• Download Mastercode
• Download and Launch the Mastercode container
• Compile & Run Mastercode in the container
• Remarks

We are maintaining a container prepared with all the libraries and tools required to run Mastercode. This is the Dockerfile for the Mastercode container in case you want to see in particular which software is installed.

1. Download Mastercode

Download or Clone using the git tools the last version of Mastercode, on your Machine, from:

https://github.com/mastercode-collaboration/mcpp

and uncompress the zip file mcpp-master.zip, for example, in your projects directory (/Users/youruserid/projects/). You will need about 1GB of space.

mkdir projects/

cd projects/

mv ../Downloads/mcpp-master.zip .

unzip mcpp-master.zip

You will end up with a directory called ./projects/mcpp-master/ under you main directory.

2. Download and Launch the Mastercode Container

Launch the Docker application if you have not already done so, and open a new terminal on the Mac. If Docker was not running previously, you should always open a new terminal for the docker commands to be available.

In this terminal, download the Mastercode container by:

docker pull indigodatacloud/docker-mastercode

Then you can simply launch the container mounting your Mac $HOME directory on the /mnt directory of the container:

docker run -ti -v /Users/youruserid:/mnt indigodatacloud/docker-mastercode

You may as well mount only the Mastercode directory, bearing in mind that nothing else from your $HOME will be available inside the container:

docker run -ti -v /Users/youruserid/projects/mcpp-master:/mnt indigodatacloud/docker-mastercode

You will end up in a terminal, in which you will appear as root (with the "#" symbol). Do not be impressed by that: it is not a fake root, you will not be altering anything.

Type

"cat /etc/fedora-release"

and convince yourself that you are now on Linux. Typing:

"ls /mnt"

you will see your $HOME directory, or the directory you decided to mount on the "-v" option.

3. Compile & Run Mastercode in the container

On the container terminal you can now compile and run Mastercode as usual:

cd /mnt

cd projects/mcpp-master

./configure

make

.....

Will take a while to compile the first time (order of 20 minutes on recent Mac models), as it downloads all the software tools of the different components, and compiles from scratch. When compilation is done, you can already run it:

./mc_point.py --run-mode mc-cmssm --predictors all --inputs 500 600 0 10 --print-mc-spectrum
4. Remarks

In the typical situation you will be editing your local copy of Mastercode to add features, etc... and compiling and testing it on the container terminal.

In this respect:

• It is not advisable to launch the container more than once. This is, you will have only one "linux terminal".
• To close the container session, just type "exit"
• The changes you may do to directories belonging to the container, will be automatically lost when you exit.

  If you want to save them, you need to save the status of the container (see https://docs.docker.com/engine/reference/commandline/commit for detailed instructions on Docker software)

• Notice that your $HOME directory is simply mounted, and has nothing to do with the container software itself. Of course, all your changes there will remain.