Depy

Depy is a dependency manager based on scoop. It automatically installs any program from a collection of over 3000+ packages.

Depy automatically manages and creates virtual environemnts similar to python -m venv that help you manage versions of apps cleanly.

A bit of documentation about the functions can be seen here if you are planning to use depy as a library

Instalation

If you already have scoop and git installed just open up powershell and run:

scoop install https://raw.githubusercontent.com/insertokername/depy/main/manifest/depy.json

If you don't have scoop installed there are just a couple of steps:

• for 64 bit First make sure you have VCRUNTIME

  And then open powershell and run:

  Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
  Invoke-RestMethod -Uri https://get.scoop.sh | Invoke-Expression
  scoop install git
  scoop install https://raw.githubusercontent.com/insertokername/depy/main/manifest/depy.json
  

• for 32 bit: Download git, make sure you have VCRUNTIME And then open powershell and run:

  Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
  Invoke-RestMethod -Uri https://get.scoop.sh | Invoke-Expression
  scoop install https://raw.githubusercontent.com/insertokername/depy/main/manifest/depy.json
  

Now you're done!

Compilation

Compilation dependencies:

• only rust

First install rust from rustup. After that download the source code from github. Finally open the in the terminal and run cargo build. The project will be compiled under target/debug/depy-cli.exe and target/debug/depy-gui.exe.

Usage

Gui

Open a terminal in the folder you wish to make a virtual environment in, run the command depy-gui and now the interface will open, you can now search the packages you wish to install. Once you find you pacakge just click on add install. Once you are done with selecting your needed packages just Install added packages and you're done!

Cli

First we need to create a depy.json file that will declare all of our desired programs.

depy.json

[
    {
        "bucket_url": "https://github.com/ScoopInstaller/Main",
        "bucket_name": "main",
        "name": "grep",
        "version": "latest"
    },
    {
        "bucket_url": "https://github.com/ScoopInstaller/Versions",
        "bucket_name": "versions",
        "name": "python36",
        "version": "3.6.8"
    },
]

In this depy.json file we declared that we need to install grep from the main bucket, by using the latest tag it will always be set to the latest version available when updating with depy , and python version 3.6.8 from the versions bucket. Buckets are explained here.

After we define our depy.json we can just run:

depy

And everything will install automatically without polluting the environment PATH(in most cases).

Last thing we need to do is run:

.depyvenv\activate

Now all of our apps will be available to us and (CURENTLY IN DEV SHELL) should appear in front of the prompt. Until you close the shell you will have access to all the specified versions of the programs. On subsequent runs you will only need o run the activate script.

Some more important info

This installation system will have conflicts if you are using the scoop "use_isolated_path" config, very recomended you unset it if you are using it.

This application does not cause conficts with existing scoop installed programs.

Buckets

Buckets are repositories that contain a number of packages. Each bucket has its purpose:

• main: contains most cli applications and frameworks and etc
• extras: contains mostly gui apps and apps that don't fit the main bucket criteria
• versions: contains older versions of packages that have changed their install system and require special instructions compared to the newer packages.

You can also make your own custom buckets

depy.json

The format of a depy.json looks like this:

[   //an array of "packages"
    
    {   //each package is defined by:
        "bucket_url":"https://github.com/ScoopInstaller/Main"   //the github repository in which it is stored
        "bucket_name":"main"                                    //an identifier for that bucket
        "name":"generic_package",                               //the package name
        "version":"1.0",                                        //the package version
    }
]

• The name and version section are self explanatory.
• The bucket_url represents the repository in which to search for the package in.
• The bucket_name represents a label for that bucket.
• The versions can be set to latest so that every time you run depy or update via gui, it will automatically get the latest versions available in scoop.

Adding a custom bucket:

First you should read a bit about scoop buckets from here. The proceed to use the bucket template repo and make sure your bucket directory is on the master branch. After that you can just put the link to you repository in the bucket_url argument of the program that requires it (something like this: "bucket_url":"https://github.com/ScoopInstaller/Extras")

Garbage clean

The garbage clean command uninstalls all packages from the cache, freeing up some space.

Cli

depy -g

Gui In the garbage clean menu, hit the Cleanup Packages button

Uninstalling

Uninstalling is done in two steps:

Cli:

depy -d

Gui: In the garbage clean menu hit the Uninstall Depy button.

This uninstalls all apps under %userprofile%/depy/scoop.

Second:

scoop uninstall depy

This just uninstalls the exe itself which is very small.

Curently working on:

As of now if you install an app that has a startmenu shortcut the startmenu will be polluted. To combat this i have made a pull request to the scoop repo with a new flag that allows apps to install a package without polluting the startmenu, curently waiting for aproval.

Resources used

While making this project I mainly had to read the scoop wiki and some crate documentations like for example druid