This is interesting but I feel like a lot of these Rust-inspired package managers are a little... too inspired by Rust. This project for instance uses .toml as a config file format, presumably because that's what Cargo does.
But I think for this project in particular, Lua for the config files would have been a better choice!
I think that Lua tries to be a good configuration language (it started as a configuration language called SOL (sun), which configured reports for lithology profiles), and in fact Luarocks uses "rockspec" for their config, which is syntactically Lua. Lux claims to be inspired by Luarocks, and yet they chose to use toml over lua for config. I'm wondering why? What was wrong with lua that made toml a better choice?
edit: Okay, I've found more information where they say they support both formats... which, I don't know if that's the right call? Seems like going with one or the other is better from a project management standpoint, although I can see why they want to give users the option.
> Not everyone may want to migrate (nor use) the TOML system for describing a project. For this reason, I’d had liked Lux to support a rockspec file alongside the TOML file (similar to the old project.rockspec format). This has finally been implemented! By creating a file called extra.rockspec in the project root, you will instruct Lux to merge the TOML and the rockspec together when performing any sort of operation.
Nope. We chose TOML as the default for various reasons:
- Simplicity.
There are use cases for a turing complete configuration language.
Lux is not one of them.
- Ergonomics.
The ability to edit it using the CLI (technically, that could be possible with Lua too, but it would be a lot more complex and not a very pleasant UX).
> which, I don't know if that's the right call?
The reason we currently support importing a Lua extra.rockspec is ease of migration for complex projects, e.g. with platform-specific overrides (not yet supported by the TOML spec).
This perplexes me. Lua was conceived as a configuration language and the whole point of a configuration language is you edit a config file. Trying to abstract this away behind a CLI seems like it misses the ethos of Lua.
It’s also a tad strange that a package manager designed for Lua isn’t written in Lua. Presumably Lua developers already have Lua installed, know Lua, and would more likely contribute to a project written in Lua.
> Trying to abstract this away behind a CLI seems like it misses the ethos of Lua.
With `lx add <package>`, you can install the package and add it to und config file in one step. And do things like fail if the package or version doesn't exist or isn't compatible with your system.
You can provide editor plugins or use LSP to give users hints if there's an update available, and use code actions to update them, etc.
> It’s also a tad strange that a package manager designed for Lua isn’t written in Lua.
Again, the fact that Lux relates to Lua is a pretty weak argument for choosing Lua as a language to write or configure it in.
Lots of Lua libraries and packages aren't written in Lua, but are built with Lua bindings.
Lua (which as you yourself just mentioned was conceived as a configuration language) is a pretty poor choice for something with the scope of Lux.
In fact, luarocks was recently rewritten in Teal.
Lux has a Lua API (lux-lua) which means it can be embedded or used as a Lua library.
> Presumably Lua developers already have Lua installed, know Lua, and would more likely contribute to a project written in Lua.
We're not worried about finding contributors. If anything, what we need are high quality contributions. Lua developers who only know Lua are not what we're looking for.
Thanks that does answer my question! Had you considered parsing a subset of lua to get the properties you want? That way users don't have to learn a whole other syntax. I'm thinking in particular of my students whom I teach lua. They struggle enough learning one language, having to teach a second with all its quirks seems like a lot to throw at them.
I wouldn't present it to them as "these config files are the same thing as Lua but without loops", but instead "these are config files and they have the same syntax as the Lua records we just learned about". And I would prefer that over "these are config files and they have a different syntax as the Lua records we just learned about." Although I can see merits on discussing that syntax differences exist between languages, that tends to overwhelm people learning their first language.
That's a neat idea, but it would mean we'd have to maintain our own library.
When editing with the CLI, you have to make sure you preserve comments, which the toml-edit crate does quite well.
> I can see why they want to give users the option
I completely dislike the practice of giving options for no reason other than to give options. Don't make me learn different ways of doing the same thing to succeed in an ecosystem. Don't make me learn differences and similarities. If one way works properly and doesn't have obvious downsides, stick with having one way. If it has obvious downsides, stick with having a different one way. Subjective format taste isn't a real downside. Pick one format and stick with it.
The line from the zen of Python about how "there should be one-- and preferably only one --obvious way to do it" is something that people all too often forget the value of.
> If one way works properly and doesn't have obvious downsides, stick with having one way. If it has obvious downsides, stick with having a different one way.
What if you're in the real world with tradeoffs? So you have both obvious downsides and obvious upsides mixed in each option, and what's more important, those depend on the user, not you, so you can't pick one best option?
That's the reason you give options, and you don't need to learn different ways, learn one you like better or just flip a coin
The irony of quoting the Zen of Python in a discussion about a package manager. I can think of at least three Python package managers off the top of my head.
> The line from the zen of Python about how "there should be one-- and preferably only one --obvious way to do it" is something that people all too often forget.
The zen of Python should be the zen of all languages.
"beautiful", "elegant", and "tasteful" have all been used to puff up various libraries, frameworks, etc, and now we have "luxurious" to add to the long list of ridiculous adjectives used to puff up tech. Lovely.
Lua developers do, indeed, deserve a bit of lugubriousity with regards to describing luxurious things.
I’m all for ‘lux’ as a tool, if it can be used as the tip of the knife that delivers the pearly oyster.
For a lot of Lua projects, there are other extremes, by the way. There are helaciously discomfiting situations with regards Lua package management in certain environments.
If Lua gets something that makes it far, far easier to deploy, that pearl gets fatter.
Honestly it makes me roll my eyes, "let's describe our software utility as if we're trapped in a perfume commercial".
But on the other hand, I think when creating something it does help to have underlying vision, even if it's abstract or doesn't quite make sense.
I think that is why despite it being eye roll inducing, there is still value to these descriptors as it explains what they are going for. In this case it tells us they are prioritizing the feel over everything else and for a package manager that is pretty solid focus.
As a Lua dev, I’m pro “anything that makes Lua easier to use/deploy”, and while do I see “lux” from the perspective of precocious naming attempts, I can also immediately overlook it by treating “LUX” as more of an acronym for “Lua User Experience”, so if it is actually able to deliver that promise, the eyes are gonna stop rolling and start focusing on using it ..
Right, my bad.
Still, being able to do more to aid the creation and maintenance of packages than just install packages doesn't make something "not a package manager".
When I tried using Gleam, I loved that it came with all the basic tooling I needed and that's what I think is so wonderful about Lux. I don't want to spend my time fiddling around with setting up all the individual tools — I just want to write code. For me, Lux makes the broader experience around building Lua projects a lot more enjoyable.
If I can get lux to deal with the package management scenarios around a few turboLua projects, I’m pretty sure I’m going to ship much more Lua code next year.
This is interesting but I feel like a lot of these Rust-inspired package managers are a little... too inspired by Rust. This project for instance uses .toml as a config file format, presumably because that's what Cargo does.
But I think for this project in particular, Lua for the config files would have been a better choice!
I think that Lua tries to be a good configuration language (it started as a configuration language called SOL (sun), which configured reports for lithology profiles), and in fact Luarocks uses "rockspec" for their config, which is syntactically Lua. Lux claims to be inspired by Luarocks, and yet they chose to use toml over lua for config. I'm wondering why? What was wrong with lua that made toml a better choice?
edit: Okay, I've found more information where they say they support both formats... which, I don't know if that's the right call? Seems like going with one or the other is better from a project management standpoint, although I can see why they want to give users the option.
> Not everyone may want to migrate (nor use) the TOML system for describing a project. For this reason, I’d had liked Lux to support a rockspec file alongside the TOML file (similar to the old project.rockspec format). This has finally been implemented! By creating a file called extra.rockspec in the project root, you will instruct Lux to merge the TOML and the rockspec together when performing any sort of operation.
> presumably because that's what Cargo does.
Nope. We chose TOML as the default for various reasons:
- Simplicity. There are use cases for a turing complete configuration language. Lux is not one of them.
- Ergonomics. The ability to edit it using the CLI (technically, that could be possible with Lua too, but it would be a lot more complex and not a very pleasant UX).
> which, I don't know if that's the right call?
The reason we currently support importing a Lua extra.rockspec is ease of migration for complex projects, e.g. with platform-specific overrides (not yet supported by the TOML spec).
> The ability to edit it using the CLI
This perplexes me. Lua was conceived as a configuration language and the whole point of a configuration language is you edit a config file. Trying to abstract this away behind a CLI seems like it misses the ethos of Lua.
It’s also a tad strange that a package manager designed for Lua isn’t written in Lua. Presumably Lua developers already have Lua installed, know Lua, and would more likely contribute to a project written in Lua.
> Lua was conceived as a configuration language
That alone is a pretty weak argument.
> Trying to abstract this away behind a CLI seems like it misses the ethos of Lua.
With `lx add <package>`, you can install the package and add it to und config file in one step. And do things like fail if the package or version doesn't exist or isn't compatible with your system.
You can provide editor plugins or use LSP to give users hints if there's an update available, and use code actions to update them, etc.
> It’s also a tad strange that a package manager designed for Lua isn’t written in Lua.
Again, the fact that Lux relates to Lua is a pretty weak argument for choosing Lua as a language to write or configure it in.
Lots of Lua libraries and packages aren't written in Lua, but are built with Lua bindings. Lua (which as you yourself just mentioned was conceived as a configuration language) is a pretty poor choice for something with the scope of Lux. In fact, luarocks was recently rewritten in Teal. Lux has a Lua API (lux-lua) which means it can be embedded or used as a Lua library.
> Presumably Lua developers already have Lua installed, know Lua, and would more likely contribute to a project written in Lua.
We're not worried about finding contributors. If anything, what we need are high quality contributions. Lua developers who only know Lua are not what we're looking for.
Thanks that does answer my question! Had you considered parsing a subset of lua to get the properties you want? That way users don't have to learn a whole other syntax. I'm thinking in particular of my students whom I teach lua. They struggle enough learning one language, having to teach a second with all its quirks seems like a lot to throw at them.
Do you think that is more difficult than explaining to the students why they can’t use loops in their lua config files?
I wouldn't present it to them as "these config files are the same thing as Lua but without loops", but instead "these are config files and they have the same syntax as the Lua records we just learned about". And I would prefer that over "these are config files and they have a different syntax as the Lua records we just learned about." Although I can see merits on discussing that syntax differences exist between languages, that tends to overwhelm people learning their first language.
That's a neat idea, but it would mean we'd have to maintain our own library. When editing with the CLI, you have to make sure you preserve comments, which the toml-edit crate does quite well.
> I can see why they want to give users the option
I completely dislike the practice of giving options for no reason other than to give options. Don't make me learn different ways of doing the same thing to succeed in an ecosystem. Don't make me learn differences and similarities. If one way works properly and doesn't have obvious downsides, stick with having one way. If it has obvious downsides, stick with having a different one way. Subjective format taste isn't a real downside. Pick one format and stick with it.
The line from the zen of Python about how "there should be one-- and preferably only one --obvious way to do it" is something that people all too often forget the value of.
> If one way works properly and doesn't have obvious downsides, stick with having one way. If it has obvious downsides, stick with having a different one way.
What if you're in the real world with tradeoffs? So you have both obvious downsides and obvious upsides mixed in each option, and what's more important, those depend on the user, not you, so you can't pick one best option?
That's the reason you give options, and you don't need to learn different ways, learn one you like better or just flip a coin
The irony of quoting the Zen of Python in a discussion about a package manager. I can think of at least three Python package managers off the top of my head.
Not to mention they had a major release that (from the sidelines) looked like "throw away everything you know and learn a new way to do things."
The zen itself is correct even when Python doesn't follow it.
> The line from the zen of Python about how "there should be one-- and preferably only one --obvious way to do it" is something that people all too often forget.
The zen of Python should be the zen of all languages.
Unfortunately in practice you see snowflake implementations left and right in python.
It’s still unclear to me if python is too expressive for its own good, or if it’s so widely used, that it’s impossible to avoid nonsense
I assume like 70% of Python developers do not even know the Zen of Python, let alone PEP-8. It's likely the size.
Yes, it should be. Sadly it's not even unambiguously the zen of Python these days.
Yes, one way to do it. Sync or async!
All I'm hearing is that Erlang is the true Zen.
Good thing we're not giving options for no reason other than to give options ;)
[flagged]
TOML is seriously ugly... nested configs are less clear in toml compared to even YAML.
Okay, sounds good, let’s see:
>+ Create and manage Lua projects
Yay!
>- Easily manage dependencies, build steps and more through the lux.toml file.
Boo! Ermm .. .toml?
You lost me at .toml.
Why is .toml being used to configure Lua?
(Disclaimer: I am a long term Lua user/developer and have a strong opinion about all things Lua.)
https://news.ycombinator.com/item?id=45626961#45630063
"beautiful", "elegant", and "tasteful" have all been used to puff up various libraries, frameworks, etc, and now we have "luxurious" to add to the long list of ridiculous adjectives used to puff up tech. Lovely.
If the biggest flaw in a project is that the name sounds pretentious to some people, it's probably doing something right
It’s okay, you know why? Because this is Lua.
Lua developers do, indeed, deserve a bit of lugubriousity with regards to describing luxurious things.
I’m all for ‘lux’ as a tool, if it can be used as the tip of the knife that delivers the pearly oyster.
For a lot of Lua projects, there are other extremes, by the way. There are helaciously discomfiting situations with regards Lua package management in certain environments.
If Lua gets something that makes it far, far easier to deploy, that pearl gets fatter.
It's just a silly pun. Search for "moon illuminance" and perhaps you'll get it :)
I have always thought that those qualities should be shown, not said. Both in software and in life.
Honestly it makes me roll my eyes, "let's describe our software utility as if we're trapped in a perfume commercial". But on the other hand, I think when creating something it does help to have underlying vision, even if it's abstract or doesn't quite make sense.
I think that is why despite it being eye roll inducing, there is still value to these descriptors as it explains what they are going for. In this case it tells us they are prioritizing the feel over everything else and for a package manager that is pretty solid focus.
As a Lua dev, I’m pro “anything that makes Lua easier to use/deploy”, and while do I see “lux” from the perspective of precocious naming attempts, I can also immediately overlook it by treating “LUX” as more of an acronym for “Lua User Experience”, so if it is actually able to deliver that promise, the eyes are gonna stop rolling and start focusing on using it ..
I don’t understand why a package manager needs lint support.
Because despite what the tagline says it's not a package manager, but a project manager.
Lux helps you install and create/maintain packages. Linting is a useful step in the creation of packages.
Pip lets you create virtual environments. Does that mean it's an environment manager, not a package manager?
(╭ರ_•́)
It doesn't, no? You create virtual environments using Python's venv module, not pip. The newer alternatives like uv do handle it, though.
Right, my bad. Still, being able to do more to aid the creation and maintenance of packages than just install packages doesn't make something "not a package manager".
It's like a package manager on steroids!
When I tried using Gleam, I loved that it came with all the basic tooling I needed and that's what I think is so wonderful about Lux. I don't want to spend my time fiddling around with setting up all the individual tools — I just want to write code. For me, Lux makes the broader experience around building Lua projects a lot more enjoyable.
I’ve come to using turboLua as my main Lua ‘Swiss army tool’, since it comes with so many things built-in, on top of a fairly functional luajit 2.0.
https://turbo.readthedocs.io/en/latest/
If I can get lux to deal with the package management scenarios around a few turboLua projects, I’m pretty sure I’m going to ship much more Lua code next year.
I recently added a linter to my Common Lisp package manager, ocicl. It seemed like a natural place to put it.
I've been awaiting the rocks.nvim team to migrate to this