Chapter 1. Organisational Requirements

Table of Contents

Background
Opportunity
Existing Systems
Where Our Solution Fits
Objectives and Success Criteria
User Classes and Needs
Charlie: The Beginner
Dakota: The Enthusiast
Jamie: The Administrator
River: The Developer
Risks

The Adélie Linux distribution has a belief that "all people, in all walks of life, deserve the very best that their technology can offer them".[1] However, at present, the only way to install the Adélie Linux distribution is via a list of complex commands, entered at a shell prompt that may be foreign or confusing to many people.[2] Further, improper execution of these commands can cause permanent loss of data[3], or even hardware damage[4].

To this end, a new system must be created. The system must allow any user to install Adélie Linux onto their computer without needing to use a command line.

Additionally, we have identified a significant user need for installation automation.

Existing installation systems in the current Linux distribution landscape do not adequately meet the needs of different Linux users. In this section, we will review current installation systems and their deficiencies. Current distributions contain one of three major types of installation systems or routines: those that hold the user's hands at the expense of flexibility (or "batteries included"); those that are rigid and don't work well outside of a limited scope (or "fixed"); and those that require the user to do most or all of the installation themselves using a command line (or "manual").

The Ubuntu distribution uses an installer named Ubiquity. Ubiquity is very user friendly; however, it has many shortcomings. It cannot be scripted - administrators must use the Debian installer with a preseed file to script an Ubuntu installation. The partitioning options offered are limited. It uses DBus and udevadm to set up networking, which is not robust. It is focused on x86 and unlikely to work well on other architectures. It uses Debian's partman internally for partitioning, which is a mess on its own.

The Debian distribution uses debian-installer, or d-i. d-i offers a lot of flexibility with plug-in installation steps. However, it suffers from many limitations. The scripting format, termed "preseed", is extremely convoluted and very poorly documented. It has been described as a "nightmare" by its own users[5] and has many warts as a result of being partially written in Perl, and partially in C, which resulted in a maintenance burden. In addition, d-i does not always support common hardware, especially on the remastered CDs required for scripting - notably, SATA drivers are not added which renders it unusable on most modern computers. It suffers from a severe lack of predictable results; given the same input, it does not always produce the same output. This is due to various race conditions throughout the system.

The Red Hat and Fedora family of distributions use the Anaconda installer. This is one of the oldest systems still in use today, and is wholly written in Python. Scripting support via Kickstart files is higher quality than Debian's installer, but still suffers from multiple issues. Repository signatures cannot be added. If packages are required that are not present in the base RHEL/Fedora repo, you must allow unsigned packages to be installed. This can present a security risk in the case of DNS or ARP poisoning. Further, until the most recent version of Fedora, the only supported wireless network authentication mechanism was WEP, which is highly insecure and is typically not available on newer wireless networks. The newest version additionally supports WPA passphrases; it does not support RADIUS authentication or certificate-based authentication.

The Calamares system is considered most likely to be a "batteries included" system, but we did not investigate it thoroughly during our requirements elicitation phase due to time constraints.

The Alpine Linux installer is a set of interactive shell scripts that offers very few options. A user can choose between using a hard disk for the entire system or for data only, leaving the OS and software in a tmpfs. There is no option to use an existing partition map, or to use an alternative to LVM2, nor to use RAID. Because it is a set of interactive shell scripts, it does not make sense to script an installation using it. Typically, users script a manual installation (such as in the "Manual" section below). There is no set of default packages: the installation system only installs a minimal base system. There is no support for using IPv6 in the Alpine Linux installer, and IPv6 support is crucial for usage on the modern Internet.

FreeBSD and OpenBSD also fall in this category. OpenBSD's installation system is very close to Alpine's, while the FreeBSD bsdinstall makes many assumptions about the user's desired choices and offers very little customisation. One area where bsdinstall succeeds is the partition interface. None of the three major BSD derivatives support any form of scripted install.

The Arch Linux distribution provides a list of commands for the user to run, and provides a very minimal environment on their boot CD to allow them to run. You must manually bootstrap the package manager, Pacman, and install the base system packages yourself. This lends itself to scripting, but it again must be entirely manual. The user would be scripting the disk manipulation, clock set up, package database initialisation, and so on, themselves.

The Gentoo distribution provides a "stage" tarball containing a base system. The user must create partitions manually before extracting the stage to the target. The user must also build and install their own kernel and bootloader. As with Arch Linux, any scripting would be automating the low-level commands themselves; there is no provided framework for automating installs in Gentoo.

We feel that there is a significant user need for an installation system that is as easy to use as any of the proprietary operating systems, such as Windows 7 or Mac OS X. Such a system would significantly further our goal of bringing high-quality, Libre software to all people of all walks of life. Additionally, such a system could gain Adélie Linux significant market-share, leading to more sponsors, which will lead to further growth and the ability to take on even more ambitious projects in the future.

We have identified a few key objectives that we feel are important for the system to meet.

For the purpose of these objectives, "completing" the installation routine means that the user has finished answering the system and that the system may now begin performing the installation.

  1. Users who have not used the Adélie Linux system before, but have prior experience with other Linux systems, shall be able to complete the installation routine in less than 15 minutes.

  2. Users who have not used any Linux system before shall be able to complete the installation routine in less than one hour.

  3. Positive reviews of our installation system in at least two publications within six months.

  4. Rated as the best installation system in at least one publication within one year.

We have identified four major classes of users, each with their own unique set of needs and requirements for Project Horizon. For each class of user, we have an example profile of a person who would belong to that class of user. We then list their reasons for using the system and the features that matter most to them.

Charlie is a 45 year old designer who has never used Linux before. Charlie is concerned about privacy using Windows 10, and would like to install an alternative operating system. Charlie does not want to erase the current OS installed on this computer.

  1. Simple user interface with defaults pre-set for most common identifiable use cases.

  2. Ability to install "side-by-side" with another OS with minimal effort, including:

    1. resizing the current OS partition;

    2. creating needed new partition(s) for Adélie Linux; and

    3. ensuring a boot menu is displayed that allows choice on whether to boot Adélie Linux or another OS.

  3. On-screen help for explaining terms used and an explanation for available options.

  4. Disk encryption to keep data private and secure at-rest.

Dakota is a 27 year old Linux enthusiast, who has used Linux for 4 years. Dakota has used other distributions and wants to try Adélie Linux on a laptop, and has opinions on package choices.

  1. Customisation of partitioning, including sizes, file systems, and mountpoints.

  2. Ability to choose init system, shell, and window manager.

Jamie is a 33 year old systems administrator at a business. Jamie wants to deploy Adélie Linux to five Web servers on a business network.

  1. Network-based, scripted installation.

  2. Ability to select additional packages to install.

  3. Assurance of security and integrity, for both the installation script and the packages installed.

River is a 20 year old tinkerer. River has Single Board Computers (SBCs) such as a Pine64 and a Quark. River also has multiple disparate laptops.

  1. Ability to choose init system, shell, and window manager. (See NEE-2.)

  2. Support for esoteric hardware.

  3. Disk encryption to keep data private and secure at-rest. (See NEB-4.)

  4. Customisation of partitioning, including sizes and mountpoints. (See NEE-1.)

  5. Ability to select additional packages to install.

We have identified the following risks associated with developing Project Horizon.

  1. User acceptance may be lower than anticipated, causing less adoption than desired for Adélie Linux.

    Impact: High; Probability: Medium.

    Potential mitigations: Ensure that user acceptance tests are done throughout the development cycle. Involve designers that have experience with user experience best practices. For higher quality feedback, involve parties from different communities whom have different levels of familiarity with Linux.

  2. Scheduling over-runs could cause severe financial losses.

    Impact: High; Probability: High.

    Potential mitigations: Raise funding for Adélie Linux from sources that are not contigent on the release of Project Horizon first. Drop lower priority features that are slated for 1.0 if schedule slippage begins. Seek more contributors.

  3. Packages and security updates could fall behind if the team puts all resources into Project Horizon.

    Impact: Medium; Probability: Low.

    Potential mitigations: Not every Adélie Linux committer will be dedicated to Project Horizon. Some committers can do important security bumps while the others work on Project Horizon.

We have identified the following risks associated with not developing Project Horizon.

  1. Users will not want to install Adélie Linux because manual installation is too complex.

    Impact: High; Probability: High.

  2. Sponsors will not fund Adélie Linux because they cannot easily install it.

    Impact: High; Probability: High.

  3. Users having technical problems due to bad installation procedures would give Adélie a poor reputation and take up a lot of time for community support.

    Impact: High; Probability: High.

  4. Potential developers and maintainers would not contribute to Adélie Linux because the barrier to getting started with Adélie is too high.

    Impact: High; Probability: Medium.



[1] Adélie Linux home page, 2019-08-23, adelielinux.org.

[2] HOWTO: Manually install Adélie Linux, 2019-01-12, Adélie Wiki.

[3] Such as improper use of parted, or even mkfs.

[5] See Eromenko 2012, for example.