Please install Docker according to the docker readme.
In order to initialize, run or attach to containers you will be using the ./exec.bash script.
You can use either base, devel or release as argument in order to determine which image shall be instantiated in your container (or leave out the argument and use the default mode specified in settings.bash):
$ ./exec.bash <mode>
Normally the exec.bash script will drop into a /bin/bash shell with the default mode set in the settings.bash if no other arguments are given.
Each subsequent call to ./exec.bash is opening a new console attached to the same running container.
If the requested image is not yet available locally, it will be pulled from the docker registry. Docker hub is the default registry. For certain images or other registries you might need to login at this point (docker login <registry>).
You can directly run commands in the Docker container. Especially for release this is useful as you only need to run one command on the host computer to execute an application inside the container.
Example:
$ ./exec.bash <command>
command can be any (shell) command found on the client's PATH.
Here, you can also optionally specify the execmode to specify whether you want to run the command with the base, devel or release image.
You can (and probably should) define start scripts to execute one or many applications from the image's workspace.
All scripts from the startscripts folder are automatically added to the PATH in the container.
You can activate some autocomplete hints on the host machine by sourcing $ source autocomplete.me.
This adds all scripts of the startscripts folder to the auto completion feature using TAB after ./exec.bash.
However, if the image was created with a different version of this cloned repository, the startscripts that are available in the container might differ.
In this case autocompletion might not work correctly. A possible workaround is to reset this repo or at least the startscripts directory, to the date when the release image was created.
When you exit (detach from) the container shell, the container keeps running in the background. You can inspect which containers are running at the time via docker ps and stop containers via stop.bash [base|devel|release|CD]. In order to regenerate a fresh container from an image, you can delete the old one by executing delete_container.bash <execmode>. The next call to ./exec.bash will generate and start a new container.
IMPORTANT: In order to apply new settings (like 3d acceleration, or new ADDITIONAL_DOCKER_RUN_ARGS in settings.bash), the container and saved settings have to be reset: delete_container.bash <execmode>
For more advanced topics, see the advanced topics readme.
In case you don't have a registry or your registry does not contain the image, you can build those yourself:
Base images contain the basic dependencies for several projects.
If you need to create them, please look at the base image readme.
Devel images should contain project specific dependencies.
If you are a developer using docker_development, please have a look at the devel image readme.
Release images are ready to run, if you already have one, just start it using ./exec.bash with the correct parameters (see "General Usage").
Please have a look at the release image readme.
Release Images can be exported as .tar.gz file and a run sripts .tar.gz file which can be imported an run on another PC.
See export image readme.