Event opening

After two years of leading a global, multi-stakeholder conversation to co-design the Open Source AI Definition, the Open Source Initiative (OSI) is getting close to the finish line. Stefano Maffulli, OSI's executive director, will explain how the co-design process works, reveal the content of the latest draft of the Definition and highlight the next steps.

This presentation delves into higher education's impact on and benefits for open source software projects and their partner communities. Open source thrives through collaboration, innovation, and knowledge exchange; all core attributes of academic and research institutions. By engaging with universities and colleges, organizations can tap into a pool of aligned initiatives, diverse resources, broad expertise, and cutting-edge research. Likewise, higher education institutions benefit from real-world applications, industry insights, and opportunities for student experiential learning. This session will introduce strategic opportunities and best practices for fostering fruitful partnerships. Participants will gain insights into the immense potential and tangible value of such collaborations and practical guidance for initiating and nurturing effective relationships in the open source ecosystem.

From its historical dedication to education and women's rights, Tunisia has laid the groundwork for a thriving inclusion in various technological fields especially for women.
We will showcase key figures and approach to gender inclusivity, our journey and contributions related to technology and specifically the open-source field and discuss the opportunities and challenges, we face in nurturing Tunisia's Open-Source landscape.

Join us in exploring the groundbreaking success story of open-source 5G technologies that has transformed the landscape of telecom technologies. This talk will delve into the wider paradigm shift brought about by open-source initiatives such as Free5GC, OpenAirInterface, Open5GS and PacketRusher, which have shattered the barriers of proprietary systems (or have they?). Discover how open-source is making telecom accessible to a broader audience, breaking down the complexities of traditional telecom, and fostering interoperability in a once-closed ecosystem. Explore the future of telecom, where innovation thrives in the hands of a collaborative, open community.

The smart device revolution has given rise to amazing products and services. It has birthed unprecedented connectivity, powered the rise of social platforms, generated new industries and empowered people.
Looking forward, we see a significant shift towards open web technologies and open mobile technologies. A shift that will liberate developers from siloed models, encourage interoperability and champion fair competition.

Huawei invites you to step outside Droid’s house, go past the walled orchard and discover the free world. Let us show you a new landscape with Oniro and OpenHarmony, a future-proof open source operating system collaboration for a multitude of devices

The Oniro Project is a collaborative effort between the Eclipse Foundation and the OpenAtom Foundation, forging the world's first cross-regional open-source operating system project. Built on the robust foundation of OpenHarmony, Oniro leverages the strengths of both organizations:
- Eclipse Foundation's Open Source Governance: Oniro benefits from the Eclipse Foundation's proven experience in managing open-source projects, ensuring transparency and fostering a thriving developer community.
- OpenAtom's Technical Foundation and Developer Base: Oniro builds upon OpenHarmony's strong technical core and taps into its vast developer ecosystem.

This unique partnership empowers Oniro to create a future-proof operating system with a rich graphical user interface (UI) for mobile and IoT device. In this session, we'll delve into the various components and layers that make up Oniro.

The goal of the FusionIAM Community is to put together open source software used to build a global IAM (Identity and Access Management) solution.

Most of the time all the softwares present in the FusionIAM community are well know separately but never in the context of a global IAM solution. FusionIAM project will provide an easy way to deploy and configure the whole stack:

• LDAP directory
• Management interface
• White pages
• Web Services
• Authentication portal
• Access control
• Synchronization connectors

FusionIAM provides containers for an smooth deployment, either on a virtual machine or on a kubernetes server.

Presentation of the Data Visualization managed service integrated in Orange Cloud systems, based on Apache Superset open source SW.
Focused on: the perimeter of use cases / ReX & when to chose the solution (it value added vs other editor BI choices), how we did implement it in Orange in order to better meet business needs, its compliance to the company rules, including security aspects (authentication, vulnerabilities checks, backup & restore)

As more and more software companies decide to abandon open source licenses to adopt source-available or non-compete licenses, it is time to take a step back.

In this talk, we will look critically at those recent developments, replace them in historical context, explain the origin and value of the permissionless innovation that we currently all enjoy, and reassert the value of software developed in an open collaboration.

Daniel Merino from the Public University of Navarra and Bernardo García from Entornos de Formación and the Apereo Foundation will share their experiences about participating in the S2U Unidigital Project, a great example of cooperation between universities to grab common resources and request public funds in order to improve their education technology based on the popular open-source LMS software Sakai.

Here is an introduction article posted on LinkedIn about this initiative:

https://www.linkedin.com/pulse/s2u-unidigital-plan-daniel-merino-echeverr%C3%ADa-6mlof/

To synchronize their work in research and education, actors in these domains use more and more shared storage space such as Dropbox, or even better, collaborative office suite, to prepare syllabus, prepare exercise sheets, write articles or just exchange ideas. However, in these domains, people usually defer to known proprietary solutions such as Office 365 or Google Workspace.

However, as these solutions relies on US based service providers, they go against the “Schrems II” judgement of the Court of Justice of the European Union (CJEU), thus not fitting GDPR requirements for mandatory tools (as in education, where minors cannot legally express their consent).

It is therefore important to provide open source solutions, that can be deployed by governmental institutes to address this need in a user-friendly manner.

As a solution to this issue, we will show the collaborative office suite CryptPad that prioritizes protecting your data privacy (which can be important for research in progress), while staying accessible to most people, thus conciliating privacy and collaboration.

CryptPad is a web-based open-source end-to-end encrypted collaborative office suite, allowing its users to securely collaborate on multiple kinds of documents: spreadsheets, rich-text, diagrams… It also includes some tools to help collaboration in a broader sense such as right management, shared folders, team management, calendars, kanban, forms…

The Architecture, Engineering and Construction (AEC) sector has entered a new stage in the digitization of the design and construction of the built environment. Today, big companies, small startups, academic labs, are investing massively towards a brand-new paradigm in AEC :« platforms ». Traditional design software are moving from the desktop to the browser. In this context of a network-oriented environment, a new kind of flow-based workflow is emerging, combining heterogeneous design softwares. At the edge of this cultural shift, open source variant of Github such as Gitea or Forgejo, combined with the open stack of the Building Information Modeling (BIM) enable the creation of a new kind of« design forges » and new ways to conceive the act of design in a more collaborative way. We present Gitaec (gitaec.com) and Sloyd (sloyd.work) as two early example of such platforms for professionals and for education.

Forty-four years ago, Richard Stallman gave birth to the Free Software movement. Personal interest, an intellectual view about the pragmatism of software openness, and a deep-set belief about the morality of software as a commons, ran head-first into the growing shift towards closed and protected software. Clever use of copyright law to enshrine software freedoms kick-started a revolution. Yet despite the rise of the more business-friendly Open Source Software movement, many for-profit companies still view liberated software as an anti-capitalist endeavour. It's a common question: "How can something given away for free, even possibly used and exploited by competitors, make any financial sense for a business?"
This talk will explore the notion of value and why, when and how freed software makes good business sense.

Xerte Online Toolkits is a suite of tools that allow you to easily create beautiful, interactive and media-rich online modules. Xerte is open source and easy to use in various learning environments using xAPI, LTI, SCORM, etc.

If you have no experience with Xerte yet, this session is very useful to get an idea of what it is and what it can do. Various examples will give you an impression of what you can do with it.

What can you create with Xerte: online learning modules, mini websites, practice modules, formative testing, math, explanation/reference work, escape rooms, language training, flipped classroom, micro learning, adaptive learning, etc. It´s also very easy to share and reuse Xerte modules with others because it is open and it´s possible to use metadata.

It's an interactive session, so if you have any questions feel free to ask :)

Worldwide, we spend over 1 trillion dollars per year for the loaded cost of software developers.

If every company spent just 0.1% of that amount to fund open source maintainers, we’d unlock a billion dollars per year to fund the maintenance of open source.

That would pay the full time salaries for thousands of maintainers, their managers, security training, etc.

That seems fairly cheap for software that accounts for 70% to 97% of our software stack depending on how you count. And just imagine the positive impact it would have on the security of our software supply chain!

We’re way overdue making a clear distinction between open source developers and open source maintainers and professionalizing the latter.

In this talk, we’ll look at the current solutions to support open source software maintenance and their limits, the negative impact of feature-driven open source sponsoring, successful example of professionalizing maintenance in an adjacent field (Open Web Docs), what this could concretely look like for open source, and the benefits it would have for the community, the contributors, the users, and the end-users.

FOSSBot ( https://github.com/eellak/fossbot) is an open source and open-design DIY robot that has been developed by Greek Free and Open Source Software (GFOSS) society and Harokopio University of Athens (HUA) for educational purposes. FOSSBot provides a practical, hands-on platform for educators who wish to experiment with teaching science, technology, engineering, and mathematics concepts (STEM). In the last four years, GFOSS has expanded FOSSBot's functionality through the Google Summer of Code and FOSSBot has been presented in various open source conferences (e.g. FOSDEM 2023) and open access journals.

For users who cannot print and assemble the physical robot, we have developed an online coding environment that allows them to interact directly with a simulation of FOSSBot. Using the same software stack one can write and test any FOSSbot-related code in the simulated environment before running it in the physical robot and this allows to scale up training and expand the FOSSBot community of users.

With this talk, we aim to introduce our newly developed online platform that expands the accessibility to the coding environment of the robot through simulation, making it open and free for everyone. The FOSSbot platform enables everyone to access a development environment with just a few clicks and start developing robotic applications using a block-based IDE (a custom block-based visual editor that builds on Blockly) or a native Python IDE (a custom version of Monaco). The new web-based simulator developed using Godot allows for direct, installation-free use of FOSSBot simulation from the user's browser. Our platform is designed to be lightweight, enabling anyone, including schools with limited or poor internet connections, to self-host it locally and start using it. We have incorporated modern technologies such as PostgreSQL, Docker, React, FastAPI, and many more to provide a modern and user-friendly experience.

Soon the platform will support collaborative scenarios among users and classes. It will also facilitate direct programming with physical devices, including the FOSSbot, Arduino, and microcontrollers supporting MicroPython.

In OW2 we want to share with the community examples and lessons learned through the development process and invite anyone interested to contribute to our project.

This presentation will focused on the return of experience of 3 organizations of TOSIT association :

MAIF : 1. Funding of individual developers and 2. Creation in open rather inner source (mélusine, shapash...)

SNCF : 1. Creation of PostgreSQL.fr interentreprises 2. Major investment in TOSIT 3. Co-creation of open rail Foundation

ORANGE : Establishment of a network of 15 partners based on recurring annual budgets such as maintenance to sustain support for Orange and in turn, for the open source ecosystem.

This presentation introduces the UniTime open-source project to the OW2 community. UniTime started back in 2000 as a research project with collaboration between universities in Europe and North America. It has become a software solution that covers various educational timetabling and scheduling needs of a university, including course timetabling, examination timetabling, student and instructor scheduling, and event management. Over 100 institutions around the world are using it.

The UniTime project successfully transitioned from a very research-oriented start to a practical, open-source, web-based solution that can be used out of the box. However, it still has a solid background in research. Many related research papers have been published, and we co-organized an international timetabling competition, allowing us to contribute valuable timetabling benchmarks back to the research community.

The UniTime project has a tiny core team: only four people. In the second part of the presentation, we would like to concentrate on how our project is being funded and the current pitfalls of running such a project, especially in higher education and their procurement processes geared toward large commercial companies.

In this debate we will explore the impact of the new forthcoming European regulation and notably the CRA (Cyber Resilience Act) on OW2 projects and open source projects in general. How projects will be able to ensure that their downstream consumers can comply with the CE Mark requirement? We will also address more largely the state of the European standard system.

Open source and research are intricately linked, with the foundational principles of open source—such as sharing, peer review, and collaboration—mirroring the core values of the research community. Despite the prevalent use of open source tools in data science, including Python, R, notebooks, and the development of a wide array of open source softwares and libraries, these contributions often go unrecognized. Research institutions frequently overlook the significance of software production and maintenance when evaluating research impacts and career advancement for research professionals. Simultaneously, technology transfer centers struggle to harness economic and commercial value from open source projects effectively.

Research groups committed to open source face challenges in gaining institutional support and fear potential roadblocks from technology transfer centers when attempting to initiate discussions. This reality prompts a reevaluation of the current situation and calls for enhancements to more closely align research practices with institutional goals.

In this presentation, we will examine the state of open source practices in research, spotlighting the barriers to recognition and the ongoing and proposed strategies for overcoming these hurdles. We will explore solutions, including the integration of open source with other components like hardware and data, and underscore the significance of the citation of software in publications. For technology transfer centers, we will discuss the potential of digital commons and the necessity for specialized training on the nuances of open source intellectual property and community-based approaches, and the role to play by the open source ecosystem.

Fake news threatens democracies, public health, and news outlets’ credibility. For this reason, tackling misinformation is an open challenge faced by governments, private companies, and the scientific communities (Vosoughi et al., 2018).

There are many proposed approaches, some based on AI methods, others on fact-checking by human experts, still others on a combination of the two (Manzoor et al., 2019; Nakov et al., 2021). However, the fact that fake news detection algorithms are often owned by private social media companies, and additionally the adoption of “black-boxed” algorithms contribute to the lack of transparency in the fake news identification and filtering process.

Debunker-Assistant is an application that allows users and newspapers to assess the trustworthiness of a news item starting from its headline, body of text and URL. Inspired by the survey provided by Ruffo et al. (2023), it adapts ideas from Natural Language Processing and Network Science to counter the spread of online misinformation.

Its centerpiece is a set of four News Misinformation Indicators: Echo Effect, Alarm Bell, Sensationalism, and Reliability. These indicators are designed on the basis of specific linguistic and network features, such as the absence of sources, non-authority of references, the presence of specific figures of speech or flames, and other stylistic characteristics.

This application works with Italian and is not bounded on a specific topic: it is designed as a general purpose tool that can extract relevant features for assessing the quality of information.
Indeed, its main purposes are: 1) displaying the indicators to deal with misinformation; 2) de-biasing the mechanisms to make trustworthy the internet; and 3) showing insights about a certain context to aid the search and discovery of information.

Given the complex and ever-changing nature of content creation and information dissemination, there are several directions for improvement. For example, users could be involved in providing anonymous feedback on the news itself and on the characterization of the evaluated articles, improving the overall performing skill of the tool, more so for those features that are less explored in the literature. In addition, this type of interaction makes the user think about important aspects of the online information, thus increasing awareness.

Over time, as users search for new URLs, the core data that feed the models will expand to cover larger and more diverse sets of domains, incorporating a richer perspective on news consumption. To help the above research directions, we plan to develop a user-friendly interface and evaluate the general user experience. Finally, a future challenge would be to scale the model for other languages starting from English.

References
Vosoughi, S., Roy, D., & Aral, S. The spread of true and false news online. Science, 359(6380), 1146-1151, 2018.

S. I. Manzoor, J. Singla, Nikita, Fake news detection using machine learning approaches: A systematic review, in: 2019 3rd International Conference on Trends in Electronics and Informatics (ICOEI), 2019, pp. 230–234.

P. Nakov, D. P. A. Corney, M. Hasanain, F. Alam, T. Elsayed, A. Barr’on-Cedeno, P. Papotti, S. Shaar, G. D. S. Martino, Automated fact-checking for assisting human fact-checkers, in: International Joint Conference on Artificial Intelligence, 2021.

G. Ruffo, A. Semeraro, A. Giachanou, P. Rosso, Studying fake news spreading, polarisation dynamics, and manipulation by bots: A tale of networks and language, Computer science review 47 (2023) 100531.

OpenSource Science B.v. founded in 2022 in Breda The Netherlands, is the first institute of higher education, solely focussed on open source technologies. We aim to train open source devops engineers in our 4 year fulltime program. In this presentation our CEO ir. Erik Mols MSc BEd will explain how theory and practice meet in reality and how any open source company can participate in our curriculem.

A free, trustworthy, inclusive and respectful internet is one of the main visions of the founders of the internet, users and the authorities (e.g. https://www.ngi.eu/about/). Research estimates that approximately 96% of all software nowadays is open source or has open source at its root. Initiated as a response to proprietary code, open source is usually defined as source code made available to the general public for modification and redistribution, where human programmers build upon the work of others, adapting excerpts of the code, often in a collaborative process. As a decentralized software development model, with its main principle peer production, open source grew out of the F(L)OSS (Free/Libre Open Source Software) movement, where free refers to ‘freedom’ and not price, with source code openly shared without restrictions. However, there are also those who promote F(L)OSS and state that open source is the resulting business model built upon it. Yet why is it necessary to understand what open source is, its histories and how public funding promotes sustainability?

The evolving genre of FOSS/open source has been offered as a solution to the problems of (corporate) proprietary code; simultaneously it opens the door to software developers worldwide, who build new infrastructures and technical solutions based on shared resources and open code. The EU’s NGI (Next Generation Internet) funds diverse projects from a range of developers: SMEs, entrepreneurs, academics and activists––one of them is NGI Search. The vision of the NGI Search project is to ‘change the way we use and experience search & discover data and resources in general on the internet and on the web’, with the condition that all code from deliverables must be made open source. (https://www.ngisearch.eu/) Ethnographic interviews/surveys will be conducted with some of the developers to comprehend the ethics and values of the forms FOSS is taking within their awarded search projects and the sustainability of open source development through various lenses (financial, ideological, mental health, community, etc). This research delves into the ethics implicit in FOSS and open source business models, thereby contributing a theoretical framework/data visualisation mapping of humanising open source software and sustainability development.

Free software is increasingly financed by large organizations, which tend to steer projects towards their own specific needs. I will explain why this is not desirable for anyone and I will give feedback on the way I try to ensure the projects I participate to continue to remain attractive and useful for everyone.

The virtuous sides of the Open Source model are well known; at the same time, and without necessarily a causal link, many companies are migrating to SaaS-type models.

In this session, we’ll explore both subjects and see that a causal link does exist, but also that this relationship between Open Source and SaaS is actually more complex than at first sight, since it forms a virtuous circle that, in retrospect, looks like a no-brainer! Are Open Source and SaaS, two very strong trends in our industry, the perfect couple?

The OW2 best project awards recognize and reward "best of breed" projects and successful implementations of OW2 technologies.

Orange Design System (ODS) is Orange’s open-source toolkit for product design and accessible digital experiences. It offers a comprehensive suite including a web framework, web libraries, mobile SDKs, and various tools, all of which are open-source and either utilize or are based on existing open-source software.

Join us as we explore this suite and its integration within the open-source community, uncovering how developers actively contribute to upstream projects. Additionally, learn about the notable involvement of some developers who are members of Bootstrap's core team; whose our web framework is a fork of.

"Performing end-to-end testing is complicated", "Maintaining end-to-end testing is expensive", "Accessibility? We no longer have the budget".

These statements suggest that, even in 2024, users who depend on digital accessibility will be neglected, and that automated verification of business scenarios will require a certain level of expertise.

In this talk, we will explore whether these preconceived notions are justified and discuss the best approach to address these issues simultaneously.

On February 7th, France and Germany announced that they were collaborating on a sovereign digital suite based on open-source and interoperable solutions, aiming to create a coherent and comprehensive range of digital tools for public agents, including an instant messaging service based on Matrix.

In France, this effort is lead by DINUM. We will present this new approach and how we intend to take part in open source communities to help finance, maintain and support the development of open source solutions used by public agents.

TBD

Sakai (www.sakailms.org) is a leading open source Learning Management System used around the world. For a LMS, digital accessibility is doubly important because it can block a student's ability to learn. Over the past four years, the Sakai project has taken a novel approach to accessibility that is well beyond paying a third party for an "accessibility certificate" every few years. Sakai involves blind and low vision team members through the entire design, development, testing, and certification processes. Our goal is fostering a fully accessible community where all can participate and contribute. The Sakai accessibility lead is statutorily blind and participates in all aspects of the Sakai effort, including bringing those with disabilities into the project and mentoring them to become contributors. The accessibility team trains sighted developers and testers to do basic accessibility testing as part of their normal processes and provides consulting to any developer who wants a new feature or bug fix to be tested for accessibility. Building on the success of our accessibility efforts to date, Sakai is producing its own VPAT (Voluntary Product Accessibility Template) for the Sakai 22 and 23 releases. This presentation will show techniques that can be applied in other open source projects to improve their approaches to accessibility and eventually do their own accessibility certification.

Today, Open Source alternatives are missing when it comes to the search of complete suites integrating the most common use-cases of collaboration: email, chat, document & knowledge management, issue tracking, user management. IT departments looking for such suites are forced to turn towards proprietary solutions such as Microsoft 365 or Google Workplace.

However, alternatives are in the building. Among them, openDesk, an initiative lead by the German Ministry of Interior, stands out.

This project aims at building a complete Open Source alternative to Microsoft 365, where the software, the integrations, and the infrastructure required for project are to be released under Open Source licences. At the core of this project are several European Open Source software vendors, including Univention (user management), Open-Xchange (email), Nextcloud (file management), Nordeck / Jitsi (video conferencing), Element (chat), Collabora (document editing), OpenProject (project management) and XWiki (knowledge management). Operations on the project are also coordinated by Dataport, a service provider for the German administration.

This presentation will focus on describing the project, its governance, and describe how its different actors organize in order to foster integrations within Open Source products, and thus create a usable platform on top of them.

Evaluate the interest to relaunch OSAi community initiative and who is volunteering to participate, contribute, or coordinate?

Join us for an exciting update from the world of Collabora Online (COOL). Let us show you how users and integrators benefit from using a security focused, truly open-source, online office suite.
In this session we’ll show you why File Sync & Share and LMS provisions are integrating Collabora Online into their products. Hear about the work we’ve done over the past year to improve integration APIs, and make COOL an even better experience for its users. See what makes Collabora Online so feature rich and powerful.
For administrators seeking cutting-edge solutions, we have interesting new Kubernetes deployment options that are easy to install and automatically-scale. From seamless deployment and integration of external tools, to easy font management APIs and tooling to enhance compatibility across platforms – COOL makes your job a breeze.
But that's not all! We have infused the underlying LibreOffice technology with a host of essential features for users, building on our foundation of strong interoperability and collaborative editing. Hear about the latest features including usability, accessibility and interoperability improvements, as well as how we have improved performance across the board.
Learn how Collabora Online brings scalable, secure, on-premise document editing to everyone – allowing integrators to provide extra functionality to their offering, and users to stay in control of their data.

An Escape Game on digital Accessibility
Orange Accessibility department have created a escape game to raise awareness among Web users of any kind by confronting them with the difficulties encountered by people with disabilities.
Why this Open source Escape Game?
Generate awareness of the importance of making digital content and services accessible and respect WCAG 2.2 accessibility international guidelines to ensure digital inclusion.
Format:
One hour and a 20 minutes breakout session.
One hour for the escape game.
Twenty minutes of discussion and presentation around the issues encountered.
Animation held by accessibility experts.

Description:
This escape game aims to offer you an immersive experience, making you particularly aware of the challenges faced by people with digital disabilities.
You will be immersed in a scenario in which you will play a character with different deficiencies with each new stage reached.
Here you will have to solve the puzzles and challenges one by one, overcoming obstacles linked to accessibility and ergonomics to advance in the game.
It's up to you to get through it, alone or in a group, against these difficulties and not-so-super-benevolent developers.

In the last year we saw the rise of AI systems like ChatGPT, Stable Diffusion, Dall-E and others. Large Language Models like GPT are enabling a lot of new innovative features and products which will revolutionise the world.
But this large autoregressive language models come with a lot of challenges that can have negative effects on the Open Source and Open Tech community. For example it’s unclear if in the future everyone will have access to the same ML models and training data. Can students, startups and open source people build innovative new products using AI in the same way the open source communities build Open Code and Open Tech. How can we make sure that the AI system are not discriminating underrepresented minorities? What is the energy consumption and CO2 emissions of this new big AI systems?
This talk will discuss the current challenges around AI. In the second half it covers the new innovative local and open source AI features that the Nextcloud community is building. At the end the talk looks into the new Ethical AI framework and what it means for existing SaaS AI solutions and the open source local alternatives. It also covers the ongoing discussion to define an Open Source AI definition in the OSI community.

For the past 50 years, the computing industry has been following Moore's law, so we end up sending to trash millions of well-working computers every year. We can do better. In fact, now that climate change and biodiversity collapse are here and getting worse, we urgently need to do better. I have a plan. The good news is that Moore's law is slowing down. If we start optimizing the software stack we have, we'll realize we're sitting on a huge pile of unused ressources. This will enable us to keep innovating without having to build new IT equipment, which is the biggest environmental issue with IT. In short, we'll run tomorrow's services on yesterday's computers? We have to, we can't wait any longer.

At the FOSDEM '24 event, OpenLLM France evolved into OpenLLM Europe, embracing the goal of fostering an open collaborative space to develop partnerships among national initiatives aimed at creating open-source, sovereign, and efficient Large Language Models (LLMs) in the respective languages of each country.

OpenLLM Europe now boasts over 450 active members, receiving significant backing from the academic research community (22% of our members) and nearly 40% of our members are either users or companies specializing in AI.

Following the publication of CLAIRE in October 2023, our new model LUCIE has entered the pre-training phase. With 7B parameters, our primary objective is to address the underrepresentation of the French language in LLMs generally.

We have also incorporated other European languages, including German, Spanish, and likely Italian soon, alongside developing code to enhance the model's reasoning capabilities.

LUCIE is slated for 200,000 hours of pre-training on approximately 96 GPUs within the Jean ZAY machine, a sovereign supercomputer provided by our partner GENCI.

By June, LUCIE will be made available, having undergone transparent and independent evaluation by research teams from CNRS and INRIA.

This presentation will offer an update on the training progress, fine-tuning efforts, and the adoption of the LUCIE model.

Therefore, today in Europe, we have all the capabilities and, I believe with OpenLLM Europe, have generated the momentum, to build a real Open Source and sovereign European LLMs that meet our regulations.

The AI Act, which is the new European AI Regulation set to be adopted in April 2024, establishes a comprehensive legal framework for regulating the commercialization of AI systems. Its rules relate in particular to the transparency of the AI systems, and their strength depends in particular on how an AI system is developed. That’s specifically where Open-Source becomes key in AI development ! By fostering collaboration, transparency, and community-driven innovation, Open-Source enables the responsible and ethical advancement of AI technologies.

This regulation, through its cardinal principle of transparency leads to the creation of an exception to its application for AI system distributed under an Open-Source licence. In this respect, by recalling the conditions under which the Open-Source qualification will be retained, the AI Act specifies the criteria(s) under which it will not apply to AI system distributed under an Open-Source licence.

The purpose of this presentation is to understand the conditions for benefiting from an exemption from the application of the AI regulation for AI tools using Open Source. In this way, we will analyze the definition of Open Source adopted by the AI Act, a definition which emphasises the need for transparency through the desire to encourage developers to create models to accelerate the sharing of information throughout the lifecycle of an AI. This will make it possible to address the specific cases in which AI distributed under an open-source licence is exempt from the application of the regulation, and the direct consequences of this.

Local-first community created a fresh new architecture for our world dominated by cloud. In this talk, Andrey, will advocate for local-first architecture, delve into its unique futures, and provide real-world production experience.

This talk is helpful for every engineer because it contains multiple steps to improve privacy. From simple first steps without rewriting the app to profound revisions utilizing local-first architecture- there’s something for everyone.

A significant source of stress for maintainers and contributors to open source projects is mismatched expectations around credit and feedback for contributions. Sometimes a person makes a contribution representing months of work, only to have a maintainer lightly rewrite it and take sole credit. Other times a maintainer is overwhelmed by work, has no time to review unsolicited contributions, and is accused of plagiarizing work they never saw. And sometimes people game the system by sending large numbers of low quality, computer-generated, or plagiarized contributions to artificially inflate their apparent qualifications.

This talk describes one way to reduce this source of stress: a written contributions policy laying out what contributions are welcome, how they will be handled, and how the credit will be distributed. We will present a contributions and credit policy creation guide written by members of the RIPE Open Source Working Group which can be adapted and reused by open source projects.

The contributions and credit policy guide is available at: https://github.com/contribution-credit/policy

Real-time group communications involve many building blocks to bring functionalities like calls, chats or video conferencing. In an open world, relying on a single actor to supply all the technology required for such services to be live is not recommended.
Video conferencing servers are key components that need to be integrated in full-scale digital workplaces. Functions such as conference scheduling or participant management need to be tightly linked with calendars or address books.
The Linphone’s team has developed an open source video conferencing server that can be controlled following RFC6503, an open standard designed “to create, manipulate, and delete conference objects”. Therefore, there is no need for proprietary code to integrate conferencing management into a wider system.
As an illustration of the matter XCON CCMP intends to solve, this talk begins with an introduction to existing conferencing servers (Jitsi, Zoom, Teams, etc.) and their specific controlling APIs. Next, a quick introduction to this protocol. Then, we present an implementation made by the Linphone team in the Flexisip SIP conferencing server and how to control it from a third-party component. Finally, we conclude this talk by emphasis how open standards improve the resilience of complex solutions based on heterogeneous building blocks.

At first glance, it may seem obvious to associate Open Source and Open Science, whose synergies seem natural. Open Science is the unhindered dissemination of research publications and data. Open Source can be seen here as one of the branches of Open Science in the same way as Open Access or Open Research Data.
Open Source, which defends the freedom of access to and modification of software, therefore appears to be a perfectly complementary approach, both in terms of providing the tools and platforms for Open Science and in terms of the transparency and openness of results to the community. There are many remarkable examples of this with initiatives such as GitHub and, historically, Mozilla Science Lab.
But a more nuanced view, informed by the experience of these projects, particularly in business and public/private relations, shows us that there are obstacles. Cultural differences, intellectual property issues (with a wide variety of licences) and the challenges of funding and commercialising projects all come to the fore.
The need for companies to achieve a return on their investment can lead to constraints on intellectual property - or even patentability - that conflict with Open Science. Similarly, laboratories are increasingly seeking to add value to their work and collaborations in order to consolidate their budgets, which are not always sufficient. These constraints can call into question the Open Source nature of tools and results.
So we see that challenges remain, including in the management of open data and compliance with publication standards. Overcoming these cultural, legal or financial obstacles in Open Science projects requires a good understanding of the issues at stake for all the partners and the adjustments that need to be made. But this is the only way to build sustainable opportunities and real leadership for the Open Source and Open Science ecosystem in research.
By way of illustration, OpenStudio will be focusing in particular on the field of AI and its European challenges, drawing in particular on the experience of the Atlas of Productive Synergies, LongRun and Mellia, research and scientific cooperation projects (four CIFRE theses are currently in progress as part of public/private collaborations).

New this year : the conference will end with an open debate, giving everyone a voice, and promising passionate exchanges. All participants (attendees, speakers, exhibitors) are invited to participate in the discussion. The vote on the topic gave a majority of votes for "Open Source Funding" (50%), followed by "the role of Open Source foundations" (32%), and "Open Source and CRA" (18%).
Please go to the page: https://l.ow2.org/OW2con24Debate, and fill-in the quick survey to give an insight of the views you would like to put forward.

Closing Session