10. Packaging Environment (sandboxing)

This document will describe how to setup and use a clean packaging environment to build packages locally on your installed VectorLinux

10.1. Assumptions

This guide will assume the following.

  • You have VectorLinux 7.1 installed.

    This setup cannot be produced in earlier versions of VectorLinux

  • You have read and followed the docker guide for VectorLinux and

    have successfully launched the docker daemon

  • You are confortable using the command line or a teminal emulator.

  • You are somewhat familiar with how to use git

10.1.1. Task Description

The clean build environment is achieved by setting up docker and importing the correct image into your docker installation.

The default docker behaviour is to discard all of the data after each container has run its life cycle. In order to save the the packages you create inside the clean environment, we will need to setup a directory structure somewhere in your path so the data can be saved there for access after the container is terminated.


If you are on a 32-bit installation of VectorLinux you will only be able to use 32-bit images. Users with a 64-bit installation of VectorLinux will be able to use 32-bit and 64-bit images.

10.2. The quick and easy way (using the vlsandbox utility)

The VectorLinux project has a tool called the vlsandbox which automates a lot of what is described below. This tool can (and should) be used as a non-root user. When launched, it will perform the following tasks.

  • Create a temporary clean build environment usind a docker image
  • Refresh (clone first, if necessary) the official vabs repository and make it available in /home/slackbuilds inside the sandbox environment.
  • Give you (the user) a shell prompt with all the builds tools and source scripts at your disposal to compile what you wish from source code.

10.2.1. Getting the utility

You can get the vsandbox utility in VectorLinux versions 7.1 and newer via the slapt-get and GSlapt package manager.

slapt-get -i vlsandbox or search for vlsandbox in GSlapt from your GUI desktop.

10.2.2. Adjusting The settings

The vlsandbox utility relies on a few settings that can be adjusted by the user to better fit their needs. These settings can be found in:

  • /etc/vlbuildbot/vlsandbox/vlsandbox.conf. This file controls the system-wide default settings. User-spefic settings are based off this template.
  • $HOME/.vlsandbox/vlsandbox.conf. This is a copy of the system-wide settings. Each user can adjust this as needed.

Some key settings can also be overridden at runtime from a terminal to temporarily switch from one base to another. For example, the VLARCH env variable tells vlsandbox which architecture to work with and the VLRELEASE tells it which version of VectorLinux to target.

These can be set in the run command. For example VLARCH=i586 VLRELEASE=veclinux-7.1 vlsandbox will give you a sandbox for VectorLinux version 7.1 in 32-bit mode.

10.2.3. Using the sandbox to contribute to VectorLinux

You can use the vlsandbox utility to test your builds if you wish to contribute to the VectorLinux project. To do so, you will need to fork the vabs tree from bitbucket and issue a Pull Request. See the graphical representation of the workflow available at https://i.imgsafe.org/d5a9dbe.png

10.3. The Technical way (complicated, TL;DR)

This will allow you to enter a sandbox similar to what is documented above in a more detailed way.

10.3.1. Preparations

As pointed out above, you will need to create a directory structure to save your packages once they have been created. For this, we will assume you are in your $HOME directory, but really it does not matter as long as you know where you placed your data. Wherever that may be, create the structure following these commands.


The following snippet assumes you have cloned the image you wish to use. Use docker clone m0elnx/vlocity-7.2-bb:latest to download the base image for 7.2 or use docker search to find a suitable docker image.

mkdir -p sandbox/data-{in,out}
pushd sandbox/data-in
git clone http://bitbucket.org/VLCore/vl71

This has just prepared a directory structure in your host that will be shared with your docker container. The git clone... line would have downloaded the build scripts collection for VectorLinux 7.1 and should be a good starting point for building many applications from source.

10.3.2. Entering the build environment (sandbox)

To enter the build environment, you will need to call the docker run command. To get familiarized with what the arguments used mean, see docker help run.

export DATA_IN=$PWD/sandbox/data-in/vl71
export DATA_OUT=$PWD/sandbox/data-out

docker run --rm -ti -v $DATA_IN:/home/slackbuilds -v $DATA_OUT:/data \
m0elnx/vector-7.1-bb:latest /bin/bash

At this point, you should see some output on the terminal and will notice the prompt has changed. you should be able to cd /data/in/ and see the directory named vl71 that you cloned with git before. You can find a lot of build scripts here ready to be executed.

A graphical representation of the workflow for contributing your changed data back to VectorLinux project is available at https://i.imgsafe.org/d5a9dbe.png

10.3.3. Working inside the build environment

Now that you are inside the clean build environment, you are free to install whatever you need inside. There is a variety of tools available to you inside the sandbox. The following is a short list of some of the ones you will most likely need to use inside the sandbox.

  • slapt-get
  • sbbuilder
  • version control applications (svn, git, etc)
  • text editors (nano, vi, emacs, etc)

If you happen to need anything else, just use slapt-get to install it.