Forked from fossabot/pgmodeler-x11 2019
Build and run the latest version of pgModeler inside a Docker container on Linux (& macOS) - 0.9.4-beta1.
This project is also an example of creating a build environment that does cross-platform UI compilation from inside Docker, compiling using GTK and Rust. i.e. it is a mechanism to build Linux and macOS binaries in Docker.
-- https://github.com/mvniekerk
-- https://github.com/bank-builder
Just cut and past this at the command line:
docker run --rm -it --user $(id -u) -e DISPLAY=unix$DISPLAY --workdir=$(pwd) --volume="/home/$USER:/home/$USER" --volume="/etc/group:/etc/group:ro" --volume="/etc/passwd:/etc/passwd:ro" --volume="/etc/shadow:/etc/shadow:ro" --volume="/etc/sudoers.d:/etc/sudoers.d:ro" -v /tmp/.X11-unix:/tmp/.X11-unix cybermint/pgmodeler
Further Docker only instructions (ie not building or source code required) can be found here.
- Because building pgModeler is harder than it should be.
v0.9.4-beta1 has some significant changes on previous versions, and some of these bits of advice may longer apply, however thanks to Michael we have a working docker release.
-
Because most distros are behind in their pgModeler version and/or don't have an easy way of getting the latest release - this gives another docker-based option to easily get a modern up-to-date pgModeler version.
-
This allows you to easily switch between different pgModeler versions without clobbering your config directory.
Change the PGMODELER_VERSION in the container to build another version to be run, add PGMODELER_VERSION as an environment variable when running the launcher image to start a different run image.
-
Because having a base Docker pgModeler image to be shared between users on all platforms helps address the security concerns raised by running binaries others have built. A company (such as a bank) can compile its own image and can audit how it was built and how it was stored - banks are funny that way!
-
Because having a reproducible way of building GTK applications for the three major desktop platforms using Docker and Rust rocks!
pgModeler provides pre-compiled binaries at a small price. While we are providing an easy way to build and run pgModeler, we will not be changing the scripts and images provided to build stable versions (only development versions) - this is so as not to erode their (paying) customer base.
We will also not be providing support.
Further, as you're essentially running pgModeler on Linux, it will look and feel like Linux. Which is nice if you're a Linux user, not so great if you're a Mac or Windows user - kindly buy the prebuilt binaries from the pgModeler project if this is a problem.
Please ensure you have read and understood the Xcode license terms before continuing.
Building for macOS should be done on Apple hardware!
The launcher compiles and links inside a container that has Macports installed. The binary is target/macos-x86_64/x86_64-apple-darwin/release/launcher. In order to be able to run this, you need to install gtk3 using either Brew or Macports. This will change in the future as we aim to make an application launcher to be put in your /Applications directory.
[TODO implement https://wiki.gnome.org/Projects/GTK/OSX/Bundling]
- Docker
- Docker
- XQuartz
In the container directory, run ./build.sh.
This will create an image: pgmodeler-docker-x11/run.
This will take ~15 minutes.
As service to the community Cyber-Mint makes this available on docker hub -
git pull cybermint/pgmodeler
** Currently Linux Only **
In the launcher directory, run ./build.sh <platform> where platform is linux, mac.
Only Linux actually installs it now (for now). For macOs the binary is built but then needs to be run from the command line and have gtk3 installed (see "macOS Packaging").
When the Linux build finishes, it will create a /usr/share/applications/pgmodeler-docker-x11.desktop entry, which means you can run pgModeler from your application launcher. See that wonderful elephant in your launcher!
Note that if you wish to use the docker-compose found here, it needs the following to be set:
# add exports to ~/.bashrc
export USERID=$(id -u)
export WORKING_DIR=$(pwd)
The launcher looks at your environment and then runs the command below and lets you know if something goes wrong.
But nothing is keeping you from doing it yourself. NOTE: this can/will clobber your config if you're not using 0.9.2-alpha1.
docker run --rm -it --user $(id -u) -e DISPLAY=unix$DISPLAY --workdir=$(pwd) --volume="/home/$USER:/home/$USER" --volume="/etc/group:/etc/group:ro" --volume="/etc/passwd:/etc/passwd:ro" --volume="/etc/shadow:/etc/shadow:ro" --volume="/etc/sudoers.d:/etc/sudoers.d:ro" -v /tmp/.X11-unix:/tmp/.X11-unix cybermint/pgmodeler
You need XQuartz installed, and 'Allow connections from network clients' selected in the X11 preferences.
Run the following two commands:
xhost + 127.0.0.1 docker run --rm -it -e DISPLAY=host.docker.internal:0 --workdir=$(pwd) -v /Users/{your_user_folder}:/home/{your_user_folder} --volume="/etc/group:/etc/group:ro" --volume="/etc/passwd:/etc/passwd:ro" --volume="/etc/sudoers.d:/etc/sudoers.d:ro" -v /tmp/.X11-unix:/tmp/.X11-unix -v $(pwd):$(pwd) cybermint/pgmodelerNote: Replace {your_user_folder} with the literal name of your home folder. Do not use the $USER variable.
© 2019, Grindrod Bank Limited. © 2020, Cyber-Mint (Pty) Ltd