The Importance of Collaborative Applications for European Digital Sovereignty
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Subtitle |
| |
| Title of Series | ||
| Number of Parts | 542 | |
| Author | ||
| License | CC Attribution 2.0 Belgium: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/61666 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
00:00
Exponential functionDigital signalCryptographyInformationMobile appBitExterior algebraArithmetic progressionOpen sourceCollaborative softwareOrder (biology)Level (video gaming)SoftwareInheritance (object-oriented programming)Type theoryAssociative propertyCollaborationismComputer animation
03:15
SoftwareInstallation artServer (computing)Virtual machinePoint cloudSoftwareDifferential (mechanical device)Open sourceIndependence (probability theory)Internet service providerDifferent (Kate Ryan album)Chemical equationComputer animation
05:33
Local ringOpen sourceService (economics)Point cloudArmWaveComputer-generated imageryCollaborationismSpacetimeDigital rights managementAuthenticationVideoconferencingOnline chatOffice suiteWikiTask (computing)EmailData storage deviceSoftwareKolmogorov complexityInternet service providerStandard deviationVideoconferencingMobile appDifferent (Kate Ryan album)Service (economics)SoftwareView (database)Firewall (computing)Point cloudStandard deviationSoftware as a serviceAuditory maskingInternetworkingComponent-based software engineeringPhysical systemRevision controlComputer hardwareStatisticsCollaborationism1 (number)Decision theoryCloud computingStack (abstract data type)EvoluteGame controllerInternet service providerComputing platformCollaborative softwareParameter (computer programming)Open sourceComputer fileRemote procedure callPresentation of a groupLevel (video gaming)Point (geometry)NumberOffice suiteVirtual machineTerm (mathematics)Type theoryOpen setData storage deviceMultiplication signComputer configurationTouchscreenComputer animation
14:49
SoftwareKolmogorov complexityInternet service providerStandard deviationCollaborationismComputing platformPoint cloudDigital signalMobile appAuthenticationOffice suiteFiber bundleCollaborationismSoftwareInternet service providerGame controllerStandard deviation1 (number)Set (mathematics)Office suiteComplex (psychology)AuthenticationRenewal theoryPhysical systemMultiplication signOrder (biology)WikiPoint (geometry)File formatData structureDirectory serviceServer (computing)State of matterSoftware industryProcess (computing)FreewarePower (physics)InternetworkingCollaborative softwareFiber bundleDifferent (Kate Ryan album)BitComputer animation
24:06
Fiber bundlePoint cloudCollaborationismBuildingLocal ringOpen sourceRevision controlSoftwareStandard deviationMobile appOpen sourceBitLocal ringSoftwareCloud computingMultilaterationOffice suiteOnline helpProduct (business)1 (number)Point (geometry)WindowWeb browserBuildingDifferent (Kate Ryan album)Open setTrailCASE <Informatik>Internet service providerInternetworkingStandard deviationMereologyLink (knot theory)Service (economics)Shared memorySoftware industryComputer animation
33:23
Open sourceRevision controlSoftwarePoint cloudLocal ringCollaborationismStandard deviationVideoconferencingOffice suiteCryptographyOpen setDigital rights managementBuildingClient (computing)BitCore dumpSoftwareProjective planeSoftware developerOpen sourceLevel (video gaming)VideoconferencingCollaborationismRegular graphClient (computing)Point cloudRight angleMoment (mathematics)Goodness of fitStrategy gameConfluence (abstract rewriting)Multiplication signOpen setElectronic mailing listType theoryProof theorySystem callAxiom of choiceGame controllerComputer fileComputer animation
39:48
Self-organizationOpen sourceSoftwareNext Generation <Programm>InternetworkingLocal GroupOperations support systemLatent heatSuite (music)WindowCodeOpen setComputing platformDigital signalInstance (computer science)Physical systemPay televisionProjective planeBitEnterprise architectureMaxima and minimaMereologyInternet service providerClient (computing)Open setOpen sourceExterior algebraSoftwareCloud computingPoint cloudPoint (geometry)CodeSoftware bugComputing platformView (database)Message passingComputer animation
44:20
Point cloudOpen sourceSoftwareMachine visionWeb serviceFiber bundleChemical equationDisintegrationSpacetimeClient (computing)Projective planeOpen sourceOpen setSoftwareBusiness modelExtension (kinesiology)Client (computing)FreewareComplex (psychology)Chemical equationPoint cloudGoodness of fitMereologyMobile appInternet service providerDifferent (Kate Ryan album)Cloud computingEndliche ModelltheorieStrategy gameSoftware developerTerm (mathematics)MultilaterationINTEGRALOrder (biology)BitPhase transitionMultiplicationComputing platformIndependence (probability theory)BlogMultiplication signCodeComputer animation
53:58
Client (computing)Open sourcePoint cloudDisintegrationOpen setCollaborationismStandard deviationCollaborative softwareOpen setSoftwareGoodness of fitClient (computing)Open sourceCollaborationismStandard deviationComputing platformCloud computingParameter (computer programming)Internet service providerTransverse waveMultiplication signMultiplicationProjective planeComputer animation
57:31
Source codeSoftwareCollaborative softwareBitFile formatDecision theoryPoint cloudOffice suiteDistanceExterior algebraState of matterOpen sourceServer (computing)Physical systemComputer animation
01:01:24
Firefox <Programm>Slide ruleDigital signalOpen setGoogolOnline service providerPoint cloudInformation privacyPhysical lawDiagramComputer animation
01:02:50
Open setWebsiteServer (computing)Firefox <Programm>Address spaceComputer-assisted translationPoint cloudComputer animationProgram flowchart
Transcript: English(auto-generated)
00:07
Let's go. OK, so the goal also of this talk is to talk also about the progress and challenges of alternatives facing the big techs. So I want to first explain a bit why apps is key,
00:25
why end user apps is important, and also where are we? What are we doing with these apps? Do they exist? So first, who am I? So I'm Ludovic Dubost. I'm the creator of XWiki.
00:40
So I created the XWiki software. So I don't know, does anybody, who knows XWiki here? Cool, that's already nice. Who knows? And we also created a second software in my company. She's Kripad, who knows Kripad? Cool. And that's pretty nice. And so we've been doing collaboration software
01:00
for 20 years now, like 19 years. And we've been competing with software from the big techs. And it's not an easy thing to do. You get lots of challenges there in order to actually build that software
01:22
and get to the level that you can compete. And so I've done this. At the same time, I created the company. So it was a business goal to be able to live from building these applications as open source. And we have 50 employees.
01:43
We're a remote first company. And we have employees in France and Romania and also now in Germany. And actually, we hire. So if you're interested, come see what we do. And we are currently having more success. And that's not happening every year to be able to hire.
02:06
So if you're interested to see how it works inside the company that has a goal of doing open source, of building open source applications, not only using them or installing them, but really creating them, come see what we do.
02:21
So as a company, we're also a member of our ecosystem. We're a member of associations in France. For example, the Sistematique, OW2, the CNLL, it's Conseil National du Logie C'est Libre. We're not member of OSBA in Germany, but we actually want to be, and it's planned. We're also a member of Clidia.
02:44
So Appel is actually a European entity over CNLL and OSBA. We're trying to be active in order to try to promote the type of things that our companies do. And I'm a strong believer that European companies
03:03
should work together more than they do together. And I know how difficult it is, because it's super difficult, because we're all very focused on what we do every day. But collaboration is actually super important. So first, one thing is in 1995, technology
03:24
was more like you buy machines, you buy or you build software, you install that software. And not all activities in our lives were actually dependent on software in 1995. It was not as important. But in 28 years, actually a lot of things have changed.
03:45
Now you're renting cloud servers. You could even run your server just on the cloud without worrying about buying any machine. Just take the service. But you could also rent cloud servers.
04:00
You could build and make your own software and run it. But also one thing is that almost everything, like there is no business that's not dependent on software. It's starting to be really hard to find anything that won't use any software. Maybe some independent agriculture people might not need any software, but even though might still
04:22
use some of them to be more efficient in their job. But you have a lot of activities that are highly dependent on software for every day work. And it can be a major competitive differentiator to use software in your business.
04:43
We also see that there is a business imbalance. Actually, these are numbers. So I didn't put the source. I should have. Basically, the GDP of the tech sector in the US is 10%.
05:01
And there is data from Europe, from Europe's websites, where you see GDP is 4%. So it's not easy to know if they're actually exactly talking about the same thing. The scoping of what is the tech sector is not always easy to match from one country to another.
05:21
But we kind of intuitively know that there is much more tech in the US. And that actually makes a huge difference. And if we look at the problem from the sovereignty point of view, so what I tried to do here is to look at the different levels of what
05:42
we use in technology and what the situation is and how Europe competes from that point of view. But there are some differences also in the way it works. So for example, you might not know
06:01
how to make a hardware component. You might not have the people that can do it. But once you bought it, the hardware component is yours. And that actually is something that is significantly different from the top, where you might not know how to do the cloud service,
06:24
but you also don't own it. You don't have it. You use it. You use it remotely. Now, it's true. Even in hardware, you're seeing evolutions where the world of software, the way software works, of SAS software works, is starting
06:41
to happen on the hardware level. You can turn on a feature in a Tesla remotely. And you can turn it off also remotely. So we might have situations where we don't own the software that is inside the hardware machine.
07:01
So you have a lot of hardware that contains software. Even a chip can contain software. Whether or not they can turn off feature remotely from the internet is another story. But software is something that you can turn on, turn off remotely. And that's something very important. Now, we're not actually looking that good all over the place.
07:24
We know in Europe how to make some semiconductors. For example, Aram, a European company, was almost acquired by Nvidia. But it didn't happen. But it's a UK company. So we still own a semiconductor company in Europe.
07:43
We have some actors for hardware. But most of it is manufactured in Asia. So that's something important. And we know that with COVID, when we wanted masks in Europe, we didn't get them as fast as we wanted, as we needed them. So being able to manufacture our technologies can be important.
08:05
So now, cloud services, we do have European actors. But the good thing is that we have standardized open source tools in cloud services, so Linux, things like that, like Kubernetes, Docker.
08:21
These are open source software. We don't always have the competences in Europe. We are not the ones that build them, most of them. But we do have access to the source. And we have actors. But it's possible to run VMs all over Europe.
08:42
That's not a big problem. Now, when you go to pass software, platform as a service, you have a major dominance from the actors. In France, there's been a complaint about the fact that the Health Data Hub, which is a national service doing
09:03
statistics on health data, was run on Microsoft Cloud. And basically, a lot of people say, what? You're running health data of all the people in France on an American system. How is that protecting our data? It's not compliant with GDPR.
09:21
And basically, the answer of the people that had made these decisions said, there are not enough good APIs on the French cloud services to run it on a French cloud service. Obviously, everybody said, no, there are. You have enough. There is enough to. Maybe you're going to miss a little bit of stuff. Maybe it's not going to be convenient for you,
09:41
but it's possible. But basically, the argumentation was that it was not good enough. The ministry in France basically bashed the French companies for not being good enough. And the answer, of course, the ecosystem, what are you waiting for to give money to the French company to allow them to be good enough?
10:02
What are you doing instead of putting that type of service on foreign systems? So the technology, it's possible to develop it. A lot of it is open source, but companies like Amazon, or Amazon Web Service, Google Cloud,
10:23
or Microsoft are very ahead in terms of number of services that they have. When I looked as a company what technologies are available to deploy XWiki software and Cripad software in an app store on cloud providers,
10:45
I mean, you have different technologies on all of them, but they all have a lot of options for how you would sell your software on their app store. And when you look at what European companies have, they don't have systems to automatically deploy
11:03
applications, and sell them, and put a price on them, and people would start buying them. So they're really ahead. And when you go to SaaS, you have like thousands of services all around the world, startups. And we do have national actors,
11:22
and a lot of these actors are not open source. So you have a lot of SaaS services, and a lot are not open source. And even the ones provided by Microsoft, Google, and Amazon are not open source, or they can be based on open source, but they're not themselves open source. That means you cannot decide to self-host them,
11:40
and take control of them. And that's actually something that we'll see is important. Now, in the collaboration space, when you ask to the big companies what they need, they don't need one collaboration software. They need a whole stack. It was interesting in the previous presentation
12:02
about the co-op, it was about deploying multiple apps. And this is actually, bigger companies want all this software, they need all multiple software, and they need them to work well together. So they need authentication, user management, they need, of course, email, chat, video conferencing, file storage,
12:21
they want to edit documents, they want collaboration tools, and then they even want to build custom apps. So they want a lot of different software, and they want all this to be simple and easy. And when you start using apps, you become very quickly locked in.
12:44
Even when the app is open source, by the way. When you put your data in a system, the way you put that data in makes you quickly locked, because it's complicated to get the data out, and to adapt the way you work to other apps.
13:01
Even if the app is standard, in 20 years of XWiki, I've heard so many people say that, how they used to the fact that the button in Microsoft Office version X something, it's here on the screen. And you give them another tool where the button is on the right,
13:22
and they say, it's difficult to use. Like, okay, it's difficult to use. It's difficult to use because the button is on the right. No, well, they don't even know why it's difficult to use, but it feels difficult to use to them, because they're so used to the way they worked before that it's very difficult to change.
13:40
And so, and we have been trying to move people from Office to Wikis. So imagine how they tell you that it's difficult to use. No, it's not that difficult to use. As you need to adapt your way of working, you need to learn a new tool. It's not that difficult to use. It's just, it takes time to adapt.
14:02
You need to try to think how it works and try to adapt your way of working. So the lock-in is very high. And on SAS software, the lock-in is actually immense. Like, I worked at Netscape in 1995, and we were talking about the lock-in
14:21
of Microsoft on software when it was on-premise software. And it's like, oh, Microsoft controls your software, like you're blocked on their software. But at least the software was in your company, and you could put a firewall around it, and potentially no data would go out. Here, now it's on the cloud,
14:42
and so they will tell you it's gonna be double the price to upgrade, but you didn't have to upgrade. You could decide to stay on the older version for a bit longer, and you didn't have to pay every year. You paid licenses for a long time. Now you're paying every year, and if you don't pay on the 1st of January
15:04
for your renewal or every month, they turn off the software immediately. So you're actually in an immense lock-in. The lock-in is actually way higher than it was in 1995, when a lot of the industry was complaining about that lock-in.
15:21
And the data is controlled by the provider. There is a huge lack of standards on the structure of the data. So how that data, your basic formats are standardized, but in collaborative apps, even in Wikis, I have to recognize that the XWiki software formats
15:42
are not that standardized. We tried at some point to work with other wiki providers to work on common syntaxes, but never happened. Our syntax was based on something called Wikicraol, which was discussed 15 years ago, but now everybody's using Markdown.
16:02
So there was something that people said that should be the standard. A lot of wiki providers discussed, especially data, that's a good common standard, but it's a completely other one that a lot of people now say, ah, it's great. But not many people discussed the Markdown standard. It was just, oh, people liked it,
16:21
and it started to become very light. So the standards are very difficult to have. Even when they exist, they're not necessarily adopted. And so if they're not adopted, they're not very useful to move data from one system to another.
16:41
The complexity of the software make is very difficult to switch. And one key aspect of the software industry is a winner-takes-it-all approach. This is basically why VCs love the tech industry. They put a lot of money, they try to win, and they like the fact that when you win,
17:01
everybody else is dead. The winner makes 95% of the revenue. All the other people fight for the 5% that's left. And they love that because that's actually their goal, trying to be the winner. What's interesting is that they don't necessarily win that much because they create 10 companies or 100 companies, and only one makes it to being the winner.
17:23
And so in the end, they don't win that much money because they drop a lot of money on the way. But this is their goal, basically, of what they're trying to do. And one of the problem of the lock-in also is unfair business practices. You have a lot of unfair business practices
17:43
because the providers are so powerful that it's very tempting to use that power to make more money. So if you ask a business lead to make 10% more revenue next year, if there is no difficulty
18:03
to increase the price because everybody's locked, they will just increase the price 10%. That's much easier than work hard to try to get new customers. And so in order to keep that approach running, they also try to, or another approach is to, OK,
18:24
let's make other people buy my second software. I have a first software. If I can force the other people, if I can encourage them to use it, I give for free the second software, then at some point, they become used to it.
18:41
It's free. And at some point, I say, now it's not free anymore. Microsoft Teams is typically that approach. And Slack has actually complained about it. They actually sued about it because Teams was free. Office 365 is very expensive. They just gave Teams to everybody that has Office 365.
19:03
And at some point, they say, now you have to pay. And all the others are dead in the meantime. So you have a huge amount of risks of lack of sovereignty associated to this. So if you can stop all the software remotely,
19:24
if we disagree with the US government, knowing that all these companies are US-based, if we disagree with the US government, they could turn off the software. And even worse than that, by the way, is that they don't even need to turn the software
19:42
because they just need to remind our government that they can turn off the software. And then the other government will start being, want less to disagree with the American government because they know it's possible, or in negotiations,
20:01
it could be vaguely reminded that it's possible. And that's actually a big problem. And our ministry in France of economy and finance said no political sovereignty without digital sovereignty.
20:20
I'm not sure what they actually do to try to avoid that situation, but they actually acknowledge it that there was a problem, that at some point, you cannot state your political opinions if you're dependent so much on other people that have a different opinion than you.
20:43
And at this point, I'm not sure how many businesses would still function if you shut down their collaboration tools. If you shut down Microsoft today, what's the impact? It's a bit difficult to deny, but how long
21:02
would it take for the companies to start being able to do work again? And so from my point of view, that makes really collaboration apps very important because we don't have the control on these apps.
21:22
And it's the primary set of tools that people use. Actually, I didn't ask that question. In your company, which ones are using Microsoft or Google? And so it's the primary tool that people are using,
21:45
and they get really used to it. And it's also the primary authentication system that people use. Most companies will put Azure AD, LDAP, Active Directory,
22:00
not LDAP, Active Directory as the authentication system. And basically, they end up also being almost ready to deploy servers. So when you have Azure AD, you have an Azure account, you're one click away from deploying servers
22:22
on the Microsoft system. So people buy Office 365. They tell people, ah, you need to connect Office 365 to Azure AD to manage your users. And then you're one click away to deploy servers. And a lot of users like simplicity. If it's simple to click, they start using the other service.
22:44
And they do a lot of work to make all this work as easy as possible together. Do they do a good job is another question. But they have the possibility to make it easy. We might be happy that they're not always very good at it.
23:02
And so that gives us some opportunities to provide other solutions. And so it's the entry point of all software. And the biggest problem, from my point of view, has been the bundle issue, is that a lot of software
23:22
becomes bundled as a package. And so when you go to Microsoft's website to buy, you say, I need Office, you will be proposed the whole Microsoft suite, not only Office 365. And you will have this software plus this software plus this software. And you'll see that the difference of pricing
23:42
is not always very high between the package with. And I'm not even sure it exists only Office 365. Only Office, you will get for free one thing, even what you don't need. This was actually the problem of computers,
24:00
bundled with Windows, which were bundled with Internet Explorer. I said I worked at Netscape. And I can remember how it killed Netscape's business around the browser when Microsoft bundled Internet Explorer in Windows. Who would buy a browser if you have one for free with Windows when you get your computer?
24:22
And at the time, you were paying Windows. Well, in the meantime, there's a lot of things that have changed because now browsers also got free because you can make money through advertisement. But that's actually potentially a problem because you can make money through advertisement, but we also got tracking thanks to that.
24:42
So we get browsers for free, but we get tracking. And that's also a problem of sovereignty. So now they bundle even more. They bundle apps, but they also bundle Office 365 with the cloud, with the pass, with the yes. Microsoft has a whole offer.
25:01
Servers, yes, a pass, and apps. And when you go see a client with you, ah, I have a wiki. Would you like my wiki? Sorry, I already have it for free in the whole offer. And I think I had a few cases like that.
25:22
For a department that's part of a big company, if they buy another software, they pay for it. If they use Microsoft software, the central entity pays for it. And it's divided, and they pay the same share of it
25:43
whether they use it or not. So basically, you come as a provider, and you have a cost. And the Microsoft software you're competing with has no cost. It's free. But it has a cost in reality, because some people will start
26:02
using some of the software. And when they renew the license with Microsoft three years later, Microsoft comes and counts everything that is being used, and they set the new bill. And so the company will end up paying it, the fact that they
26:21
started using that software. And so it's very difficult to compete. And so everything's bundled. So actually, Nextcloud has sued on this. I think antitrust.nextcloud.com, or dot com, I think. I think I have the link a bit later.
26:42
And so because it's a real antitrust issue, in the end, you cannot compete with that type of situation. And so some people say, and especially this is true in France, in our government,
27:00
they say we need to build better software companies. So the problem is not the way the competition works or the way these companies operate and so on. That's not the problem. The problem is we don't have a French one or we don't have a European one. And so say, let's build unicorns and make companies.
27:22
But in reality, when you look at unicorns, if they actually get very big, if they manage to get very big, they will behave exactly the same way. And it's not going to help the people and the small companies. It's going to be a bit less a sovereignty issue,
27:42
because maybe our government will have a little more control on them. But we are not going to get our own sovereignty, because we're going to be very dependent on this big company. The other problem is that most of these unicorns are made with VCs.
28:02
Most of these VCs are actually not even European. And VCs want to make money, so they want to sell the company at some point. And it's much easier to sell it to the ones that have the most money, which are the current American companies. So in the end, there is a very big chance that all these unicorns will either die,
28:23
because they don't work well enough, or they're going to be bought by the same companies that are currently leading the market. And another thing, they usually use the American cloud providers
28:41
for their services. And so these cloud providers see everything they do. So when something works well, they can replicate it. This is actually a complaint against Amazon, even in the US, that small businesses that sell on Amazon are showing to Amazon everything that works. And surprisingly, at some point, you get products that pop up made by Amazon instead,
29:03
and they take the market. And so they're supposed to not look at the statistics, but we don't know what really happens. Now, how can open source help to solve these type of problems?
29:23
So one thing I believe is very important is that if you try to build local companies, French ones, German ones, UK ones, even though UK is not in Europe anymore, but still in Europe physically, but if you try to build your own local companies,
29:43
and you all do it at the same time, at some point, you'll have a problem. And because suppose one of them buys the other one, you're going to have one very big French company or one very big German company, and maybe, I don't know, a Swedish one.
30:01
And the problem is that this won't be sovereignty for the countries that don't own these companies. So you might have a bit of sovereignty in France, but no sovereignty in Poland. And so if you try then to tell, so let's invest all,
30:22
let's start using the software from this European company. Like in Poland, they're not very interested to use the software from a French one. That doesn't give them that much sovereignty. So an open source works very differently. If we invest in open source in France,
30:43
it can be used in Germany. And it can give sovereignty to Germany because Germany can control the software. German companies can build competence on top of that software. And that's true in Poland, and that's true in Romania, et cetera.
31:01
So if we all, in different countries, co-invest in open source, we can all get more sovereignty out of it. And we also can get individual sovereignty out of it because any individual can then make use of it. So I'm sure I don't have to convince that much
31:21
about open source here. But it's important to remind these arguments, which can be useful if you meet people in your own countries. So also, open source is usually a bit more open to standards. And if the provider doesn't want
31:41
to adapt this software to standards, then somebody else can do it. You have this possibility. So open source will allow to adapt software to standards when providers don't want to do it because it's not in their economical interest.
32:03
It's very hard. If you look at software, you very often have a lot of import tools. You very rarely have export tools. But anybody can build the export tools on open source software. And one aspect that is also interesting is that we can reuse also US open source software.
32:27
But we need, however, to build local competence. If we use massively very large US-based software, but we're not building local competence on it, we will risk having a lack of sovereignty
32:44
because we don't have the people that can take over. So we can reuse open source software from anywhere in the world. But it's also important to invest in the competence. So if we start building massive solutions, and not only
33:03
competence about how to use them, but competence how they're built, how they're made inside the software, it's very easy to use open source. It's much more difficult to improve an open source software. You need to build competence on the software itself.
33:23
So now I want to look a bit at what actually exists. Because we're looking, and I've been maybe a bit looking pessimistic, because it's hard. It's tough. But actually, what's interesting is
33:42
that we have tons of great open source software in Europe. And I'm sure I missed a lot, because I spend most of my time on our own software. And I'm sure I missed a lot. But I mentioned the list of types of software that companies are looking for for collaboration.
34:03
And so for example, in video conferencing, we have software like Jitsi, which has been created in France. Actually, core developers are in the US now, but it was created in Strasbourg. Big Blue Button, Canadian. It's open source, not developed in Europe.
34:22
Next Cloud Talk, developed in Europe. OnlyOffice is developed in Lethony, if I'm correct. There are also issues around OnlyOffice, because it's linked to Russia. So we also need to have more control on it,
34:48
on the development of it, Collabora, on collaboration, Next Cloud, XWiki, Cripad, project management, projects like Tulip, Open Project. So we have a lot of software on authentications,
35:01
software like lemon-eldap, Univention. On email, we have OpenExchange. We have SOGO, BlueMind. And on file management, software like Next Cloud, PDO, Cripad. And on chat, we have great solutions like Matrix. So we actually have all the bricks existing.
35:26
Now, they're all provided by different companies, different providers, and they do. They're not as integrated as the other solutions will be.
35:40
And another aspect, I'll come back to that, the integration aspect. Another aspect is that actually building open source business, it works. All these software have core teams. The software I mentioned, they do have core teams.
36:00
Most of the companies, most of the software have a core team that is working for at least one company that has regular revenue out of this software, and they're able to fund the roadmap and have a regular development on top of this software.
36:21
So it's possible, and I can attest of it, that with my company, we didn't raise any money. And for 19 years, we've been able to pay the people in the team and grow, and we've even been able to create a second software out of the work
36:47
that we were doing on the first one. So it's possible to fund the company. Now, at the beginning, it's not easy. Like, the first years, you need to build critical mass.
37:02
And sometimes, you have an issue of competition. So you're not necessarily finding enough money to build it quickly at the level that it can compete. But it's possible to find business, and it progresses.
37:21
The better your software is, the more companies are interested in it. When it's good enough, the companies get interested in it. So you have moments where companies will just go to the other provider. The winner takes it all. So you get very little market share, but then at some point,
37:41
they get more interested, and you get more prospects. So how does open source business work? Clients do buy support, but you need also to have the price right. So if you put it too expensive, then clients
38:00
might just go away. But if you put the right price, you get clients that buy support. You also have clients that sponsor development. But you have issues of dumping and bundling that slows it down. You cannot sell to everybody because you have competitors,
38:23
and their software is given for free or cheaper. And XWiki knows it very well because we've been competing with XWiki with Confluence. Anybody has Confluence? Is there a company? So while you can try to convince your companies
38:40
to switch to XWiki, we have importers. And Atlassian has kept increasing the price over the time. And they actually changed a lot the strategy. So they had a pretty low price for small businesses. And now they say, small businesses, you have to go on the cloud.
39:00
We stopped doing the small price software, even though it makes money. If you look at the finances of Atlassian, it does make money. So it's a conscious choice to make more money. It's not something that doesn't work. It's something that works. But now that they've done that, we have a lot of calls.
39:20
And we get much more clients. And even though we never got the money of these clients to build our own software, so we've been able to build the software with the money of other clients. And now we're having more clients coming. So I think it's a clear proof that the dumping and the bundling approaches are slowing down open source.
39:44
And there are unfair practices. There are also some good news. For example, European initiatives, NGI. So we at TxWiki, we've been able to benefit from that on Cripad in particular, because Cripad doesn't have that many clients.
40:02
There has almost no enterprise clients. It's just starting to have enterprise clients. And it's been mostly funded with research money. We do have some clients. We have donations and subscriptions on our instance. But it would not be enough to pay for that software, which
40:23
requires a lot of work. So NGI is providing funding for companies. And that's really a great thing. There are also the EU starting to pay for bug bounties on open source software. We've been able to benefit from it also.
40:44
And that's been great. GaiaX is interesting, but there are also issues with it from my point of view. It's not easy to understand, to see how four small providers like us that are doing apps,
41:01
how to work together with the big cloud providers and big companies that are currently running GaiaX. We're not seeing also open source being as put in front in the GaiaX project. So there are initiatives in Europe to help.
41:21
And we've seen that it's becoming more and more in subject. That's interesting. In France, we have the French government decided to put some funding. So on one side, they're opening the door to American software. They also reacted to the fact that from the GDPR perspective,
41:42
there is a huge problem with the Cloud Act. This is the Max Schrems Act. And basically, they're recognizing that it's illegal to host data on American companies' systems. But it seems to be illegal, but we don't really
42:01
know if they consider it illegal for everybody. Is it illegal only for the public service? Is it illegal for all the companies? It's unclear, but they started to say, OK, for example, the National Education should not use Microsoft, should not deploy Microsoft software anymore.
42:23
But on the other side, they're proposing that French companies are going to build clouds to deploy Microsoft and Google software on them and host it instead of Microsoft and Google. And that will be OK.
42:40
So they will control the software, but they won't control the hosting, but they won't control the software. On the other side, they say, OK, let's fund a little bit alternative. So they decided to put 50 million, but they didn't put it on open source. We are partners in some of the projects to try to get funding for XWiki and Cripad
43:01
and try to build also additional solutions. So you have a bit of a mixed message. On one side, they like open source. On the other side, they don't promote only open source, so it's not always clear. So you have messages.
43:20
There are better news in Germany, where you have specific funding for open source platforms with the Sovereign Workplace Project. Actually, at XWiki, we have had the chance to be called in, and that is actually great news for us. And you have a bunch of great open source software
43:42
that are part of that project. You also have additional initiatives like the Sovereign Tech Fund, and there is a specific entity called the Centrum for Digital Superinitate. Then this. My German is good. And you have the Open Code Project, so a bit better news.
44:03
But at the same time, it's never completely won. You have internally at government some people pro and some people that will continue to push for, oh, it's not so important. Let's use American software.
44:21
So the challenges, I've talked about them. The antitrust is clearly bundling and unfair practices and trying to kill competition with aggressive pricing. I've seen cloud develop, and the typical cloud business model
44:40
is let's give things for free and then increase the price. Let's get the users when they're in, when I'm sufficiently known. Let's change the price. You could say open source does the same thing. Let's make it open source. And at some point, let's give it for free. And let's change the price, or let's stop being open source.
45:05
So some companies do it. There is a big difference. However, is that what you put as open source, you can never take away. So maybe you can change your price, stop, do open source, and do your own life and turn your back
45:23
on the people that helped you, because you have a lot of people that help you when you do open source. You give them a lot, but they also give you a lot. I can say that I have a lot of people that gave us things, gave us contribution, make the software known.
45:40
The user platform wouldn't be what it is with all the people making it known. So we have a lot of people helping. So you can do that, but you can never take away the code. And that's a very, very important difference versus a strategy of the cloud providers that gives things for free and increase the price later.
46:02
And this is also why on my side I'm a big proponent of a GPL with no CLA, so no retaining the copyright of everything that people give you, because this shows that you will not turn your back on open source,
46:21
or at least you will have to give back all the code and redo what people gave you. For Cripad, it's a GPL. XWIC is LGPL. And we don't want to change it, because this is what we did. And we don't want to change that.
46:40
But Cripad is a GPL. And we have never taken any CLA on contributions. And we believe we cannot go away of what we're doing. There is another challenge, which is open source financing. So the initial financing of an open source project
47:01
is very difficult. It's something that is tough. There is one thing that is very important, is that a lot of people, when they build open source software and they try to live from it, they question the open source business model. Is that, yeah, but somebody is going to steal my software,
47:22
somebody is going to compete with me, et cetera, et cetera. And you have difficulties. You will always have difficulties. But sometimes it's very important to understand that the main difficulty is not the business model. The main difficulty is do a good enough software. In many cases, when you tend to think
47:44
that you should be selling it, you should be selling it instead of giving it away as open source, well, sometimes that's not the problem. The problem is that the software is not good enough. You need to be better. The competition is too good.
48:00
And you need to work more on your software. We had phases where it was a bit difficult at XWiki. And what we did, we worked on the software. And things improved. So you have to look at that. We also decided at XWiki to do extensions
48:20
that are open source but paying. And so I usually have a bit of a complexity to explain that. Like I say, I even had some people that said, so they asked, but can I take it? Or can I reuse it for free?
48:40
I said, yeah, take the source, rebuild it, remove the license check, and you can use it. It's open source. So it's not open source. I cannot use it for free. No, that's not open source. It's about the source. It's not about whether or not it's easy for you to use it. So we decided to have some extensions of XWiki
49:03
which are open source. But either you go find the source, and you go and take it and build it yourself and install it yourself. Or you use our app store, and in one click you can use it, pay for it, and buy it. And it's been very useful for us because it allows also to explain
49:21
to companies that concept. The fact that the important part is the fact that it's open source. It's not that it's free. That's what is important in what we do. But it allowed to explain that, and it helped actually improve the relationship with some of our clients which then naturally say,
49:41
okay, if I don't have to pay, I don't pay. I don't help the project. I don't need to pay. So you have to find that balance also between the commercial offers and the freedom. Sometimes when you build open source, you're too idealistic in a way that, okay, everything should be free.
50:01
But the problem that when everything's free, you have a lot of people that don't understand that they actually should help the project because this project will not survive without funding. You need money to survive, to make the software. It's necessary. Yeah, another aspect is that it's actually very difficult
50:21
to make partnerships with service and cloud companies. Right now, XWiki has not been able to do good partnerships with other service companies to distribute our software. It's not simple, and I'm not really sure how to do partnerships with cloud companies.
50:41
Cloud companies can be very tempted to just reuse all the open source software available and sell it and not fund in any way the software that are in these offers. And we need a model for that. So if we want to separate the work
51:01
of cloud service and of software, we need to find a model to how to sell open source software when it's sold to companies. Otherwise, the software will not be able to fund themselves. So Matrix, for example, has made a blog post
51:20
about the fact that they have major success in terms of usage and deployments, but many of these deployments are made by other companies that are making money off of it and are not giving back anything, not even developers, to work on the project. Neither funding, neither developers.
51:42
And in the end, they had to fire people. And so in order to, I mean, if you want to do good software, you need great people. You need great developers, and so you need to find a funding model that works. One aspect for open source company
52:01
is to work with less cash. So I don't believe in VC investment because I think that doesn't allow to keep the independence of what we build, but you have to work with less cash. So even when you're growing, it's difficult to hire because you need a lot of cash
52:22
to run the company itself, like to advance the money, and you are taking risks, and it's not a simple problem. Another problem is integration and fragmentation. So clients want things that are integrated. That's actually also one of the reasons you start to see packages that bundle the software,
52:43
so trying to make them all installable together easily. That's actually something that's important. And cloud providers, providing multiple software together. This is very interesting, but there is a lot
53:01
of integration work that is necessary, and sometimes that work is not brought back to the projects, and we need to make that happen. There's also a lot of fragmentation in the EU market, so you need to go to multiple countries to sell. It takes a lot of time, so we need also partnerships
53:21
between the companies and the different countries to work together in order to make that work. I'm very happy to have partnerships with Univention, for example, Peter, because for us, it's a major win of time to go on the German market.
53:42
If a company says, look at this French company, it changes a lot the way the German market will see a French company, and it's gonna be true the other way, so we need to work more together. So that's a bit the next step I have.
54:02
So first is we need to believe in ourselves. So we have great offers, and the companies that are building software, open source software, are able to make it, and the difficulty sometimes they have is about competition, is about being considered good enough so that a lot of companies invest in them.
54:23
There is a huge amount of money sold on collaboration. Like I think we count billions of euros that are given to Microsoft and Google, and if we sum all the open source company we were talking about, we are talking about 50, 100 million,
54:41
something like that, revenue, I don't know exactly because it's not always easy to get the numbers, but we're spending billions on collaboration software to American companies, so there is a lot of money. It's just we need to convince the clients
55:01
that we're good enough as a group, not only as individual company, and open source provides good openness and flexibility, and that's useful to clients, so there are also good technical arguments for open source when it comes to company. We have had a lot of clients
55:21
that were very interested in the flexibility of our software, but they also want easy to go solution, one provider to buy from, and that's a real difficulty. We need to work more on European collaborations, I talked about it, so at XWiki this is something we're trying to do more,
55:41
so the intervention partnership we did last year, and we're also looking at partnerships with NextCloud, with OpenProject. There are multiple companies in Europe that think the same way. They just lack a lot of time to actually make it happen, and so it's going to take time,
56:02
but there are many companies that want to do that, and Sovereign Workplace Project is really great to create also this integrated offer, bring the providers together. Now we will need to find ways
56:20
to do partnerships with cloud providers. I've yet to see how this will happen, and we need more open standards. That's also another aspect that's transversal. We need more open standards. I've also been a believer that we need an open source marketplace, so a way for cloud providers to sell open source software in one easy way,
56:44
but I hope that Gaia X would be a place for that. I haven't seen that happening, and the most important thing we need is clients, so any clients that are brought to the companies
57:01
that are providing this alternate platform are helping these companies bring the software a little bit better, and it makes a difference, so everything that as individual you can do to tell your boss, employers, that maybe they should look at what these companies are doing,
57:21
that other companies are actually deploying this software is helping, so if you can help us, do it, and make it known, and that's it. If there are any questions, I think there's two minutes left.
57:49
Everybody fall asleep. One. Hi. You mentioned earlier in your talk that the French government deemed it illegal to host data on American companies
58:00
because of the U.S. Cloud Act. Could you expand on the Cloud Act a bit more? Yeah, so the Cloud Act is the fact that the U.S. government is able to access data from American companies and even from American people,
58:21
so first, I'm not a lawyer, okay, so don't take everything I say for granted, so it allows to access data even if that data is in a foreign country, so, and they can even, I understood from a paper in Holland by lawyers
58:41
that they could even ask to an American person secretly to give the data that these people can access, and they cannot refuse, basically, like they risk a lot, so Microsoft, if asked, would secretly have to give money
59:00
that they have on servers in Ireland, and that is not compliant with GDPR. That's what the Max Schrempf's judgment said, so it's been written, but then the problem is that it needs to be recognized and decisions need to be made to stop doing it,
59:23
and it seems that it takes, and it seems like everybody's waiting for Europe to sign another agreement to say that it's okay so that you don't have to make the decision to stop hosting it on American systems.
59:40
Hi, thanks for the presentation. So you were saying that actually, like open source collaborative software would help solve the issue or improve the digital sovereignty. If we suppose that Microsoft tomorrow makes Office source code available
01:00:00
on GitHub, let's say, how would that change the actual state of the digital sovereignty in the EU? Well, it will change it if people start to take that software and host it and verify that it solves the business needs. But yes, it would improve it.
01:00:21
For example, Microsoft does not give the source code of Office, so if you want to do Office editing, you have to work with software that has been built by other people, not by Microsoft, and Microsoft and Google have better software, more compatible with Office formats and so on and so on. So it would improve, it still would need to be packaged, there would still be a lot
01:00:40
of work, but it would improve the offers, the alternative offers, and then companies. So about digital sovereignty, the thing is, if tomorrow we have a political problem, they can turn off the software, so all our companies will have problems. The question is, okay, do we have alternatives to replace that software at this point?
01:01:01
Not necessarily as good, and they are not as good because we're not buying these alternatives, so we cannot fund them, so the distance between the American solution and the European solution is increasing instead of being reduced. So it's all about whether or not we would be able to replace that software. If Microsoft made it open source, it would be easier to replace their own services, their
01:01:26
online services. Sorry, one big follow-up to the first question, basically, so US Cloud Act regarding
01:01:42
it, as far as I recall, Privacy Shield was the first one that has been taken down by your court, and I'm not sure if US Cloud Act somehow replaces it or not, because, no. No, privacy, the Cloud Act is an American law. The Cloud Act makes it a problem that makes Privacy Shield illegal, basically.
01:02:05
Privacy Shield said it's okay to put data in the US, and basically, the judgment from Max Schrems, go see his website, he explains the problem much more precisely, says that
01:02:21
this is not true, Privacy Shield says it's okay, but it's not okay, because of the Cloud Act and the FISA warrants, the laws in the US that allow access to data, even if it's hosted in Europe. Yeah, and European court accepted that Privacy Shield is not valid, it's down for a couple
01:02:41
of years, I believe. Yeah, now the European government is right now negotiating, the Commission is negotiating with Biden some new paper that says it would be okay, but it's already kind of validated that this paper is not okay, that it wouldn't, but there will need to be another judgment
01:03:02
to say that it's not okay. So it's a kind of a cat and mouse, because some people in Europe don't want it to be illegal to use American system, and Biden doesn't want to change the Cloud Act. Thank you.