Node.js: Be a good dev: offer dependencies | Hacker News Reader

Node.js: Be a good dev: offer dependencies

medium.com

18 pointsleichtgewicht21 comments

Correct me if I'm wrong, but this would mean that I will be informed about missing dependencies RUNTIME? And I have to do that multiple times, if there are multiple 'dynamic dependencies' like this. The main reason for having declarative dependencies is to NOT be having to do that.

Yes, the missing dependencies would show up at runtime. Which is a valid and important concern, however: It would be possible (though probably inadvisable) to implement an automatic download of missing dependencies.

Which is as terrible idea. You have some code deployed, and suddenly this code downloads random things from internets without any control whatsoever.

Even worse if the application was deployed behind firewall and can't access internet freely.

It is definitely a bad idea when working on a server, but for a cli tool or some editor that might install it anyways it could be arguably worth a try.

It's JavaScript. If you make any thing that's merely silly in command line tools, it will end up being adopted where it's a terribly stupid idea.

Plugins are plugins and dependencies are dependencies. Don't mix them up.

Isn't that the whole reason there is a package.json? That way, people just run npm install and already have all the necessary dependencies.

Yes, a package.json is good to get the required dependencies. This article is about possible dependencies that help in cases where you all the possible would mean a huge load on the system.

If you can use a package without them then they aren't really dependencies are they?

True, how should those things be called?

A good example is Consolidate, which provides a consistent API to various template libraries. You wouldn't want to force installation of all the unused libraries.

https://github.com/tj/consolidate.js

Will add a link to the "other problems" list, thanks!

How is it a huge load on the system given that node can lazy load modules? Do you mean disk space?

Here's some discussion on a related idea:

https://github.com/rails/rails/commit/2dfff4d31634d669f4c198...

In that case, the "offered" dependency was on the mysql2 gem. The intent was similar (don't want to force users of ActiveRecord to install DB adapters they don't use) but ran into an unexpected situation when the generated Gemfile entry didn't have a version constraint but the runtime-loading code did.

Honestly, this sounds like a big step backwards in UX if implemented widely.

I would really rather not have to manually install a million packages when pulling down dependencies - "npm install && npm start" and then going to grab a glass of water feels great.

The given use case of a master parsing library doesn't really resonate. A handful of extra kb on disk doesn't bother me at all for an all-in-one solution, and I've never really had the problem of not being able to find a parsing library on npm to begin with.

You do have to choose which packages you want to install at some point. When you do `npm install && npm start` it will still work as planned. The problem here address is for when you need to deal with a new issue.

If it were a handful of extra kb, I'd agree. Of the top off my head I can think of following file formats "somehow" suited for configuration: json, cson, properties files (java), yaml, xml, conf-files (Apache), ignorefile syntax, ini files.

Supporting all of them could be a reasonable goal to the community (looking at the packages in npm: it is an actual goal already).

https://en.wikipedia.org/wiki/Configuration_file

Use js files for config. They support comments, syntax highlighting will work out of the box, and you can require them.

A problem that I have with my article is that readers seem to get hung up on the config loading bits when I tried to use it as an example for a more-general problem.

To fix this I added a section to the article: https://t.co/sXJhTYXQvy

(Using executable files for configuration can be an enormous problem for some uses.)

I think it's a poor example. I don't want to choose between 20 different config file formats. Give me a config file in a format I know.

He said that approach "d" (implement every implementation) has the problem that "It will take a lot of time to implement all configuration files properly".

Somehow that problem disappears when using his approach, and I think that's even bigger problem than the download size of the dependencies.

I'm not shure that it is a good idea to create (and use) such packages as 'parse-any-structed-file'. If you know what you need, then you can use an optimized implementation.

In my use-case I would like to offer as many structured files formats as possible. But to be sure: you can replace "structured file" with "code syntax" (example: syntax highlighting) or "natural language translation" (example: translation of a cli tool) or "database driver" (example: orm tool).

(In my project that is the exact problem case)

Reply on news.ycombinator.com