 Thank you Siren and again, thank you very much Diedmar for giving an insight into your paper but also for the applause to our study. We were happy to do together with colleagues at the Open Forum, Europe and some other co-authors. I was asked to talk a little bit more detail about how we really calculate the economic impact and I go a little bit into into depth here. First of all, that was the overall approach. We looked at the literature also on open innovation as maybe the bigger term and we also collected data and then we did different things and I will talk about just the economic impact assessment. There were also case studies by Mikko Böhm and Rukatz and we conducted also a stakeholder survey to validate the assessment we derived from more of the econometric exercises and also here Siren and Paula did the analysis of the policy policy initiative and that all went into the recommendations. I'm not going to talk about but very nicely to see that also Frank Nager to which I also really send my big thing for inspiring us both regarding how to deal with data. I will come back to that and also the methodological approach because he just published a paper where he compared our recommendation with what the US should do in a Brookings Institute paper therefore for those who are interested in that but I'm not talking about it. Okay that's the overall approach where we kind of had different types of validations. We looked at the benefit assessment and here I will just focus on the macroeconomic analysis. We also looked or asked here stakeholders to kind of try to assess the different benefit dimensions of open source and then I'm sure one has also to compare with the cost side. Here we looked at the macro level on the investments in the EU into open source and we also looked at individual contributors, the top contributors in Europe and then we validated these kind of macro and micro approaches and the cost and the benefits and my focus today will be on the left hand side where we just compare the cost on the micro level and the benefits on the micro level a little more in detail. The data sources here we follow a previous paper by Frank Nagel who looked at the impact of public procurement law changes in France on their innovation performance and he used the data from GitHub and therefore we follow this means we look both at the comments we also looked at the users and their organizational affiliations and tried as far as the data allowed that to attribute that to the different EU member states. We know that's not complete that means overall we are probably underestimating the contributions by factor two or even by three but at least by two and then we combined it with other economic data from the OCDO we started also looking at the European patent office data just to get an indication of Molochik and we also looked at at the crunch based database of startups because some other paper by Boy White and others also by Frank Nagel they also looked at that and we also kind of leveraged this insights to the EU and therefore this number already showed of having 600 additional ICT startups per year in the EU based on open source contributions is another insight which I am not really going into detail today. Okay this is the development of the comments between 2008 and 2018 that was the last year we could look at and overall we see this is going up in all our member states there are some special kind of issues here for Greece some artificial changes but even if we take this out the robust are quite robust. New figures I have seen in 2019 there was a data problem because we took the data from GT Torrents from Delft University they had some data problem in 2019 but looking at 2020 we have still an upward trend in contrast to some other two recent papers which have been published one and plus one but based on a quite questionable methodology. These are the comments by the EU member states and here also the contributors also per country and we also see it's going up there was some slowdown 2018 but overall this is again taking up in 2020 again therefore still increasing contributions although for example recently Bitcoin the German industry situation they launched the open source monitor in 2019 it was also kind of co-sponsor on that and one big problem is certainly that the skilled shortage in this area and this is going to become more well and therefore maybe later in the policy discussion we can take this topic up. That means that's the data and what we did where we looked then okay if we really then try to calculate what are the cost of these contributions we looked both at the member states level and then also we looked at the efforts by the most active companies located in the EU and as I already mentioned this is really lower bound an assessment because not all of the contributors really disclose where they are located. Overall the basic assumption is that these contributions are going into the public domain this is this innovation comments what Dietmar presented before and that these investments will outweigh the cost but due to knowledge below us the benefits are much higher now but if we if we really take these numbers and we put them a little bit in context we have around three million employees in the computer programming sectors in the EU and we found in 2018 about almost 10 percent are contributors to GitHub and probably we are it's probably more 15 to 20 percent and if we kind of use then some methodology to calculate the labor cost we get up if they would really full time work on that we would see labor cost on investment of 40 billion if we take the comments instead and and use the so-called constructive cost model we get to 16 000 full-time equivalents and this would be then an investment of one billion labor cost or personal cost in the EU in 2018 that's that's the cost side. Then looking a little bit more into the structure who contributes we find that here these these companies which are most active these are counting for 12 percent of the contributors and one third of the comments and overall they employ more than one million employees these companies but if you're looking in detail these are in general very small companies in micro companies that means 75 percent of these top contributors have less than 100 employees and we also see that the smaller the companies the more contributors are listed and the more comments they provide it means the contributions are done by 50 percent of companies which have left than 50 employees and on the other hand especially the the small companies here five percent of their full-time equivalents are involved in contributing to open source that means that's a that's a little bit different structure than we find for the US but especially really the large players the big tags are making really more major contributions to open source. However if we really put these figures together with the macro figures we get a quite a validated picture on on the cost side. Okay the cost side is always easier than than the the benefit side and here we rely on the traditional macro economic model the so-called but called Douglas production function. Frank Nager also used that for his work on on US companies on the company level but we can also do it and on the aggregate level and the the idea is okay if you're looking at the GDP at the output of a country it's it's more or less the the product of capital labor plus some knowledge stock and the knowledge stock we use R&D research and development plus some previous stock and here we take the patent data and and overall this changes that means this the logarithm of these values and here a special thanks to my colleague who did the chronometrics is based on R&D on previous on the patent stock as a knowledge stock but then on capital labor and the contributions to open source also the contribution of their own country but also the contribution by the rest of the world that was the overall approach and what we did then is we we kind of modeled something just for the you taking the you together and looking at the total knowledge pool of the you as as the base and calculated then the so-called elasticity the elasticity is the the the measurement of the percentage change of one variable in response to another one that means how the GDP in the you member states changes according to the contributions to open source and what we find are two different types of elasticity one here we use the comets as as an input and the measurement for open source and the other one is is using the contributors and the elasticity is 0.04 and 0.06 what does it mean this means that and and we we have seen these 10% increases in the in the comets in the last year especially from 2017 to 80 that means then if you have an elasticity of 0.04 and this means that the contributions to Github or to open source are contributing 0.4% of GDP in the you and if we take these numbers then we get to about 60 billion contribution of open source in 2018 if we take the number of contributors as an indicator then this 10% increase in the number of contributors would push the GDP in the you by 0.6% and then we would have even a total contribution in 2018 up to 95 billion I mean overall this is really a significant contribution from the global pool of open source into the you GDP and then for the future we could even assume that maybe we reach 100 billion we use this approach already 20 years ago to address or and calculate the impact of standardization to the German economy and and by then 20 years ago we found a contribution of 15 billion to the to the German economy and therefore the numbers are are not really too different and in the last step we we put the different numbers together and at first think you say okay these are very high cost-benefit ratio but we have to to assume that you rely on on the the open source contributes also in previous years there are some figures about the lifetime or the halftime of open source coded projects and we assume the kind of linear relationship of 10% depreciation and therefore we have to multiply the the contribution 2018 by 5 and then we get to a cost-benefit ratio of 12 to 1 and in a in a last step we also consider the hardware cost and then overall we we are quite conservative in our calculations we get the ratio of one to four and this ratio of one to four is is quite close to a ratio which has been recently published by Jones and Summers where they they calculated the impact of R&D spending on GDP by by really kind of doing a meta-study comparing different different approaches and they get also to cost-benefit ratio of if you spend one one you or one dollar into into R&D the the economy kind of benefits by five dollars or or yours and that's depending on some scenarios it goes in up to one to ten but overall this is this is our approach and and this is quite consistent with what has been done before then we also derived some power policy recommendations I'm not going into it maybe we can talk about them later these are the the overall kind of suggestions looking at institutional capital capacity in the in the public sector but especially it's about knowledge creation and diffusion because here the economic impact is is strongest it's also about entrepreneur aspects regulatory issues and so on but overall and to to come to the end and now we see there's a really a large economic impact of open source for the EU and we see also a potential impact of the open emerging open source hardware but one has really to to push this and this is also in line with the Diet Maher was kind of innovation policy recommendation from their paper that we have to incentivize the contributions in order really to reap and exploit the benefits for before the European economy and we have also to to be aware that different policy areas have to be coordinated in a quite kind of complex way you know in order really to be able to exploit these these benefits thank you and thanks again also to the coordinated co-authors and the contributors also to the stakeholders surveys and the case studies great thank you Knud very much then I think Johannes we can continue with your presentation we're gonna make you present also in a second hi everyone can you hear me yes we can hear you okay then let me also share my screen and get started right away just one moment so can you see my screen yes I assume so yes it's coming up yes we can see your screen let's just hope right this whole screen also works yes it works fantastic great thank you very much thank you very much Simon thank you to the organizers for this wonderful opportunity to speak at this this event thank you also to Knud and Diet Maher for a very interesting talk so far I want to present some recent work today that's currently under revision hopefully there will be a publication out of this soon no working papers available but I'm interested in the geography of open source software in particular and I think there are some nice complements between what what we found in our study and and this this wonderful report that forms the basis of this event just to introduce myself I'm Johannes Wachs I'm an assistant professor at the Vienna University of Economics and Business I'm also a faculty member at the complexity science also in Vienna and let's get to it so as Knud just mentioned there is this emerging body of literature which has been greatly strengthened by this report that tries to measure and I think does quite successfully the effect of open source software on the economy just to quote from the report I think both of the previous speakers have had this on the slide before but I think it it certainly merits repeating that an increase of 10% in contributions would generate between 0.4% to 0.6% additional EU GDP per year and more than a 600 additional ICT startups per year in the EU so there are some mechanisms discussed in the report of course this is not all of them but I think they it helps to drill down into the mechanism and this motivates why I'm interested in geography so much so some of the mechanisms for these costs for this these gains in productivity include the idea that open source software is a public good so we can reuse code and build on the work of others this is links to the idea of the open source software as a kind of commons that we can all benefit from using open source software saves money so especially in the public administration this is a great way to save some costs to cut costs and open source software is open and transparent and this leads to this idea that that it may have a higher standards and quality than proprietary or closed source software this is kind of referencing the the old idea that with many eyes all bugs are shallow but this leads left me asking a question does it matter where open source software is created so none of these mechanisms for for why open source software may benefit the economy say anything about where this is being made beyond at maybe a country level and I think that geography actually matters a lot more than we might think or hope so so to quote some geographers despite the digital age and the digital economy we all live in the notion of distance is certainly not dead so there are some papers that show that the likelihood of collaboration on github decays exponentially with distance this is both of these papers fit a kind of gravity equation to the likelihood of collaboration as a function of distance and in 2010 and both 2020 the same result came out that we are much more likely to collaborate with people on github if they live close to us of course many many reasons why we might expect that but it is certainly an sort of empirical fact on the other side we also may think that many of the benefits of open source software so the economic benefits might accrue or build up locally for several reasons for several mechanisms and here I also build a lot on the work of Frank Nagel first we know that firms learn and gain feedback by contributing to open source software firms using open source software also tend to become more productive and finally this is a kind of information story firms workers and investors use the information revealed by open source software contributions to make better choices so let me say a little bit more about that last point so firms use open source software contributions to decide who to hire hiring software developers is a very difficult problem for in the software industry it's very hard to to predict a person's ability at software and how good of an employee they will be developing software and open source software contributions give a strong and credible signal of ability that smoothens out this this rather difficult labor market developers or workers as I call them in this in this slide also benefit from this they may see a firm that has made open source software contributions and they may benefit from this information they may they may better be able to choose the firms they like to work with and for example they also might enjoy the fact that they'll have if they join a firm that contributes to open source software they can contribute to open source software and keep this kind of visibility that is not unfortunately not so not so easy to do when you're working at a company that uses only closed source finally investors use information revealed by open source software contributions to make investment decisions so if you're a venture capitalist if you're an investor it can be quite hard to evaluate the quality of a firm the quality of technology but if you can look into at least part of their court base or see what kind of contributions they're making to open source you might have an additional information edge and be able to make a better decision so these are ways in which benefits of open source software and their impact on the economy actually happen very locally so of course an investor can invest money across a great distance but the benefits accrue to specific firms people working in specific areas of course we're in the age of the explosion of remote work but still most people do work in the same city as the as their employing firm I also want to reference a very recent this is a very fresh paper from just this is a pre-print posted just over a week ago by Chattergun and Curve that shows how important technology and software is in the broader world of innovation so in the diagram on in the chart on the left on the x-axis you have years been into five year five year bins on the y-axis you have the share of all patents made so this is the United States is the universe and we have four lines each representing a different group of cities what we see on the left is that six tech cluster cities so these are San Francisco Boston Seattle San Diego Denver and Austin contributed something like 10 percent of patents in the late 70s by now one in three patents made in the US is filed in one of these cities one of these six cities at the same time the largest cities so New York Chicago Los Angeles and two others have rather declined and the chart on the right breaks this down into software and non-software patents in the US you can patent software and we see that most of this difference most of this changes due to to work in the software activity activity in the software field pardon me so really this this the signals that that geography does matter so that software and open-source software development may be taking place in very specific areas so we investigated these these questions and well here they are let me go through them first does open-source software activity actually cluster significantly in space does it cluster more than other kinds of high tech or innovation intensive activities if so where are these hotspots third can we explain the ingredients needed for places to promote and attract open source software development and developers and then can we translate these into policy ideas so with that view we we set out to do the following well we wanted to measure geography the geography of open-source software developers in 2021 so we did the geocoding in the late winter around march and this is how we did it so we this is joint work by the way with my co-authors marioche william and axel we built a pipeline to generate geographic data on open-source software developers we looked up what we call active software developers on github using the gh archive database this is a bit different than the gh torrent database for the technical details see the see the pre-print and we set the inclusion criteria for developers at 100 commits across 2019-2020 so while we don't have 32 million while there are 32 million github accounts according to this activity threshold of 100 commits over two years we are left with about 1.1 million developers and then we geolocated them using being maps api and some heuristics applied to user provided locations on github twitter and commit email suffixes so here's an example developer x on her github profile she mentions that she lives in austria on her linked twitter profile she mentions that she lives in veen which is the german word for vienna and she frequently makes commits from an email address at wu.ac.at that's the email suffix of my university and from this we use some heuristics to infer that she developer x actually lives in vienna and this pipeline lets us geolocate over half a million active open-source software developers around the world before I get into the regional analysis let me let me say one or two words about what we found at the country level first we found that country shares of all active open-source software contributors on github in 2021 compare in an interesting way to previous snapshots so we took two snapshots from the literature one created in 2008 using data from sourceforge and one created in 2010 also using data data from github what we observe is that between countries since 2010 we have a much more even distribution around the world of the share of open-source software developers so while the united states alone accounted for more than one third of all open-source software developers just 10 years ago now it's down to about a fourth and which are the countries that are that are gaining these are the biggest gainers are china india brazil russia japan south korea so southeast aja latin america is also doing doing much much better relatively so in some sense we can say that open-source software development is spreading out through the world which is a i think a great thing and i also want to note that this is not just the story about rich countries versus poor countries so here on the x-axis i plot um so these are these are all countries on the x-axis i plot the country's gni per capita it's a logarithmic scale on the y-axis i plot the number of github contributors uh per million inhabitants also on a logarithmic scale and though we see a clear and significant correlation we all we can only explain around 40 of variants using economics using economic development um and we also see some interesting outliers so who do we have here above the regression line we have ukraine brazil uh syria belarus bogaria also let me point out estonia so these are countries that are doing much better than expected given their level of economic development and then we have some unfortunate ligards for rain syria arabia cuates these are the gulf oil states um for given how rich they are there are relatively few github contributors living in those countries so um what we can do is we can explain up to 75 percent of the variance between countries using just uh two two interesting features economic development so gni per capita and the un's human development indicator as features in a regression model so what does this mean it means well at the country level uh it's relatively easy to predict how much open source software activity happens in a country using just a few features so if you give me an imaginary country with an economic development level and a human development indicator i can predict roughly speaking how many open source software developers there will be of course there will still be deviations there will still be residuals but at the country level it seems that this is a this can many of the things can be explained with structural features so let's zoom in to the regional level that's the point of my uh my talk um so let's look at this is a uh the report is about the eu about europe um we also looked at european regions we also included uh the uk norway switzerland uh and some other uh non non-eu countries but certainly european countries and what we see when we plot this map is that we see immediately intense regional concentration at the nut stew level the winning region uh it may not surprise you is berlin and these are per capita numbers berlin has 175 active open source software contributors per 100 000 inhabitants that's a quite impressive number and the other uh the other regions and these are not stew regions um with greater than 100 contributors per 100 000 inhabitants are london syrich oslo prags dotcom and amsterdam um so one thing we can observe from this map uh is that in some countries it seems that that there there's an intense concentration within specific parts of countries and by the way these colors are pinned in a sort of logarithmic way these are called jens caspal bins so the there is indeed a very large difference between the brighter shades and the the darkest ones so let's compare metropolitan france with the check republic the differences between regions seem larger in the latter that is in the check republic and the question is can we can we measure this of course we can look at the map and we can squint and we can say well uh open source activity is much more concentrated in the check republic in france but can we actually quantify this so we can make comparisons so what we do what we did is we adopted a measure of geographic concentration uh from the oecd called adjusted geographic concentration or agc for short it ranges from zero to one it takes the value of zero if open source software developers are spread exactly uh according to the population so this measure also takes into account that different regions have different uh populations different densities and it goes to one if all open source software contributors are concentrated in the least populated region in the country so basically what we see is that france has a low agc relatively low agc score of 0.28 indicating that the open source software developers are spread out a bit more in france versus in the check republic they're quite concentrated and we can do this for all the countries in our data set by the way we don't just have data on regions from european countries we also have the united states japan china russia brazil india and we can also calculate this agc score for different kinds of individuals and their distribution through a country through the regions so for university educated individuals or individuals working in high tech fields i take these data from euro stat by the way what we see is that the concentration of open source software developers exceeds the concentration of uh either university educated people or individuals working in high tech fields and we see this the dark blue uh points are with one exception this is greece um are significantly above the concentration measures for high tech workforce and edu and university educated people this goes to say that open source software activity seems to be extremely concentrated within regions and this is not just the european story so these are the top 10 us metropolitan statistical areas in terms of contributors per capita these are msa's with at least a million people um and actually we can use our data to estimate that 34 percent of all us-based open source software developers are living in the six tech hubs of chatter good and car that was that study i mentioned a few slides ago uh and just to remind you that these six tech hubs account for a third of all us patents and 45 percent of software patents so the so this regional clustering is not just the not just the european story can we explain uh can we explain using structural factors this concentration so recall that at the national level we can explain 75 percent of variants in open source software activity between countries using human and economic development indicators use in a regression framework we can only explain 50 percent of the variance between nuts two regions with a similar approach uh the conclusion from this is that local open source software activity is much more idiosyncratic nevertheless we found some interesting uh features which are highly correlated with nuts two open source software activity namely tertiary education attainment so university graduates living in the place uh employment and high tech industries and interestingly um there's this nice measure of uh a general social trust measured by the european value survey this is available for many european nuts two regions this connects nicely with the story that contributions to open source software are a kind of uh public good provision and certainly there's a rich literature on the link between social trust and contributing to public goods i won't dive into that uh time is tight so let's revisit our questions first does open source software activity actually cluster significantly in space and the answer here is a resounding yes what are the hotspots well in europe you might not be surprised but these are london berlin prog etc um can we explain the ingredients needed for a place to promote and attract open source software development and developers it seems not very well it's rather an idiosyncratic thing at the regional level and then the open questions can we translate these into policy ideas um well regional and and and city level policy is quite different from national policy and that's why i think it's it's worth really thinking about this a bit harder so even though regions and cities generally don't have as much legislative regulatory power as as uh countries or national parliaments they have the advantages that they can often be more flexible so while they may not be able to to to uh tune the tax levers quite as quite as powerfully as the national governments they can focus uh a bit more on what's needed locally and we can also draw on a rich literature on what's called cluster policy this is the idea behind cluster policy is how can we encourage agglomerations of specific kinds of activities so this is usually applied to to innovation intensive uh industries and sectors some ideas coming from this are that we should try to foster informal networks and this is actually been a key to silicon valleys flourishing this is already uh uh there's a great paper by sexanian in in 1990 from 1990 that documents the informal networks in silicon valley in the 80s that kind of kept it going through a bit of a rough patch in hard times um a second idea is to give people opportunities to meet in person encourage mentoring relationships so mentoring is one of the the has been shown to be one of the most effective ways to get people into open source a third idea which would be effective at the regional or city level is to advise firms on the benefits of open source software perhaps they could cite these nice papers from uh frank nagle and another another arm here is to involve local universities and this goes into the direction of helix models of innovation and development so i can't say everything i want to about policy indeed this this is the the end of my presentation i just want to mention that the data and code are available on github and we have a preprint out posted to the archive i also want to thank briefs in my co-authors uh who you can see here and if you have any questions of course we're happy i'm happy to have a chat now but feel free to reach out to me directly after the conference or workshop thank you thank you johannes and also thanks to knud um well i'm inviting you both to come back so well johannes to stay in essence and knud to come back um let's see if we have a few questions i think some positive feedback that i'm very happy about um uh if some people are still gathering the the confidence to ask questions um one of the things i guess i was wondering a little bit about is um how uh kud if you see if some of the um let's say some of the um more specific more local ways of analyzing contributions could also be used to make an analysis on the EU level um to kind of um add to the analysis of on the EU level kind of give it maybe a little bit more um a local aspect first of all thanks johannes great work which is nice kind of nice complimenting what what we did on the on the accurate level um uh and and i was not even aware that this could have been also an opportunity for us to to dig down but but you you have done the work and therefore therefore i think um this is this is really done pushing the um also the the knowledge kind of body on on open source uh further we have the macro level and now we have the regional level but we have the the company studies by frank um in us i mean it goes down on the micro level um and therefore i think we have now already a good starting point and uh what what i'd like to to do is is a little um looking at your your policy um ideas and and indeed um and we did that i mean we trust finishing a study for the the german government on uh how to kind of optimize their um kind of uh entrepreneurship policies startup programs and and so far uh we we we we visited kind of what's there and there's nothing which is pointing to the role of open source and uh and uh thanks to what frank did but also now related to your work i think there's there's empirical evidence that um these these hubs you identified are also startup hot spots uh berlin is in in in germany they i think they get more than half of the funding um private and public that means uh and and i know also startups in the field and and indeed they they rely on on on open source and and but this this is not yet in the at least at the national level that's not yet reflected and we at least included one recommendation specific recommendations towards considering the open source aspect i think that's that's initially another thing uh and and this is other work um we did some work about berlin because berlin is the only city in the world which has its own innovation panel um and together with colleagues from tv we we looked at the innovation activities around the the research institutes here in berlin and what we see distance matters um but if if you are looking at the and universities have the third mission the transfer uh but here again um uh we we as as uh technical university of berlin we published a transfer strategy where we kind of expanded the the channels uh towards first the initiation but also open source contributions to open source but um so far in the other um documents i i couldn't find that universities are really taking this transfer channel really actively on their on their list i think that's that's an another point to uh to really at least communicate with the universities okay if do you think about it because i also know from colleagues here to belin that this plays plays an important role um and and even for regional policies um uh cluster policies i think uh open source should at least get into the the the mainstream thinking it's not only about maybe doing co-publicizations or co-parenting we we push some uh some ideas towards maybe doing collaborative stabilization activities but also collaborative open source activities should be at least maybe eligible for funding in such cluster policies and uh and since since open source is also entering new new areas like like biotech it's not not not only the traditional IT sectors but but it's really kind of entering more or less all other areas of of the economy and therefore this is this is uh kind of what you did together what we did together i think this is also uh a lot of food for thought then also for for game changing or at least uh expanding um considerations at the policy sides regarding uh taking this this great opportunity on board because on the other thing would we would we see more and more and the this bit com survey really kind of uh we really set and confirms that that the skills shortage is is getting is getting uh especially in the IT area a major challenge especially for Europe yeah i agree wholeheartedly um maybe maybe something i could say about the universities uh that we observed in our data is that um so besides the the traditional tech hubs which again i think aren't we're not so surprising some surprising regions that did very well uh are um uh regions without very large cities but with uh large and famous technical universities so in norway around trontheim where the large norwegian technology university is located in germany around aachen where where a rwth aachen which is a very large and successful german technical universities located and in kalzwood these were these were kind of hot spots where you uh of course reflecting on the fact that these are locations hosting these technical universities perhaps not totally surprising but at first glance surprising that kalzwood is is is doing so well so i think the universities play a really key role i think that um vienna is a very good example of a place where a city uh doesn't wait vienna by the way is its own region is a has a quite independent government um vienna is a place where they don't wait for the national uh national government to to do the right policy to take steps towards innovation to take it in their own hands and they also work very very closely with with universities there's a actually a vienna level funding agency that that distributes quite a substantial amount of funding each year so uh yeah wholeheartedly agree it's uh it's quite interesting because simon simon first just makes a point and there's a little discussion now happening in the chat but uh because simon simon says i'd also expect redhead locations to have an undue magnetism as the developer sort of convert to being company people and of course um you showed the Czech republic before and you showed of course there was Prague and then there was the region that bono is located in which i know has a big redhead uh office uh well that simon brilliant yeah same thinking um and they surely have also some impact there may be also university there i'm wondering if you have any thoughts on um the discussion regarding uh also as contributing um you know well the discussion is happening essentially on um ability of developers and job mobility here yeah so this is the classic the classic chicken or egg question so are the developers uh going to a place um because there are other developers and they like it there or are um people going to a place and then or living in a place and then they they're more likely to become developers because many people are developing around them and i mean open source software developers i think i think it's a mix of both i think there are feedback loops i don't think it's just that you know people are equally likely to get into open source everywhere and then they make the decision to to join uh the tech hubs like berlin and and prog um i think it's a mix of both i think we need to do a better job measuring it and i i take the point completely from uh simon and and jan that um we need to we need to measure this to make the right policy decisions i think a good a good null hypothesis is 50 50 but um it seems the the chat thinks that it's um maybe 80 20 in favor of mobility i'm happy to see a study that that that looks into this that's uh that's a future research area i'll make a note of it um i'm also seeing um mark who has a question for knut and read it now um yeah but i mean i'm sure you can react to it uh maybe i can i can briefly respond to mark yeah indeed um it's a it's a simplified approach to look at the contributions and and it would be much more interesting to see what was the value added the contributions generally but um this is uh this is on the macro level this is this is not a feasible approach you can do that maybe uh on the micro or on a case-based level to see a kind of what what is really the value added but but in these macro models uh this is this is not not not an approach and and also that's also an issue the the diffusion or use of of open source is not so easy to measure especially at the macro level there there are some uh the the studies frank nagel did where he used kind of propriety uh databases close and was an approach um but it's based on the quite limited number of of of companies um uh for for the us um if if that's data is is public available it would have been good to use it but um so far it's it's not um um it was not feasible for us uh i'm also seeing uh dick's comment on china and we in the study we do some analysis on china and some interviews and some with some experts um who said that uh due to the language language and some some good culture barrier there um and also view a particular reasons today that's of course uh an increasingly important point now um there's a large kind of um a group of chinese developers who are not on github and who are therefore not included in this analysis i mean they i definitely agree also with what dirk said um and i think you know this kind of point on what what is the databases um you know has been raised a few times and i think it's it's a very very fair point um on the other hand at least for our study considering that we're analyzing in europe is probably still a pretty comprehensive data set uh considering what is you know realistically used in europe just seeing if there are any further questions now if you still have something i guess that's on the discussion that simon kicked off yes uh okay let's see carlo asked a question yeah maybe one last issue regarding the the universities um indeed and that was also um one of our observations um teaching open source is is not so common yeah um in in universities here we we are happy to have kind of on board who gives courses on on this topic um but we also a little bit screened master programs in europe and uh we really didn't find a lot and this is this is uh also i think um big uh we are suggestion and to uh to really get uh also the the the students especially at the technical universities which are obviously the the the hot spots um and for contributions but also then at the end for for startup hubs um that they they they get the you know how not only to to to write the code but and then also maybe to commercialize it in in a second step with with the kind of viable business models um for for open source based startups um if i could say a word or two to that also i think i think um there are two kind of immediate ways i think universities can do a bit more with open source um the first is that um that student projects and thesis work should be should be posted on on github or git lab or some open source software repository so my my university does this already at least my department um of course that doesn't mean you're contributing to an open source project but it at least gets people uh one step closer to doing that um the other idea would be to tie funding to uh so funding uh research funding is often tied to or or makes it conditional that you have to publish your papers open access or you have to share your data uh if it's possible and i think that uh adding uh requirements to share your code could be an important nod or push in the right direction uh and that's something that at the national level uh the research uh research funding agencies could could enact quite quickly i think not everyone would be too happy about it it's hard to take your code um and put it uh put it under the watchful eyes of the crowd but i think we'd all be better off for it and i think you would have a virtuous effect and and lead to a bit more open source um so that's maybe the second way i think about it yeah i think that's in the in the addition of open signs with open access which which has been put forward by by moe does um you know previous commissioner and um but it's uh yeah it's sometimes a little hard to kind of go go this way but i think in the long run this is just the way to go yeah um i think possibly we've answered then all questions i see an interesting discussion on uh hard open hardware and uh osmos which is something that i think we haven't uh we've discussed yet of course osmos uh uh a concept is gaining gaining a lot of traction uh and open hardware at the same time also something that's gaining a lot of traction um but it hasn't been kind of or isn't in a very um structured i hope isn't happening in such a structured way so i think it's an interesting point from here um i think deep mass not with us anymore just check the list so i do think then that we maybe patent first yes this is a good one i think it's maybe a discussion for later something that nobody i mean finally the person who has solved the question how do you how do you measure this uh you know the issue of measuring innovation through patents and then open source because it does seem i remember couldn't you look that it there's a relationship there patents and open source still um but it's difficult to to uh to attack this from an economic econometric point of view as five and if that i'm not an economist so i'm gonna uh there will be new where we have just finished a master thesis on this um on the on the individual level very interesting insights but uh we'll take some time maybe next next year's symposium i will present the paper on this all right i think we are um then probably um out of time with this session so i would say that we uh move on again thanks a lot to Johannes and Knud i think that was really really interesting