowly and feels stable.
It's a shell, aren't those two things supposed to be the same basically? Or - do you mean the interaction with the terminal/command-line?
The language is what you write in an $EDITOR. Here Elvish scripts can be nice, succinct and powerful. I like how I don't have to worry about strange "bashisms" like argument quoting etc. Everything feels consistent.