5. How can Open Science and IP better interact?

In this Section we have included two Subsections which correspond to different typology of findings after the literature review. Subsection 5.1 will analyse the interaction between IP and three components of OS, namely Open Access to publications, data management and Free Software. Subsection 5.2 will bring four specific IP issues that scientists may face when performing data science in the ICT context that unbalance the relation between OS and IP, producing unaware IP infringement risks.

5.1. OS components and IPR

In the previous sections, the components of Open Science and IP that have to play together have been identified. On one side, OS is supported by the fundamental 'right to science' and, on the other side, IPR protects the rights of the creators. Nevertheless, when analysing the balance between both OS and IPR, the default legal requirement for the transmission of a work is either an explicit consent, or the usage of an exception, or the existence of public domain. In case of conflict, it is the user of the work who has the burden of proof that one or more of these requirements exist. Therefore, using works for a scientific activity may produce risks that should be avoided by a correct use of the permissions designed in IP regulations.

As it has been mentioned in previous Section 1.2., in the literature review there are very few analyses focused on how OS and IPR can coexist. EARTO has produced some abstract and general recommendations, encouraging the translation of IPR 'into the commercial world, especially through patent filing' (EARTO, 2020, p. 9), but as said, the literature is nearly non-existent.

The European project AGINFRA+ proposed five domains where transformations should be made to evolve from the current paradigm of research to OS. These were:

• From open access to open science.

• From human-readable to machine-readable content.

• From open data to FAIR data, data sharing and data re-use.

• From traditional publishing to technology-driven service.

• From semantic enrichment of content to semantic publishing.

Decomposing AGINFRA+ analysis into scientists' practices, we find that all the activities mentioned in the transformations are subject to intellectual property regulations. As AGINFRA+ states, these activities, which are necessary to shift into OS, affect the whole ecosystem of producing, communicating, linking, semantic tagging and reusing research results, within data, information and knowledge levels, by using machine readability techniques, automated harvesting and text mining. A comparison between scientific and intellectual property activities can be found in Table 7.

Scientific activity	Intellectual property activity
Collecting data	Creating a database/Create a work
Communicating	Public communication
Linking	Public communication
Semantic tagging	Public communication (a tag may include an hyper reference to the tagged object, thus, linking)
Reusing research results	Reproduction
Use machine readability	Reproduction
Use of automated harvesting	Reproduction
Text mining	Reproduction1

Table 7. Comparison between scientific activities and IP corresponding issues.

In addition to AGINFRA+ identified challenges, UNESCO's comprehensive enumeration of OS components should be considered, which include the following items: Open Access, Open Data, Open Source software/Open hardware, Open Evaluation, Open Educational Resources, Open Engagement of Societal Actions and Openness to diversity of knowledge (See Table 2.1). From this aggregated base we may begin to analyse the challenges OS and IPR have to face.

5.1.1. Open Access to scientific publications

Open Access is one of the best studied and most necessary transformations for OS. In the 20 years since the Budapest Declaration (BOAI), an overwhelming amount of publications, specific and thorough literature reviews, etc. have been published, describing and typifying Open Access from different perspectives frequently challenging IP. Some of the most cited and generic approaches to Open Access are Peter Suber's contributions (Suber, 2012), but there are also important literature reviews from CREATe project (Frosio, 2014) or the systemic and comprehensive analysis of legal action in light of international IP that (Scheufen, 2015) summarizes in his book, covering the debate over copyright law and its impact on the distribution of scientific knowledge from an economic perspective. The EC Commission Recommendation on Open Access and Preservation of scientific information was first published in 2012 and updated in 2018², and since then it is clear that all MS 'should ensure, in compliance with the EU acquis on copyright and related rights, that as a result of these policies or action plans: all scientific publications resulting from publicly-funded research are made available in open access as from 2020 at the latest'. 2020 is over, but the objective remains applicable for immediate Open Access to research publication. Plan S, and other newer strategies and approaches to OA, try to obtain that objective by retaining the necessary intellectual property rights by the researchers, allowing the possibility of licensing their works with open licenses.

The activities regulated by copyright (to copy, to distribute, to communicate to the public and the creation of derivative works) are forbidden by default. The impact of these activities on knowledge transmission is crucial, therefore their interdiction by default creates a hostile environment for science as it imposes access barriers, as it was underlined in the Budapest Declaration in 2002 (BOAI).

It is instead a question about copyright that places OA to publications at the heart of the revolution, since publications are still at the heart of research evaluation. OA questions the appropriateness of, the need for, and ultimately the success of, copyright itself, in the future of scholarly publishing (Harper, 2009). Nevertheless, several political, rather than technical, decisions are pending, mainly on how to avoid the current oligopoly on journal publishers. The regulatory approach should take into account competence and the necessity of EU's digital sovereignty (Burgelman, 2021). This regulatory approach, which seems to be followed by the proposals for the Digital Single Markets Act³ and the Digital Services Act⁴ and the recent Horizon Europe Regulation (EU) 2021/695 of the European Parliament and of the Council of 28 April 2021⁵ should be the way to dismantle the business model over science developed by big publishers, as Buranyi details in his article published in The Guardian, Is the staggeringly profitable business of scientific publishing bad for science? (2017).

A recommendation will not be made in this subsection, which will instead follow the different initiatives made by Universities and researchers around the globe to liberate academic literature production.

5.1.2. Data. The data cycle. FAIR (Open) data

When it comes to data and IPR, it could be summarised that data and facts do not have protection under copyright, but as it will be described below, databases do. The EU has a specific regulation contained in the Directive 96/9/EC of the European Parliament and of the Council of 11 March 1996 on the legal protection of databases⁶. The legal regulation defines 'database' in its article 1.2 as 'a collection of independent works, data or other materials arranged in a systematic or methodical way and individually accessible by electronic or other means' and determines that 'databases which, by reason of the selection or arrangement of their contents, constitute the author's own intellectual creation shall be protected as such by copyright. No other criteria shall be applied to determine their eligibility for that protection.' This definition could be reasonable in a world where the digital was not ubiquitous, the use of laptops was not universal and fax machines were the most advanced method of transmitting documents, but nowadays it is no longer acceptable. Under the Directive, legal concept of a database falls under the directory and file structure of a computer. The result of any classification in a computer is therefore copyrighted. As will be described in subsection 5.5. APIs, taxonomies and ontologies, ICT driven OS basic activity is to create taxonomies and ontologies through harmonisation of different concepts used in science, defining a common realm. If Linneo would have created his taxonomy of living animals today, he would be subject to these regulations and by default his classification would have been unable to be used.

Additionally, the Directive establishes for the databases an extra layer of rights, known as sui generis rights, in its article 7.1, 'a right for the maker of a database which shows that there has been qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents to prevent extraction and/or re-utilisation of the whole or of a substantial part, evaluated qualitatively and/or quantitatively, of the contents of that database.'

This regulation, as Ohad Graber-Soudry et al have insightfully studied at both the theoretical and practical levels, 'From an EOSC perspective, the protectionist approach of the Database Directive is problematic because it automatically frames access to data as a threat. This is counter to legal interoperability and in general to the Open Science idea. There should be more balance between the protection of databases on the one hand and access and re-use of data on the other' (Graber-Soudry et al., 2021, p. 29). These authors make a complete analysis of the relations between the two domains, identify properly permissive IPR and conclude with 31 very specific recommendations. Under the same opinion of the relation between databases and OS, Vera Lipton states that

'The examination found that copyright law poses serious challenges to data release and reuse in all three jurisdictions under examination---the United States, Australia and the European Union. The problems arise due to uncertainty surrounding the scope of copyright protection as it applies to the various forms of data, especially databases. The situation is even more complicated in the European Union which provides a double layer of sui generis and copyright protection. Therefore, using the data created by European research organisations carries an inherent risk of IP infringement. Another source of legal uncertainty is the ownership of data and the inability of users to identify data owners, which poses challenges to data licencing and subsequent reuse due to lack of clarity around the conditions governing data reuse.' (Lipton, 2020, pp. 28-29).

Under this legislation lives FAIR data. As asserted by (EOSC Executive Board, 2021, p. 70): 'The FAIR principles were born with research data. Today, applying FAIR principles has to be extended to the whole research lifecycle, to ensure transparency, assessment, attribution and reproducibility. For this to happen, all outcomes of science, such as data, software, other products and services, have to be FAIR.' Hence, FAIR guiding principles are not some kind of theoretical static photograph that can be taken of a project in a certain moment of its life, but rather they constitute a lighthouse that guides permanently the activities that compose the whole research lifecycle. When studying the data cycle of research projects the following phases are applicable:

occurrence (discovering, designing, authoring, etc.), transmission (networking, distributing, accessing, retrieving, transmitting, etc.), processing and management (collecting, validating, modifying, organizing, indexing, classifying, filtering, updating, sorting, storing, [normalizing] etc.), and usage (monitoring, modelling, analysing, explaining, planning, forecasting, decision-making, instructing, educating, learning, etc.) (Floridi, 2010, p. 8).

The four phases of occurrence, transmission, processing--management and usage exist in all the activities where data, information or knowledge constitute the building blocks. Therefore, they are applicable to the daily work of all researchers 'whose activities are symbolic--based [and] are continuously creating, transmitting, managing and using data in a net woven and twisted with other participants' (de la Cueva, 2018, p. 10). This scheme of information management is valid for an analytical approach to the study of the data activities exercised by any type of organisation, where the relationship between agents is multiple, where it is useful to understand the data cycle as a threefold entity composed of three scenarios where data related activities are held: the first scenario being data-in, the second data-inside and the third, data-out. The scenario data-in is where an organisation collects data. This data may come either from a third party or it may be built by activities held by the organisation through surveys, sensors, personal interviews, observations etc. In the second scenario, data-inside, is where an organisation transforms, orders, catalogues, analyses and deletes the data while the third scenario corresponds to private delivery or public dissemination of the data. The three scenario approach may help to understand where the different activities regulated by IPR⁷ are being held and thus where to take care of possible conflicts.

Fig. 3. Reuse of Research Data. Source DMLawTool (https://dmlawtool.ccdigitallaw.ch) Graphic design: Mattia Pera. Under CC-By-NC-ND.

The flow of information in the three scenarios has limits that depend on factors external to the involved agents; limits that enable, impede and/or hamper the information cycle. These limits are legal, economic and technical. In table 8 below we propose a data management scenario applicable to all organisations that manage digital data in their activities. The table uses the three scenario approach and includes the activities normally exercised in every scenario and the legal, economic and technical aspects that may appear in each of the different activities.

Fig. 4. Activity diagram of data in, data inside and data out.

• In a data-in scenario, the only activity is to collect data, which may be from new creation (survey, direct observation, sensors) or come from third party materials (data either structured or unstructured, personal interviews, analysis of images, texts, audio, code).

• In a data-inside scenario, all the activities are produced inside an organisation. The typical activities consist in transformation of the data such as process, curate, keep, analyse, normalise or delete and to link it. Special care must be made to document the provenance of the data.

• In a data-out scenario, the main activity is the dissemination of the data, which may be done either privately to a known number of persons or to the public. According to IPR European laws, public communication exists since the data may be accessed publicly (through a link to a web site, P2P, API, FTP, etc.) even though it is not accessed by anybody.

Activities	Legal issues	Legal checkings	Economic barriers	Technical barriers
Collect (Scenario Data-in)	Privacy (GDPR) Intellectual Property Rights Patents Company secrets Trademarks Personal image Respect to religion M&M clauses1 National security Unknown origin (orphan works)	Apart from the ownership of the data, list the different rights and obligations that receiving the data may have. These limitations may not come only from Intellectual Property Rights but also from a varied set of parameters (patents, religion, national security limits, hate speech, privacy, commercial secrets, etc.)	Financial burdens in favour of a third party: Price, levies, fees, etc.	Hardware and software barriers. Sensors. Proprietary formats. Superseded formats.
Process, curate, custody, analyse, link, normalize, delete (Scenario Data-inside)	International transfers between bodies of the same organisation. Consent extra limitations. Anonymisation. Pseudo-anonymisation.	Analyse if the activities done inside the organisation are allowed by the legal conditions imposed by the data licence or transfer agreement	Payments between parent organisations and subsidiaries. Costs in personnel/hour to clean, massage, order, structure and organise the data for its internal use or external dissemination. Hardware and equipment costs necessary to handle the data.	Inexistent protocols. Frankenstein architectures. No standardisation inside the organisation or between parent bodies and subsidiaries.
Dissemination, distribution (Scenario Data-out)	Privacy (GDPR). Licence. Terms and conditions.	Analyse if the dissemination is allowed by the legal conditions applicable to the data.	Financial conditions in favour of the organisation not acceptable/affordable for third parties.	APIs documentation. Obsolete/rotten websites. Technologies soon to be superseded or changed (Apple USB stick as example).

Table 8. Data activities and their barriers for interoperability (Source: FNS-Cloud project).

In order to be able to produce a more granulated study of IPR conditions of a project, the analysis of the data cycle may be done taking into account the FAIR Guiding Principles (Table 9) and their exercise in the three scenarios.

• Data-in: Analyse the data to be used under the recommendations set forth by the EOSC FAIR Working Group. Grade the FAIRness of the data.

• Data-inside: Analyse if the activities done by the organisation will affect the FAIRness of the data.

• Data-out: Analyse if the organisation has contributed to comply with more FAIR principles than the ones accomplished when the data came in. Grade the FAIRness of the data so as to build an indicator of the contribution of the organisation.

FAIR guiding principles impose certain conditions that could oblige the execution of certain activities to be carried out by the data manager, as studied specifically for reusability by Ignasi Labastida and Thomas Margoni (Labastida & Margoni, 2020), in the general and deep analysis of all FAIR data principles provided by Ohad Graber-Soudrey, Timo Minssen, Daniel Nilsoon, Marcelo Corrales, Jakob Wested and Bénédict Illien (Graber-Soudry et al., 2021), in the also complete Legal Interoperability Of Research Data: Principles And Implementation Guidelines by the RDA-CODATA Legal Interoperability Interest Group (RDA-CODATA Legal Interoperability Interest Group, 2016) and in the section 2.4. of the recent EOSC Interoperability Framework (Corcho et al., 2021).

Findable: The data and metadata can be found by the community after its publication, using search tools.	Accessible: (Meta)data are accessible and can therefore be downloaded by other researchers using their identifiers.
F1. Assign the (meta)data a globally unique and persistent identifier.	A1 (Meta)data are retrievable by their identifiers using a standardized communications protocol.
F2. Describe the data with rich metadata.	A1.1 The protocols have to be open, free and universally implementable.
F3. Register/index the (meta)data in a searchable resource.	A1.2 The protocol must allow for an authentication and authorization procedure (where necessary).
F4. The metadata should clearly and explicitly include the identifier of the data described.	A2 The metadata must be accessible, even when the data are no longer available.
Interoperable: Both the data and the metadata should be described following the rules of the community, using open standards, in order to allow for their exchange and reuse	Reusable: (Meta)data can be reused by other researchers, since their origin and conditions of reuse are clear.
I1. (Meta)data must use a formal, accessible, shared and broadly applicable language for knowledge representation.	R1. (Meta)data have a plurality of accurate and relevant attributes.
I2. (Meta)data use vocabularies that follow FAIR principles.	R1.1. (Meta)data are released with a clear and accessible data usage license.
I3. (Meta)data include qualified references to other (meta)data.	R1.2. (Meta)data are associated with information on their provenance.
	R1.3. (Meta)data meet domain-relevant community standards.

Table 9. FAIR data principles underlying the role of licensing in Reusability.

To conclude this section, it is clear that when operating with data, there is an inherent risk related to IPR infringement. Hence, in order to guarantee interoperability, it is necessary to check the validity of the consent of the rightholder or whether an exception/limitation applies.

Recommendation for practitioners

It would be recommendable that all organisations, when using data, analyse the terms and conditions of each dataset. If these are not clear or no consent has been given, then it should be treated as an 'all rights reserved' piece of information.

5.1.3. Free software

Apart from the already mentioned studies made by Professor Yochai Benkler in his book The Wealth of Networks (Benkler, 2006), Professor Christopher M. Kelty has also made an important contribution to the study of Free Software in his work Two Bits: The Cultural Significance of Free Software (Kelty, 2008) where the object of his research has been the massive production of software under free licences. He also studied the scientific newsletters and their importance for Open Science (Kelty, 2012). The relevance of Kelty's book is explained in its introduction:

'Two Bits explains how Free Software works and how it emerged in tandem with the Internet as both a technical and a social form. Understanding Free Software in detail is the best way to understand many contentious and confusing changes related to the Internet, to "commons," to software, and to networks. Whether you think first of e-mail, Napster, Wikipedia, MySpace, or Flickr; whether you think of the proliferation of databases, identity thieves, and privacy concerns; whether you think of traditional knowledge, patents on genes, the death of scholarly publishing, or compulsory licensing of AIDS medicine; whether you think of MoveOn.org or net neutrality or YouTube---the issues raised by these phenomena can be better understood by looking carefully at the emergence of Free Software.' (Kelty, 2008).

It is important to deny some extended theories that assert that Free Software creates 'largely outside IP' as some authors defend (Jones, 2014, pp. 28-29). As per this author, 'The fundamental principle of the open source approach is that either no one owns the IP, or that if they do, they grant everyone else royalty free licences for its use, development and wider application'. This is a common error because every free work is really owned by its author, in a conscious way. Authors exercise their rights in a gradient of possibilities that escape the traditional 'all rights reserved'. The question here is not if there is no IPR, because it is evident that it exists and the authors are aware of their rights (why then should they licence their works?) but how the rights are exercised. Jones continues: 'As Wikipedia itself puts it: "open source refers to a program in which the source code is available to the general public for use and/or modification from its original design free of charge"'. But this really does not mean very much. A book has its source code available (letters, words, paragraphs...), it is the paradigmatic case of a work where the user can access what the author desires to transmit, but its 'openness' to be accessed does not imply the possibilities to exercise the legally interdicted activities of copying, altering, public communication or distribution. Even though the source code may be available, if it does not have a licence then it will be 'all rights reserved'. Due to this, for many years there has been a rich flourishing of different open source licences,⁹ which have adapted dynamically to new technological situations, as for example the GPL did when it upgraded from version 2 to 3 due to the appearance of Digital Rights Management (DRM) and patent deals¹⁰.

This generalization, that Free Software is against copyright, is a common misunderstanding. Free Software is another way authors have to exercise their copyright and then manage their intellectual assets. It does not mean that a work is free of copyright; it expresses that those rights are exercised in a different way than in the traditional 'all rights reserved' paradigm. What these theories miss is that Free Software advocates use copyright to force the openness of their creations, which is a legitimate way to manage intellectual property, often based on ethical considerations (Carver, 2005, p. 445; de la Cueva, 2013). The most important point that should be raised here is that the traditional IP concept of a 'work' was a fixed creation; that is to say, after 'the sweat of the brow' a static final result was obtained (a book, a film, a song, a photograph, an invention, a logo, etc.). This final result was a static product where all the creative actions were solidified. But the Free Software IP concept protects not a work per se, but a process, guaranteeing that this process could not be stopped. Therefore, the traditional copyright is underpinned by a view of protecting static products, whereas Free Software creation is based on protecting the dynamic process.

That Free Software creates 'largely inside IP' is also evidenced by the vibrant discussions held inside developers' communities, where two opposite poles emerged. On one hand, the users of the GPL license, drafted in 1989 by Richard Stallman, which includes what has been called a 'copyleft' clause¹¹, a provision that permits reproduction and distribution of the source code of a programme, 'but does not allow anyone to place further restrictions on them' (Heffan, 1997, p. 1491), requiring at the same time that derivative works from the work under GPL license must be as well licensed under GPL. For Stallman, who channeled his efforts through the Free Software Foundation (FSF) the main focus of importance was the developer's freedom. On the other hand, a group concerned about Stallman's anti-business message, decided in the spring of 1997 to use the term 'Open Source' instead of 'Free Software'. The licences adopted by this group allowed 'the use and redistribution of open-source software without compensation or even credit' (DiBona et al., 1999, p. 3), which included the possibility of appropriation of prior code inside a derivative work, therefore, the possibility to close the initial work. Their focus was the expansion of open-source works. Works made by both groups are now known jointly under the terms Free and Open Source Software (FOSS), whose principles have been widely studied by many scholars (Benkler, 2006; Carver, 2005; DiBona et al., 1999; Heffan, 1997; Kelty, 2008; Lessig, 1999; Olwan, 2013; Raymond, 1999; Stallman, 2002; Torvalds & Diamond, 2002). Olwan analyses this new paradigm under the following approach:

'The Free and Open Source Software (FOSS) methodology has challenged traditional approaches to copyright management by using copyright to build "openness" and reuse. In doing so, it not only has shown new ways for managing copyright but also has underpinned new methods for constructing and disseminating knowledge. To some extent, it has been seen to put the power to develop and share knowledge and technology in the hands of users rather than of an elite or cloistered group of experts. To this end, it has great significance for developing countries.' (Olwan, 2013, p. 266).

'Many copyright systems, particularly those in common law jurisdictions, are based on the "stimulus" or "reward" or "utilitarian" theories. These legal systems justify giving limited property rights (copyright) to the author for the creative work that has been presented to the public. Without adequate compensation and protection of the work, the author would not invest time and effort in producing creative work and society would suffer as a result. FOSS challenges this long-established justification theory of copyright by proving that this is certainly not true for certain collaboration projects where the author produces software without being stimulated by the "reward theory", as known in copyright. Furthermore, FOSS challenges the "personal theory", which justifies giving protection to authors in continental European civil law system jurisdictions by not considering their work to be a stamp of their singular personality but accepting that it might be altered by others for the purpose of contributing to a larger discourse and common goal. The FOSS movement challenges the traditional notions of copyright by describing alternative approaches to software copyright protection and licensing as 'copyleft'. This logic, which challenges the "reward" and "personal" theories of copyright, and the philosophy of encouraging authors to be more tolerant toward others in terms of sharing and building on their work, is an interesting subject of research in FOSS and IT.' (Olwan, 2013, pp. 267-268)

The reference made by Olwan to 'certain collaboration projects' where the author produces software without being stimulated by the 'reward theory' is not only applicable to software. As Peter Suber states, 'scholarly journals generally don't pay authors for their research articles, which frees this special tribe of authors to consent to OA without losing revenue. This fact distinguishes scholars decisively from musicians and moviemakers, and even from most other kinds of authors' (Suber, 2012, p. 9). What is relevant is that IP is a tool that may serve both to protect the fixation of a work using the all rights reserved system or, as the case may be, to protect the openness of a stream where the work is inserted as one of its constitutive elements.

There is an additional approach not found in the reviewed literature but that provides elements for ICT analysis. As stated before, data and information in the digital world is just a list of ones and zeros. Some lists are executable, they are called software, and some lists do not have that function, they are just information which is then managed by software, although Grimmelmann's opinion is that 'a Word document is also a program, one that tells Word itself what characters to display and how to format them' (2016, p. 409). In any case, software is written in formal languages and, according to Charles W. Morris,

'A language, then, as a system of interconnected signs, has a syntactical structure of such a sort that among its permissible sign combinations some can function as statements, and sign vehicles of such a sort that they can be common to a number of Interpreters. [...] so a language is completely characterized by giving what will later be called the syntactical, semantical, and pragmatical rules governing the sign vehicles' (Morris, 1971, p. 25).

Morris is considering that a language has three layers: The Signs (or the syntactical), the Semantics, that 'deals with the relation of signs to their designata and so to the objects which they may or do denote' (1971, p. 35) and the Pragmatics, which refers to the usage of a linguistic sign in combination with other signs by the members of a community. 'In short, to understand a language or to use it correctly is to follow the rules of usage (syntactical, semantical, and pragmatical) current in the given social Community' (1971, p. 48). When dealing with knowledge, it can be said that the scientist work may be done in any of the three layers: either operating in the syntactics (or the source code of software), either in the meanings of the information, either in the social relationships which is, as John Searle puts it, building the institutions (Searle, 2010, pp. 90-122).

An example for jurists could be as follows. The source code of law is the word. Law is built with words that make sentences (syntactics) and a long tradition gives these words the meanings (semantics). Through these laws, institutions are built. Fortunately for jurists, they may use the words and sentences they consider appropriate when regulating a domain. They do not have to include in the legislative proposal whole immutable blocks of texts that come from a company placed in Silicon Valley which contain content they will never be able to know. The more granulated the jurists may operate, the better they can tailor a norm to the needed policy. This immutable block system, which seems irrational when creating laws, is simply what is happening within the digital environment.

When it comes to science, access to source code is not only a matter of IPR but also the necessary requirement to operate at all knowledge levels: syntactical, semantical and pragmatical. Therefore, access and operability on source code (copy, modification, dissemination) is one of the core aspects of Open Science not because of IP, but rather due to epistemology and technological sovereignty (Lafuente et al., 2009). The effects do not stop here: the access to source code is necessary for cybersecurity¹², to create societal, economical and political advancement and for privacy respect. Carissa Véliz asserts that 'Privacy and ethics have to be requirements from the very start of any tech project' (2020, p. 199), which is in accordance with the needs of the access to the algorithm source (source code), as is fought by privacy concerned activists in Europe¹³.

Hence, recommendations that assert that

'Supporting Open Source licences as the preferred licensing model for software involved in digital standards could hamper the development of the Digital Single Market. It would prevent the creation of technology-based start-ups in the digital field' (EARTO, 2020, p. 9)

should be carefully taken, as standards, by its own nature, must always be open, notwithstanding the speculative condition of the recommendation, as there is no evidence of what affirms, and the necessity for transparency when the source code is used by a public organisation and it manages private data.

This possibility of the scientist to tinker with the code must be guaranteed at any stage of the process. Due to the movable nature of the subject matter of any scientific discipline, the most appropriate IP licenses are the ones that promote openness, because they give permission to develop the scientific research while simultaneously allowing the sustainability of the project by all stakeholders, and not only by a subgroup of agents either authorized by the rightholders, either using an IP limitation, either operating in a scientific field where the works are in public domain, implying this last possibility the contrary to the state- of-the-art.

Finally, a recent report¹⁴ entitled The impact of Open Source Software and Hardware on technological independence, competitiveness and innovation in the EU economy (Blind et al., 2021) includes an extensive analysis regarding 'the economic impact of Open Source Software (OSS) and Hardware (OSH) on the European economy. It was commissioned by the European Commission's DG CONNECT'. As per its abstract:

It is estimated that companies located in the EU invested around €1 billion in OSS in 2018, which resulted in an impact on the European economy of between €65 and €95 billion. The analysis estimates a cost-benefit ratio of above 1:4 and predicts that an increase of 10% of OSS contributions would annually generate an additional 0.4% to 0.6% GDP as well as more than 600 additional ICT start-ups in the EU. Case studies reveal that by procuring OSS instead of proprietary software, the public sector could reduce the total cost of ownership, avoid vendor lock-in and thus increase its digital autonomy. (idem, p. 14)

The recommendations of this last document are structured along three dimensions: a digitally autonomous public sector, Open R&D enabling European growth and a digitised and internationally competitive industry (idem, pp. 314-343). There is no doubt that a better balance between IPR of free software authors and OS may be acquired through the development of public policies.

Recommendation for practitioners

FOSS communities have demonstrated for years a set of practices that might supply good examples for OS and intellectual asset management, which serve as recommendations:

• Licence diversity. The possible activities that the creator may allow the users to exercise are innumerable, although the side effect of using different licences is their possible incompatibility.

• Awareness of the necessity of including a licence to avoid the 'all rights reserved' by default system.

• Inclusion of licences within the source code. The licence is not a reference to an external document, but it is included as a text file in the source code trunk. This practice raises in the author the awareness of the necessity of a licence.

• Awareness that the licence may need an update due to changes in the technological or legal context.¹⁵

• Building tools for standardising the references to licences and to make them both human and machine readable.¹⁶

• Existence of communities who take care of projects' sustainability. Every project has at least one side communication channel, apart from the source code main version, which fosters contacts between the developers.

• Release early, release often. (Raymond, 1999, p. 28).

Recommendations for policymakers

Institutionalise FOSS through the design of European organisations to increase European Open Source institutional capacity.

• Creation of an Office for Free Intellectual Property Rights and Open Science (OFIPROS) inspired in the office subject matter of the Regulation (EU) No 386/2012 of the European Parliament and of the Council of 19 April 2012¹ and in line with the new IP Action Plan as stated in the New ERA and the New Industrial Strategy for Europe communications. If under a hostile context, free intellectual property models have built the Internet, the capacity of this model to work in a favourable context within the new strategy for knowledge valorisation should be explored.

5.2. Specific IP issues

In this Subsection four specific IP related issues are analysed. They represent four items where IPR have an impact in daily activities held by scientists when managing data. Awareness of these points would avoid the risk of IP infringement faced by scientists or the organisations they belong to.

5.2.1. APIs, taxonomies and ontologies

One of the most common ways to interchange data in ICT is through APIs (Application Programming Interface). An API is an endpoint in a server where a computer client connects and after making a petition, the server allows the client to copy data. This connection may be with or without authentication. The copied data is usually structured.

When an API is designed, several intellectual activities must be exercised. A necessary condition for an API is that the information to be shared must be structured into its relevant entities, that is to say, the developer creates a 'taxonomy' and decides how its integrants relate together, which in this case it is called 'ontology'. Once there is a decision made on the structure and the components of the information, it is necessary to determine which parts of the structure will be exposed, so as to be shared. Therefore, there are decisions on what 'categories' are relevant for the specific knowledge domain the API is designed for and which ones are chosen for sharing, building a 'pattern' of information. When a client machine contacts the server making a petition to the API, the server responds by passing the categories with an attached set of attributes that correspond to the defined categories.

An example of the data referred to a person can be self-explanatory: The relevant entity is a person, the categories would be the name, the surname, the age, the address and the nationality. When the API is called, the structure of the data is passed from the server to the client with the attributes that correspond to a specific person.

    {
    “name”: “Alice”,
    “surname”: “Alison”,
    “age”: “50”,
    “Address”:
        {
        “street”: “10, High Street”,
        “city”: “Brussels”,
        “country”: “Belgium”
        }
    “nationality”: “European”
    }

Fig. 5. Example of information and its structure collected from an API.

Therefore, in the functioning of the APIs, it is not only data what is passed to a third party, but also a way to represent (to structure) some information¹⁷. As it is known, one of the challenges ICT driven science is facing is how to build taxonomies and ontologies that represent the scientific domain. This materialized representation of scientific domains represented as taxonomies and ontologies and disseminated through APIs is an intellectual work regulated by IPR.

Recommendation for practitioners

In order to avoid friction points, it is recommendable that the legal conditions to access an API are expressily mentioned, regulating the IPRs through free licences to allow the use of the API (de la Cueva, 2008).

5.2.2. Hyperlinks and PIDs

Hyperlinks need special attention. When enriching data with metadata to accomplish the FAIR guiding principles, the data scientist adds some meaningful information to a dataset. The information is added, in a very significant number of cases (data provenance, PIDs, DOIs, ISBNs...), as a hyperlink in the form of 'http://...'.

Differently to the United States, a hyperlink has been considered in the EU an activity that implies public communication of the linked object. Therefore, a pdf version of this report read in the EU communicates to the public all the works that are referred to herein with a hyperlink, that is to say, the bibliographic items in the References section that include a 'http//...' and the hyperlinks in the footnotes.

In order to understand the problem involved, a series of antecedents must be explained. One of the IPR battles that were finally resolved by the CJEU was whether a hyperlink was an activity that served to communicate a work to the public. The problem arose due to the existence of websites¹⁸ whose only content were hyperlinks to files hosted in user networks connected through peer-to-peer software. The nature of a hyperlink is of metadata, it is data that represents other data, but it is not an intellectual work per se due to its nature of a pure reference, such as ISBN, DOI, the serial number of a computer or a car plate. Thus, a hyperlink is a linguistic element, a reference, that represents another object different from the hyperlink. US courts have understood this nature and have not included hyperlinking as an activity that produces public communication of a work although it may contribute to infringement, which is a different category. In the United States, linking is an activity that does not constitute a copyright infringement. The case of Ticketmaster corp. et al. v. Tickets.com Inc. decided on the 27 March 2000 by the Central District Court of California¹⁹, stated that:

'Further, hyperlinking does not itself involve a violation of the Copyright Act (whatever it may do for other claims) since no copying is involved. The customer is automatically transferred to the particular genuine web page of the original author. There is no deception in what is happening. This is analogous to using a library's card index to get reference to particular items, albeit faster and more efficiently'.

Also, in Canada the Supreme Court²⁰ resolved in the case of Crookes v. Newton, dated 19 October 2010, that

'Hyperlinks are, in essence, references, which are fundamentally different from other acts of "publication". Hyperlinks and references both communicate that something exists, but do not, by themselves, communicate its content. They both require some act on the part of a third party before he or she gains access to the content. The fact that access to that content is far easier with hyperlinks than with footnotes does not change the reality that a hyperlink, by itself, is content-neutral. Furthermore, inserting a hyperlink into a text gives the author no control over the content in the secondary article to which he or she has linked'.

In the EU, the judgment of the Court (Fourth Chamber)²¹ dated 13 February 2014, gave a different response in the case Nils Svensson and Others v Retriever Sverige AB. Even though a group of reputed scholars of the European Copyright Society (Bently et al., 2013) remembered the inventor of the web Tim Berners Lee's words, 'a standard hyperlink is nothing more than a reference or footnote, and that the ability to refer to a document is a fundamental right of free speech'²², the Court of Justice of the European Union (CJEU) declared that 'the provision of clickable links to protected works must be considered to be "making available" and, therefore, an "act of communication"' but, continued the ruling,

'None the less, according to settled case-law, in order to be covered by the concept of "communication to the public", within the meaning of Article 3(1) of Directive 2001/29, a communication, such as that at issue in the main proceedings, concerning the same works as those covered by the initial communication and made, as in the case of the initial communication, on the Internet, and therefore by the same technical means, must also be directed at a new public, that is to say, at a public that was not taken into account by the copyright holders when they authorised the initial communication to the public.'

A recent decision dated 9^th March 2021 of the CJEU²³, in a case that involved the German cultural heritage foundation Stiftung Preußischer Kulturbesitz and a German collecting society, concluded that the activity of framing (including in the display of one's website parts of another) is also an activity that requires consent from the copyright holder. The activity analysed in the conflict was described as follows:

'(10) The DDB website contains links to digitised content stored on the internet portals of participating institutions. However, as a "digital showcase", the DDB itself stores only thumbnails, that is to say smaller versions of the original images of the subject matter. When the user clicks on one of those thumbnails, he or she is redirected to the page concerning the particular subject matter on the DDB website, which contains an enlarged version of the thumbnail concerned, with a resolution of 440 by 330 pixels. When that enlarged thumbnail is clicked on, or the "magnifying glass" function is used, a further enlarged version of the thumbnail, with a maximum resolution of 800 by 600 pixels, is overlaid by means of a "lightbox". Further, the "Display object on original site" button contains a direct link to the website of the institution providing the subject matter, either to its home page or to the page relating to that subject matter'.

The importance of this ruling is that it summarises the case law of the CJEU related to hyperlinking. For the CJEU, hyperlinking is a communication to the public activity, but it understands that in Internet,

'(38) [...] in a situation in which an author gives prior, explicit and unqualified authorisation to the publication of his or her articles on the website of a newspaper publisher, without making use of technological measures restricting access to that work from other websites, that author may be regarded, in essence, as having authorised the communication of that work to all internet users'.

Therefore, the key element to understanding that there is no infringement is that the work is publicly available on the internet, and the access to it does not contravene any technological measures.

The CJEU understanding of hyperlinking as an act of communication to the public is bad news for ICT data driven science²⁴, because it produces, again, a hostile environment for OS. As stated by one of the core OS UNESCO's principles, Quality and Integrity: 'Open Science should support high quality research by bringing together multiple sources of knowledge and making research methods and outputs widely available for rigorous review and scrutiny'. There are two sources of IPR risk: first, OS projects are built to bring together multiple sources either by hosting or by linking to them; second, the purpose of data driven science is to establish connections between disciplines and this is made by linking ontologies to build what is known as Linked Open Data (LOD) (Fig. 6). Under the CJEU theories, when an ontology links to another, then an activity of communicating it to the public is exercised. It is not the main intention of this report to criticise how legal operators understand technology but to identify the friction points, and this is one. Therefore, special care is required when 'bringing together multiple sources of knowledge' using hyperlinking technology to ensure the source is not protected by technological measures. Nevertheless, as Canada's Supreme Court stated that 'inserting a hyperlink into a text gives the author no control over the content in the secondary article to which he or she has linked' then a hyperlink which is legal today can be illegal tomorrow due to a change of the linked object. This risk does not exist if the hyperlinking is made in Canada or in the United States to content hosted in any of these countries.

Fig. 6. The Linked Open DataCloud (Source: https://lod-cloud.net, 05-May-2021, CC-BY).

This IP risk is embedded at the core of the development of the EOSC. The report A Persistent Identifier (PID) policy for the European Open Science Cloud (European Commission. Directorate General for Research and Innovation & EOSC Executive Board, 2020) asserts that a PID (Persistent Identifier) is resolvable when it allows both human and machine users 'to access [...] a digital object, a digital representation, or information on how the object can be accessed'. PIDs are used 'to support a functioning environment of FAIR research' and they are a type of hyperlink with a defined scheme that 'can identify many different entities. These can be born digital (e.g. documents, data, software, services -otherwise known as digital objects -and collections made of them), physical (e.g. people, instruments, artefacts, samples), or conceptual (e.g. organisations, projects, vocabularies)'. Under the CJEU doctrine, a PID that identifies a digital object performs an activity of communicating the object to the public.

In conclusion, when hyperlinking to a third party content, it is not necessary to obtain consent except in those cases where the link circumvents technological measures. Of course, this rule does not apply to content whose licence allows its public communication. Nevertheless, special attention must be taken when hyperlinking, because as the CJEU case law demonstrates, the number of dubious examples is not small²⁵, therefore it implies an inherent risk of IP infringement.

Recommendation to policy makers

The recommendation of this report would be an EU legislative change to define hyperlinks as a mere linguistic reference.

5.2.3. Text and Data Mining (TDM)

Before the copyright reform included in the 2019 Directive, TDM has been discussed and studied largely in the EC context in different reports from expert groups (Geiger et al., 2018; Hargreaves et al., 2014; Triaille et al., 2014). In 2015, lead by the European Association of Academic Libraries (LIBER), a group of experts also made The Hague Declaration on Knowledge Discovery in the Digital Age²⁶. This declaration underlined the benefits of content mining for current research and claimed the exception of TDM, on the statement that IP was not designed to regulate the free flow of data, ideas and facts, but has the key objective of promoting research activity. In 2018, the European Parliament's Policy Department for Citizens' Rights and Constitutional Affairs at the request of the Committee on Legal Affairs (JURI-Committee) undertook an in-depth analysis about the proposed Exception for Text and Data Mining in the then-proposed Directive on Copyright in the Digital Single Market, which introduced a mandatory exception to allow carrying out TDM of protected works, assessing its positive and negative impact (Geiger et al., 2018).

According to Jean-Paul Trialle et al. (2014) 'mining of text' is just one subset of 'mining of data' and to keep it simple, data mining has been defined as the extraction of knowledge from data according to data science principles (Provost & Fawcett, 2013, p. 2). This technique may involve the activities of reproduction of text and data, and 'reproduction, translation, adaptation, arrangement, and any other alteration of a database protected by copyright'. Thus, it 'might infringe sui generis database rights, in particular the extraction---and to a minor extent the re-utilization---of substantial parts of a database' (Geiger et al., 2018, pp. 6-7). It is a technique widely used and necessary in data-driven science.

The copyright reform introduced by the Directive 2019/790²⁷ included an exception in article 3.1 'for reproductions and extractions made by research organisations and cultural heritage institutions in order to carry out, for the purposes of scientific research, text and data mining of works or other subject matter to which they have lawful access' and another exception in article 4 'for reproductions and extractions of lawfully accessible works and other subject matter for the purposes of text and data mining'. This proposal was criticised by The European Copyright Society²⁸ and the Max Planck Institute for Innovation and Competition²⁹ among other reasons due to the closed number of organisations who could exercise the activity, being only research organisations and cultural heritage institutions (but not data driven journalism), and the obligation of these organisations to store the databases in such a way as to prevent unauthorized uses by third parties. The final wording of article 3.3 reads as follows: 'Rightholders shall be allowed to apply measures to ensure the security and integrity of the networks and databases where the works or other subject matter are hosted. Such measures shall not go beyond what is necessary to achieve that objective'.

In the other hand, the UNESCO OS core value of Inclusiveness and of diversity, have the following meaning according the already mentioned Draft Recommendation (UNESCO, 2021):

'(iv) Diversity: Open Science should embrace a diversity of practices, workflows, languages, research outputs and research topics that support the needs and epistemic pluralism of diverse research communities, scholars, knowledge holders and social actors from different countries and regions;'

'(v) Inclusiveness. In the common pursuit of new knowledge, Open Science should meaningfully engage the whole scientific community, as well as the wider public and knowledge holders beyond the institutionalized scientific community, including indigenous peoples and other traditional communities, engages the scientific community as a whole, as well as the wider public and knowledge holders.'

Update

UNESCO Open Science Recommendation was finally adopted in November 2021. See https://unesdoc.unesco.org/ark:/48223/pf0000379949

The final text related to this matter is as follows:

'13. The core values of open science stem from the rights-based, ethical, epistemological, economic, legal, political, social, multi-stakeholder and technological implications of opening science to society and broadening the principles of openness to the whole cycle of scientific research. They include the following:

[...]

d. Diversity and inclusiveness: open science should embrace a diversity of knowledge, practices, workflows, languages, research outputs and research topics that support the needs and epistemic pluralism of the scientific community as a whole, diverse research communities and scholars, as well as the wider public and knowledge holders beyond the traditional scientific community, including indigenous peoples and local communities, and social actors from different countries and regions, as appropriate.'

As commented before, the exception for TDM regulated in EU Directive 2019/790³⁰ requires that the activity is conducted by research organisations and cultural heritage institutions. As per the Directive wording, these organisations are the only ones allowed to legally conduct TDM activities as the norm only includes these two types. But as it is widely known and fostered from EU policies, science is not only made by professionals who carry activities under a formal organization but also by a myriad of other agents who produce knowledge, including citizen scientists. This is evident in many EC funded projects, like Plastic pirates³¹ and it is defended by associations like the ECSA (European Citizen Science Association), but also in other countries like the US (Citizen Science Association) and Australia (ACSA). Science must be open to be produced by everybody, without exception, and the participants in citizen science are excluded by the current Directive to perform TDM activity. Therefore, its current wording is contrary to UNESCO's core values.

Recommendation to policy makers

The recommendation of this report is an EU legislative amendment to expand the TDM copyright exception to match UNESCO's diversity and inclusiveness values.

5.2.4. Levies on remuneration rights

As it has been evidenced through this text, IP norms, organisations, institutions and scholars mainly think of commercial IP, forgetting that all of them use for free (in both senses of gratuit and no IP restrictions) the internet protocols. This epistemic blindness has a logical side effect in the design of the IP exceptions that involve remuneration. Padawan case³², resolved by the CJEU in its ruling dated 21 October 2010, demonstrated that all the digital reproduction equipment, devices and media of a Member state were subject to the levy on private copy³³. As the ruling stated, 'the indiscriminate application of the private copying levy to all types of digital reproduction equipment, devices and media, including in the case expressly mentioned by the national court in which they are acquired by persons other than natural persons for purposes clearly unrelated to private copying, does not comply with Article 5(2) of Directive 2001/29'.

Another example may be found in the articles 5.2 'The right to obtain an equitable remuneration for rental cannot be waived by authors or performers', and 6.1 'Member States may derogate from the exclusive right provided for in Article 1 in respect of public lending, provided that at least authors obtain a remuneration for such lending. Member States shall be free to determine this remuneration taking account of their cultural promotion objectives.' of the Directive 2006/115/EC of 12 December 2006 on rental right and lending right and on certain rights related to copyright in the field of intellectual property. As it is evident, OS dissemination is subject to this remuneration that is collected through private agencies. The limitation to the exception included in article 10.1(d), related to the sole use 'for the purposes of teaching or scientific research' forgets the necessary scientific public dissemination. Therefore a commercial understanding of IP produces a burden on OS, with no compensation back from the collecting agencies to science. It is paradoxical that Wikipedia, an IP work where thousands of authors contribute, is absolutely ignored by the collecting agencies that are supposed to promote culture.

As final analysis related to this topic, the Directive 2001/29/EC includes in article 5 the following exceptions and limitations:

Limit	Right/Activity	Levy
Temporary acts	Reproduction	No
Reproductions on paper (photographic technique)	Reproduction	Yes
Reproductions on any medium - Private copy	Reproduction	Yes
Reproductions libraries, education, museums	Reproduction	No
Ephemeral recordings of works made by broadcasting organisation	Reproduction	No
Reproductions of broadcasts made by social institutions	Reproduction	Yes
Illustration for teaching or scientific research	Reproduction	No
Uses for the benefit of people with a disability	Reproduction and communication to the public34	No
Press within noticeable works [under conditions]	Reproduction and communication to the public	No
Quotations for purposes such as criticism or review	Reproduction and communication to the public	No
Public security, administrative, parliamentary or judicial proceedings	Reproduction and communication to the public	No
Use of political speeches and extracts of public lecture	Reproduction and communication to the public	No
Religious celebrations or official celebrations	Reproduction and communication to the public	No
Works of architecture or sculpture located permanently in public places	Reproduction and communication to the public	No
Incidental inclusion of a work in other material	Reproduction and communication to the public	No
Advertising a public exhibition or sale of artistic works	Reproduction and communication to the public	No
Caricature, parody or pastiche	Reproduction and communication to the public	No
Demonstration or repair of equipment	Reproduction and communication to the public	No
Use of a work for the purposes of reconstructing a building	Reproduction and communication to the public	No
Dedicated terminals on premises of libraries, educational or museums	Reproduction and communication to the public	No
Analog uses of minor importance	Reproduction and communication to the public	No

Table 10. Exceptions included in article 5 of the Directive 2001/29/EC.

Recommendation to policy makers

Adapt the current legislation to include clear and stronger exceptions for OS.

5.3. Basic science and IPR

Understanding IPR in only the commercial sense will miss an essential point for OS: the importance of basic science. Literature is full of examples on the importance of basic science, but two will serve for the purpose of this report. The first one refers to CRISPR, the second to PCR.

Around the early 90s, Francisco Mojica, a researcher at the University of Alicante, Spain, was trying to understand how archaeal organisms, Haloferax mediterranei, could survive in the coastal salt marshes of Santa Pola. When sequencing its DNA, he discovered repetitions which he first thought were errors that arose from the methodology they used. On the 21^st August 1992, Francisco Mojica was writing down the letters dictated by his assistant Francisco Soler so he asked him to be careful as he had repeated some sequences. Mojica checked the results and discovered 14 repetitions in the Haloferax mediterranei. Years before, in 1987, Japanese researchers under the direction of Atsuo Nakata had also discovered repetitions in the DNA of the bacteria Escherichia coli and in 1991, Dutch researchers under the supervision of Jan van Embden, discovered repetitions in the Mycobacterium tuberculosis. The three organisms were so different that an interchange of DNA between was out of question, so Mojica devoted his time to finding a plausible reason. He found it in 2003, when he discovered that the spaces between repetitions were part of the immune system of the organism. He coined the term CRISPR (clustered regularly interspaced short palindromic repeats) in an email interchange with the Dutch researcher Ruud Jansen. Years later, these repetition patterns were used by Jennifer Doudna and Emmanuelle Charpentier to create a tool to edit genes. Due to their tool, they received the Nobel Prize in Chemistry in 2020. (Montoliu & Martínez Mojica, 2020, pp. 40-59).

The second example begins with the North American microbiologist Thomas D. Brock discovering a high-temperature bacteria in Yellowstone National Park which he named Thermus aquaticus. This organism lived between 50 and 80 °C and his discovery was published in 1967 in the journal Science, which spurred the research on thermophiles. This research led to the understanding on how the enzyme Taq DNA polymerase works and its use in the technique known as PCR (Polymerase Chain Reaction). In 1993, Kary Mullis was awarded with the Nobel Prize in Chemistry due to his invention of PCR (Briones Llorente, 2020, pp. 82-83), which has been one of the most common terms in the recent COVID-19 pandemic.

The two above examples lead to the same conclusion: the importance of basic science as a propeller of future IPR (patents in the above cases). A good practical example about the importance of OS-OI in the context of new knowledge valorisation is the EU project SPOMAN³⁵ where researchers from different disciplines work together with companies and SMEs to translate industrial challenges into basic research. Promotion of basic science is key, due to its essential importance for applied science. Nevertheless, the evaluation of basic science through IPR (copyrights or patent) indicators should be further analysed because basic science belongs to a stage where the results are non obvious.

---

1. Although articles 3 and 4 of the Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market mention 'extraction', an extraction is only relevant under IPR if it constitutes reproduction (Geiger et al., 2018, p. 6; Triaille et al., 2014, p. 31) ↩

2. C/2018/2375: http://data.europa.eu/eli/reco/2018/790/oj (replaces recommendation 2012/417/EU). ↩

3. https://ec.europa.eu/digital-single-market/en ↩

4. https://ec.europa.eu/digital-single-market/en/digital-services-act-package ↩

5. http://data.europa.eu/eli/reg/2021/695/oj ↩

6. http://data.europa.eu/eli/dir/1996/9/oj ↩

7. This analytical approach has been first proposed and is being used in FNS-Cloud. Food Nutrition Security Cloud (2021) Deliverable D8.2. Report on FNS-Cloud Sustainability. The author of the deliverable is co-author of this report. Parts of the deliverable are used in this subsection. ↩

8. The M&M clauses take their name after a clause required in the agreements to be signed with the Van Halen music group. The agreements set forth the conditions of what was to be provided in the backstage of the concerts and one term was that there should be M&M candies but 'WARNING: NO BROWN ONES'. Although this seemed to be a pop star irrational condition, it was not: it was a way to check the care taken by the other contractual part in the accomplishment of the agreement conditions. Since then, it has been a tactic used by some lawyers to verify the attention paid to an agreement. See a copy of the Van Halen clause in http://www.thesmokinggun.com/file/van-halens-legendary-mms-rider?page=8 ↩

9. https://opensource.org/licenses/category ↩

10. https://www.gnu.org/licenses/rms-why-gplv3.en.html ↩

11. 'Copyleft is a general method for making a program (or other work) free (in the sense of freedom, not "zero price"), and requiring all modified and extended versions of the program to be free as well.' See https://www.gnu.org/licenses/copyleft.en.html ↩

12. See the US Recommendations of the National Institute of Standards and Technology Guide to General Server Security https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-123.pdf ↩

13. The resolution of the Dutch Rechtbank Den Haag, dated 5 February 2020 is a landmark in algorithm transparency. It obliges the Dutch Government to allow access to the source code of software used by the public administration. The court bases its decision in article 8 of the Charter of Fundamental Rights of the European Union (privacy).

    https://uitspraken.rechtspraak.nl/inziendocument?id=ECLI:NL:RBDHA:2020:865 (in Dutch). ↩

14. See https://digital-strategy.ec.europa.eu/en/library/study-about-impact-open-source-software-and-hardware-technological-independence-competitiveness-and ↩

15. General Public License version 2 was updated to version 3 due to 'tivoization' and software patents.. See https://www.gnu.org/licenses/rms-why-gplv3.en.html ↩

16. See 'The Software Package Data Exchange' https://spdx.dev/ ↩

17. A more complex definition of an API can be found in the United States Supreme Court judgment in the case Google LLC v. Oracle America, Inc. This ruling does not enter into the 'transmission' of information through copying, but on the legal nature of the act of Google copying approximately 11.500 lines of code. The US Supreme Court finally declared that the act was covered under the doctrine of fair use resolving thus that it did not constitute IPR infringement. https://www.supremecourt.gov/opinions/20pdf/18-956_d18f.pdf (5 April, 2021). ↩

18. Pirate Bay is the most notorious. It is still online. ↩

19. http://www.internetlibrary.com/pdf/Ticketmaster-Tickets.com-CD-Ca-Mtn-Dismiss.pdf ↩

20. https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/7963/index.do ↩

21. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62012CJ0466 ↩

22. https://www.w3.org/DesignIssues/LinkMyths.html ↩

23. https://eur-lex.europa.eu/legal-content/en/TXT/?uri=CELEX:62019CJ0392 ↩

24. And good news for the entertainment industry. Although all rulings from CJEU insist on the rights of the 'authors', it would be very interesting to know how many of the exclusive rights remain in the authors domain and have not been waived to the organisations or companies they work for. No data analysing this fact has been found. ↩

25. 13 February 2014, Svensson and Others, C‑466/12, EU:C:2014:76; 21 October 2014, BestWater International, C‑348/13, EU:C:2014:2315; 8 September 2016, GS Media, C‑160/15, EU:C:2016:644; 16 November 2016, Soulier and Doke, C‑301/15, EU:C:2016:878; 14 June 2017, Stichting Brein, C‑527/15, EU:C:2017:456; 29 November 2017, VCAST, C‑265/16, EU:C:2017:913; 7 August 2018, Renckhoff, C‑161/17, EU:C:2018:634; 29 July 2019, Spiegel Online, C‑516/17, EU:C:2019:625; 19 December 2019, Nederlands Uitgeversverbond and Groep Algemene Uitgevers, C‑263/18, EU:C:2019:1111; 2 April 2020, Stim and SAMI, C‑753/18, EU:C:2020:268; 28 October 2020, BY (Photographic evidence), C‑637/19, EU:C:2020:863. ↩

26. https://libereurope.eu/the-hague-declaration ↩

27. https://eur-lex.europa.eu/eli/dir/2019/790/oj ↩

28. https://europeancopyrightsocietydotorg.files.wordpress.com/2015/12/ecs-opinion-on-eu-copyright-reform-def.pdf ↩

29. https://www.ip.mpg.de/fileadmin/ipmpg/content/stellungnahmen/MPI-Position-Paper_TDM_2017-01-14-corr_def.pdf ↩

30. https://eur-lex.europa.eu/eli/dir/2019/790/oj ↩

31. https://www.plastic-pirates.eu ↩

32. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62008CJ0467 ↩

33. Even the CDs for the recordings of courts' public hearings paid the levy. ↩

34. Rights of reproduction and communication to the public on these exceptions can include the right to distribute the work. ↩

35. https://spoman-os.org ↩