Improving a Large FOSS Project Sustainability
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 |
| |
| Title of Series | ||
| Number of Parts | 38 | |
| Author | ||
| License | CC Attribution 3.0 Unported: 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 | 10.5446/66678 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
FOSS Backstage 202325 / 38
7
8
11
23
27
31
33
00:00
Source codeProjective planeOffice suiteSoftwareMusical ensembleComputer animationLecture/ConferenceMeeting/Interview
01:04
Suite (music)File formatStandard deviationProgramming paradigmHorizonOffice suiteProduct (business)Video gameSoftware industryMoment (mathematics)SoftwareSlide ruleNumberWeb pageCycle (graph theory)Software developerVideo gameProduct (business)Point cloudOffice suiteComputer animation
02:29
SoftwareOffice suiteMountain passSource codeSuite (music)GoogolSoftware suiteOpen sourcePoint cloudPlanningAttribute grammarScalabilityEnterprise architectureAreaOpen sourceSoftwarePurchasingProjective planeNumberMathematicsTerm (mathematics)Office suiteFreewareShared memoryTextsystemStrategy gameConnectivity (graph theory)Moment (mathematics)DiagramComputer animation
06:48
Domain nameInternetworkingSoftwareSelf-organizationBusiness modelSuite (music)Office suiteProduct (business)Open sourceScale (map)Digital signalWage labourOperations support systemSource codeBlogProjective planeCodeProduct (business)Category of beingMulti-core processorRevision controlScaling (geometry)SoftwareChainOpen sourceMathematical analysisCASE <Informatik>Chemical equationMoment (mathematics)Metropolitan area networkCountingBridging (networking)InternetworkingSoftware developerEntire functionOffice suiteSoftware bugEnterprise architectureBlogContext awarenessBusiness modelComputer animation
11:38
State of matterPointer (computer programming)Medical imagingPoint (geometry)Projective planeData managementDiagram
12:49
Term (mathematics)Business modelOperations support systemType theorySoftwareThumbnailOpen sourceEquals signDifferent (Kate Ryan album)Revision controlSoftwareFormal languageBasis <Mathematik>VideoconferencingUniqueness quantificationLine (geometry)Projective planeAlphabet (computer science)Term (mathematics)CodeLocal ringSource codeArmNumberMobile WebProcess (computing)Open sourcePoint cloudInstance (computer science)Associative propertyComputer fontTranslation (relic)CoprocessorMoment (mathematics)BitVisualization (computer graphics)WindowComputer animation
17:30
Strategy gameBusiness modelChemical equationGoodness of fitOpen sourceSoftwareProjective planeOpen setXMLUML
18:14
Office suiteOpen setControl flowRepresentation (politics)Visual systemContinuous functionOracleBitProjective planeMoment (mathematics)Special unitary groupOffice suiteComputer animationDiagram
18:46
Machine visionIndependence (probability theory)Object (grammar)Projective planeSoftware developerComputer animationDiagram
19:19
Core dumpRegular graphBitSoftware developerBasis <Mathematik>CodeCore dumpSoftwareNumberLevel (video gaming)Diagram
20:13
Core dumpNumberSource codeSoftware developerProduct (business)Enterprise architectureFocus (optics)UsabilityUser interfaceInternationalization and localizationIdeal (ethics)RootStrategy gameSoftwareService (economics)Computing platformSuite (music)Revision controlStress (mechanics)Reduction of orderConsistencyCodeFactorizationMobile WebPoint cloudModul <Datentyp>Process (computing)Enterprise architectureMixed realityPosition operatorSoftware developerNumberProduct (business)Focus (optics)Local ringRight angleFormal languageProjective planeUser interfaceComputer animation
21:46
Computing platformPoint cloudMobile WebSuite (music)Open setOffice suiteCore dumpODF <Format>Different (Kate Ryan album)outputOffice suiteAndroid (robot)Strategy gameComputing platformPoint cloudMobile WebComputer animation
23:03
Enterprise architectureProduct (business)PurchasingFreewareRevision controlMessage passingSoftware suiteNumberSoftware developerCASE <Informatik>Strategy gameLocal ringProduct (business)Self-organizationOnline helpGame theoryUniverse (mathematics)Term (mathematics)FreewareInstance (computer science)Enterprise architectureMultiplication signMereologyOcean currentComputer animation
25:28
Thermische ZustandsgleichungOpen setSource codeProjective planeOpen setSelf-organizationInstance (computer science)CASE <Informatik>Open sourcePlanningProduct (business)Core dumpAreaSign (mathematics)Universe (mathematics)Office suitePresentation of a groupStudent's t-testDifferent (Kate Ryan album)Multi-core processorSoftware developerCodeHypothesisOrder of magnitudeTerm (mathematics)Internet service providerState of matterTheory of relativityRevision controlEmailOnline helpLecture/ConferenceMeeting/InterviewComputer animation
Transcript: English(auto-generated)
00:03
OK, while it's loading, I'm one of the founders of the LibreOffice project.
00:25
So I'm one of the founders of the LibreOffice project. I'm Italian, as my first name says, although there are more Italo's in Brazil than in Italy, which is funny. But this is the reality.
00:41
I will try to explain what we are doing at LibreOffice to improve the sustainability of the project, because the project is rather large. Office Suite are one of the most widely available software.
01:01
It's basically on every desktop. There has been a move to the cloud, although there is still a lot on the desktop. And the majority of users at the moment are using two products and not just one as in the past.
01:21
In the past, it was clear that you had one software to produce all your documents. Today, you probably have one in the cloud and one on the desktop. Creating large documents on the cloud is not as easy or as possible as on the desktop,
01:44
especially if you talk about documents which are over 50 page long, which is a large number or a presentation of over 20 slides.
02:01
It's not always easy to manage them. The life cycle of Office Suite has been questions for years, but the reality is that they're still there. And two of the largest software companies
02:21
in the world, Microsoft and Google, are investing quite a lot of money on the development of these products. If we look at the market today, we are in the market of around $28 billion, billion not million, billion dollars.
02:42
And the reason it's easy to understand why companies are trying to protect their share of the market with not completely ethical
03:01
practices by using a lock-in strategy, which is based on different components, but mostly on the document format. The growth is stable. As you see, we had a decrease during the pandemics.
03:22
Although the reality is that the usage of the software increased, but of course purchases decreased, while now the growth is around 5% per year. If we look at the usage of the software, these data are from September 2017.
03:43
They have not been updated formally, but let's say that informally the numbers are more or less the same. So let's say that open source software in this area has around 15% to 17% of the market.
04:01
If you think that the market is $28 billion, it's easy to do the mathematics. The reality is that we have more users than the revenue associated with those users. The number of enterprises, especially enterprises,
04:23
that use free open source office suite without giving back anything are the majority. Unfortunately, they are the majority. We, as LibreOffice, at the moment, we get more or less the same amount of money
04:40
from individual users and from corporations. And you understand that this is completely unbalanced. In terms of the importance that corporations give to the software. They use the software strategically. There are governments that are using the software
05:00
strategically that are not paying a single cent to the project. We don't want, of course, the same amount of money of proprietary software. But to maintain the software, to keep it competitive, we would need more attention. Let's call it attention from enterprises
05:21
than we actually get. And this is, I'm showing you just because there are some funny answers from users. The Microsoft Office is the yellow and the orange.
05:43
And you can see they are up in the judgment apart from the cost. And what is funny is that they are first and second for customer support. Has anyone ever tried to get customer support from Microsoft, from about Office?
06:03
They are the first and second for reliability. The reliability of Office 365 is historical. It's never been Office 365 so far. It's always been Office 360, 359, because it was out for six, seven, eight days a year
06:25
about the reliability. They should call it for what it is. 365 means 365 days. So last year it was 359. So they should recall or at least correct the name
06:42
at the end of the year, because people are paying for it. But apart from the jokes, the LibreOffice project was born with a rather clear model. There is a foundation that is owning the,
07:04
not the property of the code, but it's coordinating the development. There are companies, there are individuals in the community, volunteers, and there are companies in the community, in the ecosystem. The development is done by, of course, developers.
07:22
Some of them are volunteers. Some of them are paid by companies. Companies, of course, serve those enterprises that are willing to support the LibreOffice either by paying developers, either by buying a long-term supported version of the software, either by solving bugs
07:43
or making other improvements or developing new features. TDF is a charity, is based in Germany, but could be based in any country of the world,
08:01
and as such cannot sell any kind of software. So the software provided by the Document Foundation will always be free. We try as much as we can to have the entire value chain based on open source licenses.
08:26
We don't have dual, at the moment, we don't have dual core solutions. Of course, the company that pay, and the ecosystem companies,
08:42
they pay for support development. They don't pay the code. They pay for the value added associated with the code, which are the stakeholders. Community members, they have a very high involvement, and they give a project a lot of value.
09:03
And users, they don't have a lot of involvement because the product is free, and they don't give the product a lot of value because Office Suite are a commodity. Today, if you buy a PC, you usually find a trial version of Office
09:23
which tells you that that product is a commodity in any case. And so one of the challenges is to create more involvement of the users in the project.
09:43
The problem of open source sustainability, especially for large project, but some of the issues we had in the last few years are not always related to large projects.
10:01
We have always had some solutions or some proposed solutions. In 2014, the Art Bleed bug was probably the episode that turned the, made people aware
10:24
that you have to invest in open source, especially if it's a strategic software. In 2016, Nadia Ekbal is a researcher, published a very long paper, Roads and Bridges.
10:40
It's available on the internet. It was sponsored by the Ford Foundation. If you have not heard about it, download it and read it. It's quite a long paper, but it's very interesting because it's a thorough analysis of the situation.
11:01
In 2019, Dries Buitert, who is the man behind Drupal and Acuya, published a very interesting, a very long, and I think very much thought about
11:21
blog post is balancing makers and takers to scale and sustain open source. I will use some of Dries' comments because I think they are extremely interesting. So the, basically the first stage is that
11:44
the roads, this is a summary and is an image that summarizes the idea. If you take roads, the roads are created by volunteers. So you go from point A to point B and you create the road.
12:02
Of course, if there is a meaning, the road is improved by the business because there is money behind it. If you think in the middle age in Europe, to go from point A to point B, you had to pay taxes. And that was used to maintain the roads
12:22
to go from point A to point B. And then they become a commodity for the community and the state owns the road. So the community owns the road through the state because they are an advantage for everyone. This is one idea of how the project can be managed.
12:49
But the problem is that, of course, open source communities are not always willing to stick to a strict code of behavior
13:05
to make the software easier to sustain. Volunteers want to develop what they want to develop, not what it makes sense to be developed or makes economical sense to be developed.
13:25
And of course, this is what makes open source unique on one side, but makes also a challenge maintaining the sustainability of projects.
13:40
So, Bing Libroff is a large project. The code is seven million lines of code. And the number of users that we have is estimated between 100 and 200 million. We have around one million download per week.
14:03
And of course, this makes the project quite large, not easy to maintain. At the moment, we provide versions for Windows, for Intel and ARM processors, for Macintosh, for Intel and Apple processors, for Linux.
14:22
We have versions for mobiles and for the cloud from ecosystem companies. So the number of versions that have to be released is quite significant. We provide a new release almost every month
14:41
and a major release every six months. So the challenge of maintaining the project is not trivial. And the last idea of DRIS, which is what I think is extremely interesting, is the dividing users between or people
15:04
associated to open source between makers and takers. So makers are the people that actually make the software, develop the software, but not only develop the software. For instance, for LibreOffice, you can imagine how big is the localization effort.
15:24
LibreOffice is available in 120 languages. Is the software available in most languages in the world? More than any other software. There are other 30 languages in process.
15:42
So we have around 5,500 localizations. Localizers on a global basis. And of course it's not easy. Languages have different needs. In Asia, you have a lot of languages
16:04
which are written right to left and top to down, which is completely different from our languages that are written left to right and one line after the other.
16:22
And you can imagine what alphabets are making in term of efforts because not only you have to guarantee the translations, but also the compatibility of the software with the different fonts,
16:41
which are shown on the video with different styles and with different needs, each one. So what Dries says is that there is a large number of makers, but there is unfortunately a larger number of takers.
17:03
Makers are also companies, companies that invest on open source. Takers are company that take open source, maybe do a little bit, but really a tiny bit, and take most of the open source code. So basically, at the end,
17:22
this means that takers can damage makers and this visual will tell you why. If a company invest 50% of the revenues in improving the software, so it's a good maker of open source,
17:41
of course it will risk its business model in front of a taker that will invest only 5% of its revenues on open source. So these guys will have more opportunities of advertising or investing money in commercial strategies than the makers.
18:01
But on the other side, the makers are the people that are really making open source. So we have to find a balance and the LibreOffice project has the same issue. When we launched the project, we wanted to relaunch, this happened 13 years ago in 2010.
18:23
We wanted to relaunch the innovation. It was OpenOffice that was a little bit stalling because of the size of the project on one side,
18:40
but because also of the issues that sun the company behind the project at the moment. So we decided to create a foundation. A foundation, we created a foundation because there was not another foundation that could serve our objectives. We didn't want to create a foundation
19:02
for the sake of creating one because it was a lot of work behind it. We wanted to have a project that, where the community was volunteers and ecosystems. We wanted to have a development that is more or less balanced.
19:23
These are the last two years. 62% are from the ecosystem companies, around 30% from volunteers and TDF is increasing. TDF is the Document Foundation is increasing a little bit the development.
19:41
If you look at contribution, core developers account for 75% of all the code. It is important to have core developers. It is important to keep the number of core developers at the same level. So ecosystem companies are key in this effort.
20:02
But also some regular developers are members of ecosystem companies or are volunteers that are working on the software almost on a daily basis. So basically 62% of development is paid by ecosystem companies customers.
20:22
These are the good friends of, the enterprises that are good friends of our project. And the rest is done by volunteers or by the team.
20:40
And the problem is this gives you an idea of where focus is. Enterprise ecosystem companies is new features mostly. Volunteers are mostly user interface and localization.
21:04
And TDF developers are mostly doing accessibility, right to left languages and all feature requests. Features that over 10 years have never been developed by anyone. So today we are still in a good position,
21:24
but the reality is that in the next 10 years we should really increase the number of enterprises that are paying to develop LibreOffice. The concept that we have developed is that LibreOffice has evolved
21:42
from a product to technology. The reality and this shows you better, the visual as usual shows you better the difference between LibreOffice and the other office suites. So LibreOffice has the same engine independently from desktop, mobile or cloud.
22:05
Because the engine has been developed and tweaked to be flexible. All the other office suite, you can put any name on that. Microsoft Office, only Office, WPS Office, whatever Office,
22:22
they've all developed one engine for each platform. So the engine of Office on the desktop is not the engine of Office 365 and it's not the engine of Office for Android or iOS. This means that the documents will be different
22:42
the document are actually different apart from the differences that are, the hidden difference that are added to the XML to increase the locking strategy. But they are different because the engines are different.
23:04
So we want to stress in these advantage, we want to increase the number of people that are investing on LibreOffice. The current users as you see in the community are individuals who donate and some universities,
23:23
some enterprises, some government organization that are paying for LibreOffice. In the future, we should, we cannot increase the individuals. I mean, we would like to, but of course individuals are already making a lot of effort to fund LibreOffice.
23:41
So we will accept some as free riders, universities, enterprises, a few of them is part of the game. We don't want each company to pay 100% of the products. They can pay 60%, 50%, but they have to pay something. So we have to increase, we have to educate.
24:02
In the case of the Italian Ministry of Defense, for instance, we managed to have them pay for 40% of the 100,000 desktops they're using LibreOffice on. And that is okay. These are the 40% of strategic desktops.
24:24
The other 60% is used once a week. The other one are used once a day. So they have bought 40,000 licenses. This is okay. If all the large corporation that are using LibreOffice would pay 40%, we could really invest,
24:43
probably triple the investment on LibreOffice in term of development. And communicate with others to stress the importance of having them pay some money or paying developers
25:03
if they are developers that work on LibreOffice that's absolutely okay for us. We don't want to get the money. We need to have development or localization or quality assurance help.
25:21
And time for question if there is, otherwise, thank you for listening. Thanks a lot, Italo. Yeah, other questions? Thanks for the very insightful presentation.
25:48
So you say that you want to increase the organization's paying. Do you have any idea how you're gonna do that? Actually,
26:00
I think that the only chance we have to do is through education. Because otherwise, one possibility, and this is why I stress this at first, sorry, one solution could be ever dual core
26:22
or an open core product. But that is not open source. So we would like to resist that as much as possible. Because we want to stay loyal to open source. So we, the first idea is educate.
26:41
If we manage, okay. If we don't manage, there will be some organization that have to go open core. There's no other way. I mean, we have a bank in the state that is deploying LibreOffice on 300,000 desktop and it's not paying anything.
27:02
That is, I mean, we talk with them and we say, oh, we don't have budget now. You don't have budget? Come on, a bank. And the money that you have saved by deploying LibreOffice instead of Microsoft Office,
27:22
let's consider 10% of that money. In some cases, it's really difficult to really, to forgive this organization. I can understand universities. Universities, according to the governments,
27:42
they may have more or less money. There are universities that have really not that much money even around Europe, okay. But for instance, if a university starts working in the engineering area on getting students
28:05
that elaborate their thesis on LibreOffice, that would represent an help for LibreOffice because a thesis would either solve some bugs or improve some code, provide something. So we would like to stay really loyal to open source as much or as long as possible.
28:24
Then we'll see. It would be a real pity, I think, for everyone if LibreOffice was stopped, development was stopped because of the economics. That would be a real pity for, I think, for me, for sure, because I use only LibreOffice,
28:42
but for many people around would be as well, at least for all Linux users. Say, oh, okay. What's the first? Okay.
29:04
Yeah, thank you for the presentation. I have one question, one idea. Do you think it could be possible, it depends on the relations you have with them, but to go through the companies that are providing LibreOffice, to educate the organizations.
29:21
So to say, okay, we have some companies that are partners of the project, of the Document Foundation, and they can, you could sign an agreement and say, okay, they have also to educate the organization and to say, well, 1% of what I'm going to build for this project is going
29:41
to the Document Foundation, for example. But we are already doing this because we work with ecosystem companies. Of course, the ecosystem companies have a mixed business plan. They have to pay their employees first. But usually, by staying loyal to open source,
30:04
they release with an open source license. So let's say worst case, a feature that is developed by an ecosystem companies after six months is only LibreOffice community, which means it's available for everyone.
30:20
So we are already working. The problem, of course, that our resources, human resources, are limited. We are not thousands of people going around and explaining this. And therefore, the community can help us. I mean, if you have a chance of talking to someone
30:41
and you understand and know that they are using LibreOffice on many desktops, try to tell them, at least send an email and say, what can we do together with the project?
31:01
She's making a terrible sign there. No, this is better not. This is okay, but. Just only for clarification, what's the price of LibreOffice?
31:21
When you talk about like, they are deploying it on 300,000 desktops, so it's like the free version? Yes. Of it, so you are not putting a price tag on it or something? No, no. So what would you expect from them to pay you? Because I think it's very.
31:41
I mean. Go to every company and say like, hey, this is the price for you, so do you have any? No, first, as foundation, we cannot, because we are a charity, so we cannot ask a price for anything. And second, this is a question for the ecosystem companies.
32:02
They have a price for LibreOffice. I can tell you, I'm not allowed to, of course, to make any advertising for these companies, but I can tell you that is around 1 10th of what they would spend for Microsoft Office.
32:23
The ratio is one order of magnitude less. So you can imagine the difference in term of investment from a company.