Source code1
       Roberto Di Cosmo
        INRIA, Paris, France

       "Programs must be written for people to read, and only incidentally for
       machines to execute." (Harold Abelson)

   In a computer system, there are generally two parts: the hardware, which is the
physical part of the system (processor, memory, disks, screen, network card, sound card,
keyboard, mouse, etc.), and the software, which designates the set of instructions that
can be stored and executed by the machine. It is the software that gives life to a
computer system.

    Instructions that can be directly executed by a machine (called "machine language")
are often very low-level, represented by simple sequences of bits and difficult to
understand by a human being. This is why software is almost never produced directly
in machine language, but "written" by developers using a "programming language",
which can then be automatically translated into machine language.

   As an example, here is an excerpt from the executable program that prints a simple
"Hello world" message:
       4004e6:   55
       4004e7:   48   89   e5
       4004ea:   bf   84   05 40 00
       4004ef:   b8   00   00 00
       4004f4:   e8   c7   fe ff ff
       4004f9:   90
       4004fa:   5d
       4004fb:   c3


   Here is the program written in the C programming language, from which the
executable program from which we extracted the fragment presented above was
produced:
       /* Hello World program */#include<stdio
       .h>
       void main()
       {
       printf("Hello World");
       }

1. This text is licensed under the Creative Commons CC-BY 4.0 license. It has been
contributed to the collaborative work “Dictionnaire du Numérique” published in 2022.
    The "source code" of software is generally understood to be the program that was
written by the developer and from which the executable code is obtained on a
machine. Also, one might be tempted to consider as "source code" any program written
in a programming language. As technology has evolved, the situation has actually
become more complex: developers have sophisticated tools that can "produce"
programs in one programming language (such as C) from programs written in higher-
level programming languages, to the point that it is not enough to look at the language
in which a program is written to know whether that program was written by a developer
or generated automatically from a higher-level program.

    That is why the definition of "source code" for software, found in the GPL, is "the
preferred form for a developer to make a modification to a program".

     Source code is a special form of knowledge: it is made to be understood by a human
being, the developer, and can be mechanically translated into a form to be executed
directly on a machine. The very terminology used by the computer community is
telling: "programming languages" are used to "write" software. As Donald Knuth, one of
the founders of computer science, wrote, "programming is the art of explaining to
another human being what you want a computer to do".

    Software source code is therefore a human creation, just like other written
documents, and that is why it falls within the scope of copyright law. Software
developers thus deserve the same respect as other creators, and it is essential to ensure
that any changes to copyright law take into account the potential impact on software
development. This was not the case in the drafting of EU Directive 2019/790, the first
draft of which seriously endangered the collaborative development of open source
software, and which required a significant effort to make the necessary corrections.

   As software source code becomes more and more complex, it is regularly modified
by groups of developers who collaborate to make it evolve: to understand it, it has
become essential to have access to its development history.

    The software source code is thus incorporating an important part of our scientific,
technical and industrial heritage, and thus constitutes a valuable heritage, as already
argued by Len Shustek in an excellent article in 2006.

   This is one of the missions of Software Heritage, an initiative launched in 2015
with the support of INRIA, in partnership with UNESCO, to collect, organize, preserve
and make easily accessible all the source code publicly available on the planet,
regardless of where and how it was developed or distributed. The goal is to build a
common infrastructure that will allow a multiplicity of applications: of course, to
preserve the source code in the long term against the risks of destruction, but also to
enable large-scale studies on the code and the current development processes, in order
to improve them and thus prepare a better future.

    At a time when it is clear that software has become an essential component of all
human activity, unrestricted access to publicly available software source codes, as well
as qualified information on their evolution, is becoming an issue of digital sovereignty
for all nations. The unique infrastructure that Software Heritage is building, as well as its
universal approach, is an essential element to meet this challenge of digital sovereignty,
while preserving the commons dimension of the archive.

Bibliography
Abelson, H. (1984). Structure and Interpretation of Computer Programs. The MIT
   Press, Cambridge.
Di Cosmo, R. (2019). Saving software development from the European copyright
   reform. Dicosmo.org [Online]. Available at: https://www.dicosmo.org/
   MyOpinions/index.php?post/2019/04/17/Saving-software-development-from-the-
   European-copyright-reform.
Di Cosmo, R., Nora, D. (1998). Le Hold-up planétaire : la face cachée de Microsoft.
   Calman-Lévy, Paris.
Shustek, L.J. (2006). What Should We Collect to Preserve the History of Software?.
   IEEE Annals of the History of Computing, 28(4), 110–112 [Online]. Available at:
   https://doi.org/10.1109/MAHC.2006.78.
Software Heritage.       Site    [Online].    Available    at:   https://www.softwareheri
   tage.org.