Chapter 3. General ideas

Table of Contents

3.1. Looking beyond
3.2. Motivation
3.2.1. Profile of target users
3.2.2. Profile of target administrators
3.3. Status of specialised Free Software
3.4. General problem
3.5. Debian Pure Blends from philosophical point of view

3.1. Looking beyond

Commercial Linux distributors sell certain products that try to address special user needs.

Enterprise solutions
  • Advanced Server - RedHat

  • Enterprise Server - SuSE

Small Office and Home Office (SOHO)

There are a couple of workstation or home editions, as well as office desktops built by several GNU/Linux distributors.

Special task products
Mail server

SuSE Linux Openexchange Server

Firewall

SuSE Firewall on CD, ...

Content Management System

RedHat

Portal Server

RedHat

This is only a small set of examples of commercial GNU/Linux distributors addressing specific user interests with certain products.

Debian solves this problem with Debian Pure Blends.

3.2. Motivation

3.2.1. Profile of target users

The target user of a Blend may be a specialist of a certain profession, (e.g. a doctor or lawyer,) a person who has not (yet) gathered a certain amount of computer knowledge, (e.g. a child,) or a person with disabilities (e.g. a visually or hearing impaired person.) Moreover, the customisation might deal with peculiarities of certain regions where users have needs that differ from Debian as a whole.

It is not unusual for these target users to be less technically competent than the stereotypical Linux user. These people are often not interested in the computer for its own sake, but just want it to work for them. Imagine the frustration of a doctor who has to move the focus of interest from the patient to his stupid computer that does not work as expected.

Because of limited knowledge or time, the target user is usually unable to install upstream programs. This means that in the first place, they must find out which software packages in their distribution might serve for a certain problem. The next step would be to download and install the packages they choose, perhaps requiring a certain amount of configuration effort. This problem is nearly impossible for a user with limited technical competence and perhaps poor English language comprehension, which prevents the user from understanding the installation manual.

The language barrier in this field is an important issue, because we are targeting everyday users who are not compelled to learn English, like Free Software developers are, for everyday communication. So the installation process has to involve the least possible user interaction, and any such interaction has to be internationalised.

Furthermore, most target users have no or little interest in administration of their computer. In short, the optimal situation would be that he would not even notice the existence of the computer, but just focus on using the application to accomplish the task at hand.

Common to all groups of target users is their interest in a defined subset of available Free Software. None of them would like to spend much time searching for the package that fits his interest. Instead, the target user would prefer to immediately and effortlessly locate and access all material relevant to solving his own problems.

There is an absolute need for easy usage of the programs. This is not to say users expect to not have to learn to use the software. Adults generally accept that they must spend a reasonable amount of time in learning how to use a piece of software before they can do something useful and productive with it. But a simple-to-learn environment greatly enhances the value of the software, and if you consider children as target users, they just want to start using it right away without reading any documentation.

The more important part of the request for easy usage is a professional design that is functional and effective. To accomplish this, the programmers need expert knowledge, or at least a quick communication channel to experts to learn more about their requirements. One task for Debian Pure Blends is to bring programmers and experts who will use those special programs together.

Last, but not least, we find certain requirements beyond just which packages are provided in each target user group. These may differ between different Blends. For instance, while a doctor has to protect his database against snooping by outside attackers, the privacy risk for a child's system are of lesser importance. Thus, the Debian Junior project cares more for ensuring that the user himself does not damage the desktop environment while playing around with it than about remote attacks. So we find a "defined security profile" for each single Blend.

3.2.2. Profile of target administrators

In the field that should be covered by Debian Pure Blends, we have to face also some common problems for system administrators. Often they have limited time in which they must serve quite a number of computers, and thus they are happy about each simplification of the administration process. The time required to make special adaptations for the intended purpose has to be reduced to a minimum.

So, administrators are looking for timesaving in repetitive tasks. While this is a common issue for each general GNU/Linux distribution, this could have certain consequences in the special fields Debian Pure Blends want to address.

Another problem administrators face is that they are often not experts in their clients' special field of work. Thus, they may need some specialist knowledge to explain the use of special programs to their users, or at least need to be able to communicate well with the experts about their special needs, and how the software can be used to address them.

3.3. Status of specialised Free Software

Programs like a web server, or a mail user agent are used by many different users. That is why many gifted programmers feel obliged for this kind of Free Software - they just need it for their own. So you normally find a fast, growing community around Free Software packages that have a wide use. This is different for specialised software.

In this context, the term "specialised software" refers to the kind of software that is needed by some experts for their job. This might be a practice management system that is used by doctors, a graphical information system (GIS) that is used by geographers, a screen reader that helps blind people to work with the computer, etc. The difference between such software and widely used software like office suites is that the user base is relatively small. This is also true for certain software that supports special localisation issues.

  • Specialist software is used only by a limited set of users (i.e. the specialists). There exists a set of software tools that work perfectly in the environment where they were developed. If the developers catch the idea of Free Software, and just release this software as-is, people in the new, broader user community often run into trouble getting it to work in their environment. This happens because the developers did not really care about a robust installation process that works outside their special environment. As well, installation instructions are often badly written, if they exist at all. But these problem can be easily solved by shipping the software as policy-compliant binary packages, which not only ease installation, but also require documentation to be included. Thus, mere inclusion in Debian benefits the whole user base of any specialised software.

  • The trouble often continues in the maintenance of the installed software.

  • When it comes to the usage of the specialist software, it often happens that it perfectly fits the needs of the developer who wrote it for his own purposes, and who is familiar with its quirks, but in many cases such software does not comply with ergonomic standards of user interfaces.

  • Several existing programs that might be useful for specialists are not really free in the sense of the Debian Free Software Guidelines (DFSG). Programs that are incompatible with the DFSG cannot be included in Debian. This is possibly a drawback for those programs, because they could profit by spreading widely on the back of Debian over the whole world.

  • A certain number of programs are developed at universities by students or graduates. Once these people leave the university, the programs they developed might be orphaned; i.e., not actively maintained anymore. If their licenses are too restrictive, it may be impossible for anyone else to take over; sticking to AT: We should find a way to avoid printing the URL in PDF output. DFSG-free licenses avoid that problem.

  • In special fields, often "typical" (not necessarily Intel-based) hardware architectures are used. Debian currently runs on 11 different architectures, and automatic build servers normally compile software packages as necessary. If auto-builders for other architectures show problems, Debian maintainers will normally fix them, and send the original authors a patch. Moreover, users can report run-time problems via the Debian Bug Tracking System.

  • Many programs that are written from scratch use their own non-standard file formats. However, it is often important for programs to be able to share data with each other.

  • Often there are several programs that try to solve identical or similar problems. For instance the Debian Med team faces this in the case of programms claiming to serve as a medical practice management solution. Normally, all these programs take very interesting approaches but all of them have certain drawbacks. So, joining programmers' forces might make sense here.

  • Sometimes the tools or back-ends used in Free Software are not appropriate for such applications. For instance, sometimes database servers that do not use transactions are used to store medical records, which is completely unacceptable. Other programs use web clients as their front-end, which is not really good for quick (mouse-less) usage, a great shortcoming for repetitive tasks.

3.4. General problem

Free Software development is a kind of evolutionary process. It needs a critical mass of supporters, who are:

  • programmers and

  • users

Because specialised software has a limited set of users, (specialists,) this results in a limited set of programmers.

Debian wants to attract both groups to get it working.

Debian is the missing link between upstream developers and users.

3.5. Debian Pure Blends from philosophical point of view

Debian currently grows in several directions:

  • Number of involved people

  • Number of packages

  • Number of architectures

  • Number of bugs

  • Number of users

  • Number of derivatives

  • Time span between releases

So several features are changing at different rates their quantity. According to Hegel a change of quantity leads into a change in quality. That means that Debian will change at a certain point in time (or over a certain time span) its quality.

"To determine at the right moment the critical point where quantity changes into quality is one of the most important and difficult tasks in all the spheres of knowledge." (Trotzki) This might mean that we just passed the point in time when Debian changed its quality. At one point we even observed a change once the package pool system was implemented to cope with the increased number of packages while trying to reduce the time span between releases. Even if the plan to increase the frequencies of releases failed Debian became a new quality. People started using the testing distribution even in production which was not really intended and in a consequence even security in testing was implemented for Sarge.

According to Darwin evolution happens through quantitative transformations passing into qualitative. So Debian has to evolve and to cope with the inner changes and outer requirements to survive in the Linux distribution environment.