Enthought Deployment Manager (EDM)

28 Mar Enthought Deployment Manager (EDM)

Published by on March 28th, 2017 9:10pm

The Next Generation Python Environment and Package Manager

Enthought Deployment Manager, or EDM, is Enthought’s new environment and package manager. With EDM you can quickly create multiple self-contained Python environments and install and manage sets of Python packages with consistent dependencies. The “Deployment” in Enthought Deployment Manager indicates that EDM is managing more than just packages or distributions; the long-term view for this tool is to also deliver applications.

Environment Management

EDM creates self-contained environments with their own Python executable and set of packages. This allows users to reproduce a colleague or partner’s environment for co-development, troubleshooting, or testing or to work on different projects with different sets of dependencies.

Package Management

Dependency information is included in Enthought’s curated packages to allow EDM’s dependency solver to create sets of reliably consistent packages. This allows users to have one set of package versions in a known production state while developing an experimental state for the next release.

Enthought Deployment Manager Features

EDM implements the best principles of virtualenv managers such as virtualenvwrapper and pew as well as the Ruby Version Manager (rvm), but avoiding the pitfalls of virtual environments by providing robust single-layer Python environments similar to legacy EPD. It also derives ideas from Nix.

The Enthought Deployment Manager provides:

  • The ability to create multiple self-contained environments
  • A state-of-the-art dependency solver with support for resolving conflicts and other features available in advanced package managers, such as the ability of multiple packages to satisfy a requirement (the “PROVIDES” clause), etc.
  • A well-defined deployment API useable by third party programs (e.g. IDEs, GUIs)
  • A small distribution footprint, ~10 Mb single file binary
  • Native installers for Windows (.msi), Mac OS X (.pkg), RHEL/Fedora (.rpm), and Debian/Ubuntu (.deb)
  • Support for multiple Python versions (both Python 2 and Python 3), implementations (cpython, pypy) and even other languages (Julia)

Legacy EPD users will recognize the edm command line tool as an update to Enstaller’s enpkg.

Download Enthought Deployment Manager

Below are the available downloads for the latest version of EDM (currently 1.8.2). Please download the proper package for your operating system.

See full documentation.

Operating SystemFile FormatInstaller*
Mac OS X.pkg64-bit .pkg installer
Windows 7 & later.msi64-bit .msi installer
Debian 7 and above.deb64-bit .deb package
Centos 5 and above.rpm64-bit .rpm package
Generic glibc Linux.sh64-bit .sh installer

* EDM itself is a 64-bit program, but it can create and maintain both 32-bit and 64-bit versions of Enthought Python
* The .deb and .rpm require admin privileges (e.g. sudo), the sh installer does not. The .rpm and .deb install edm in /opt/edm, and symlink the edm executable in /usr/bin.

Getting Started

Below are examples of how to get started with Enthought Deployment Manager on your machine:

# Make sure edm is installed and available
$ edm --version

# Install package of your choice, e.g. NumPy
# (Creates a default Python environment)
$ edm install numpy

# Create a shell with the default environment activated
$ edm shell

How to install in a different environment:

$ edm install -e test-old-numpy "numpy < 1.10"

# Create a shell with the test-old-numpy environment activated
$ edm shell -e test-old-numpy

For more information or to get help:

# See a list of available commands and help topics
$ edm help

# Get help on a specific command, e.g. install
$ edm help install

# Get help on a topic, e.g. configuration
$ edm help configuration

See full documentation here.

No Comments

Sorry, the comment form is closed at this time.