Scott Moore | Radical Ideals: Experiments in Sustaining Open Source Infrastructure | CrossLink2019

excluded commercial entities for using this out here it could be a license simply says that anyone can use this we in charge may come pay for it but enjoy your license just needs to be maintained I see what Danny went into and then we go so why do we actually care through software like 96% of applications use open source software and as it turns out 57% of all kind of like every year is using their software so it's pretty common it's a pretty big part of our ecosystem it's a pretty print like the instance or a software ecosystem and so the question is you know given that it generates almost 400 billion dollars of value why is it that we have maybe one the two maintainers project who aren't they getting hit in a day before and it turns out the main reason for this is actually money so this is a bit hard to see but essentially it shows that about 62% of open source software is self-funded most of those people are not going to get anything through their work and the remaining portion of people that are working on the source are actually like getting paid by their employer or something who really has no stake in the open source software through writing it's great that's what sponsors that affect your number and so the types of things get funded by the employer are not the kinds of things that most people care about in the open source operating system so something like open SSL is never going to see working on it even though that powers such a large portion of the way and in fact it's kind of there's an interesting tragedy comments and the problem we're a more critical something in I can't like core in the structure the best thing need is actually funded by the current mechanism said society and this is particularly problematic because there's actually a huge difference in developer basically even the museum in South Korea in 2001 two hundred and eight thousand people use SourceForge not a huge number now 31 million plus people use github this is like just in terms of downloads actually like daily by the users and so it's not surprised in large part most open-source maintainer is end up getting burned at most open-source maintainer is end up spending a bit of time maybe three years or so we have a project that's like extremely got to the core system this was semantically why which was intended for new learning yep Fernan framework which was similar to reactor ejs and then Jack essentially just said no this was a good effort I really try to make this work but I'm getting because this is not really paying my bills it's not really actually not even to be even though it's valuable to the ecosystem this is another again probably hard to read example but this is actually open SSL robbery how many here familiar with hardly in late 2014 decided familiar yeah there's a few people in the audience so in heart lead was this critical wonder hoodie that affected a very large part the way up cost about two plus billion dollars in economic damage and that's because open SSL is our student aid by the one person and that person is getting paid another thing was trying to work job which didn't support doing that in addition to writing this piece of software that grant the entire like SSL like and actually as a result this is being like dedicated like has been less probably used as a great value librarians more recently but the reason that campus is the first place isn't due to one or to be in temerity were part of this and so it's funny than that like you know there's all this values for hundred million dollars or so approximately probably more to get honest it's probably an underestimate a value this great open source software and yet most of people are working on a nurse woman's condition this is a more recent one how many of you are familiar with like a ministry how many of you actually are in like the JavaScript ecosystem are familiar with their JavaScript like note in particular so there's a library called the dentistry men stream was essentially heavily dependent upon and the thing with the been stream is Dominic art who's maintaining it at the time sort of got tired of it he decided to do it off to another maintainer someone he didn't know and that person injected a series of wishes needs to the code base which then resulted in some pretty catastrophic consequences now you can blame tonic tar for this and you can say oh like it should have really tried to be was getting to do but the fact that matter is that didn't actually really have anyone to turn to to help him maintain this just ended up giving this off to someone who would take it which was like this random person after trying for like months to get someone to actually help him out so this is just as from anecdote that I turned out here but and he was basically saying like okay I worked as a dishwasher at restaurant you know I made a mistake need to confidently I'm going to cook it was like a fifty cents an hour pay rise with like more responsibilities and then kind of makes a comment here that like writing plan for the module is like that but without the pay rise so it's actually like kind of funny to me that we're sort of and you know these are both obviously great like paths but it turns out that the wine that's having this outsized impact on the world that we live in today for millions of people billions of people is just not getting support and so after being the downer like let's try to think about how we can solve this problem and obviously so I work for a project on Bitcoin gameplay and inserted and some of my assistants or suffered it's not a coin actually this is important though it's very much just a marketplace for people to find and others to work on their projects and I'm obviously biased it's a solutions that provide but I'm going to try to give a wide range of solutions including some that we're not bursted working today but in general I want to touch on the fact that in this industry and sort of watching ecosystem in a 3m in particular there's a lot more interest in open source software than there was you know ten twenty years ago I'm about to so I think that the reason for this is because we actually kind of ability to one program sort of our values into money and to to align incentives between people in sort of the investment side of the ecosystem and the developer side of the huge system and the reason for that is you know if you go and find some more pieces infrastructure in his face ideally that improves quality of ethereum activity that improves your interest in the ecosystem and that wasn't really something - and so all these projects is when three actually have money for open source software and open source by default and that's the thing that it just pluses the case in what - that I think is really powerful and says that we should be like a capitalized run for web three and so yeah it was like up to 900 million market cap everyone needs dubs everything is open source everything is filtered mode and that's the standard that sort of led us to this project which is largely about trying to increase this funnel for a user to contributor to maintainer so lots of people use open source software we've already established I've not as the need those people actually contribute to open source software and then even if you err those people actually need to know so it's like I think the actual stat I mean you can go back to the chart I showed you earlier but it's like one point zero one percent of people were actually core maintainers as we gave it library versus people who are just like using open source software and so the question first is okay how do we get more users to become contributors ideally you know that's the starting point just getting people aware of the fact that they can't give back to the consumer side where that they're already using and that's where this comes in this is like largely just a county model so how many of you from they were thinking around a pain or resource or similar sorts of projects that they try to sit in their model they try to do what - and I think they sort of failed was because they didn't have this aligned incentives excessive between the project and like and the like the investors and so this is really fundamentally where we think gay boy actually improves there's sort of just like this very simple flow from a funder take Sofia Vishnu and puts that into our are basically website and then a cover who goes to the finds or services issues with it like art Explorer let's try again and so this is just an example of one of the issues so you've essentially got just the details about like whether the work has been started and you'll note it's very hard to see here but like this is a very small dollar about it's something that's very granular so that's sixty dollars for work and the reason this can happen is because people are working on a whole bunch of these at once and finding these issues as they need them and so because this marketplace is so that kind of granular it's easy to put up these smaller tasks and still link in the meeting point value of the people and that's actually probably the most valuable for projects because it's very difficult to find someone who's willing to jump in to working on I have a core feature or something it's like a huge part of it wrong to remember support pieces of technical infrastructure it's easier to get people started out being smaller is used so that they can get a sense of what your project is about and then start working with you on like this path towards becoming a maintainer and so today these are just like some status basically we've got somewhere in the range it's like you're 500 plus funders 20,000 developers we watch'd we're 2017 usually this piece is like a mesh network graph of basically well actually it's the literary graph of funders to computers but it's it's actually just static right now because we're not right but in sure in short that's certain how that flow works for these are to deteriorate our minds and then we've got this contributed maintainer flood and so that's really where we get into some work we've done with grants and what grants do this is very challenging currently what grants do is basically give you the ability to fund someone on a monthly to defend for their work and give them the opportunity to become sort of maintainer over a defined period of time before you bring them on like a third party or at least oh and we've actually seen this go really well within certain projects as a theramin ecosystem that were looking to expand that to my their ecosystems as well we eventually want this to be just a general open sources rule and answer to do that we do need to think it's like fiat payments is it so forth eventually we hope that simply payment activism but we're still working towards that so this is just one example of the sort of matching funding we did which is based on this paper that italic so we went last year in short what this paper says is a like quadratic funding or quadratic remoting or quadratic mechanisms are interested let's take that and apply it to open source software funding or just generally to do translated Commons type problems in order to democratize climate and to allow people to have more to say in how these common goods are actually funded by by giving and so a good example is just like if two people were to donate $40 then actually that would be it's very good forecast this burger for us where some of this is where 16 and then similarly if you have 8 people donating $1 teach the matching amount would be 64 and seen it like I think I actually have a chart here even though the total amount of funding and stay the same the actual matches in our group has grown quite a bit and the point of that is even though people are donating the same or even like you know even if the matching amount is to same across the board if you're working for any polymer and especially then it's more democratic there should be more funding for that project and so we've done a few rounds of this these are the projects that you're spending today kinetic labs like this flop when we start and a few others this is her second round which had actually department funded regardless of what you think of that hopper prismatic labs and possibly a few others and we just finished a third round which actually had about 250,000 in total they needed to source from about 2,000 different contributors we haven't formally compiled the results of that round as yet but we should have some results in the coming weeks on our blog and our site for your particular day but a lot of the problems I've talked about here are hatred into models they meaning that like we need people that give us money it's like philanthropic for this is actually to see so how do we go past that how do we actually get people to do something more than just giving money in terms of their hearts I talked a bit about aligned incentives I think that like that helps I think it that means we'll donate but effectively they are still sort of donating and you know it's it's not quite a sustainable longer and so a few options here which are increasingly controversial our installation funding lines this is like regardless of what you think of it just an interesting model for how am I going to find base level development this is some of its cash does that if you ever changed to very well and this was one that was proposed by what our co-founders title walkie yeah King 1789 for inflation funding you can find the API standard it's really Bo and it goes into a lot of detail about why basically a block reward or some kind of going to developers would be really fun to play of interest similarly we have yeah yeah yeah IP 2025 which is very much of the same thing but for the one point X I won't go into too much detail there and then we have support contracts so for us it's bgg booty DJ story runs about a hundred different and cam packages and in total I think they have somewhere the range of 100 million downloads and he decided okay so like even if he was actually not able to get funding for his work in a sustaining fashion which is pretty well just given how many people it's his work and so I said okay I'll have a support contract and just offer consulting offers his border offer even insert promotion of my vote and this is sort of an experiment that he's running in order to try and understand where the market lies in terms of like enterprises work where like corporate me a judge decides now the problem that this goes back to what I mentioned sort of beginning which it's like a lot of these funders aren't really concerned with like large pieces of open sources and starts for the Congress to let their concern replay docker or like individual pieces of software that they use in their companies which is totally fine but it's not sustainable like it doesn't provide us with a way to get his larger like important core pieces and stretch the funding and actually this is a very interesting experiment that for Austria just after it his support contract experiment which is advertising in terminal and he's got some backlash but he makes you decided look at people that could have like fun to my work let's just put an advertisement while you're doing your npm install and let's just see how people react to it which it turns out really poorly but basically this was a good experiment in like an advertiser type model and I do think there's something here I don't think advertising used like it explicitly that I think that the way that Enterprises has been done historically is bad we shouldn't have the same level and sort of get your fear it's as advertises but I think there's a level by bit chicken still valuable and so unbiased saying of course because we do actually have one of these like as our server models we're trying this advertising model is essentially an ethical advertising model that works ideally both on his memory those ironies and on various sites across the web like orgies - and so forth we're developing focus sites there are in waitlisted and are well-known can use this to essentially fun basically fund their development fund their work through advertisements from other development centric companies and offering just like some random spam company which is when they would usually end up getting is they'll pay the most if you're insisting like for example or like actually yeah sorry donut and so this is actually extremely I think important to have probably a discussion around that the climber ties it it's not really the topic by my talk so I'll leave that off for someone else to like going to but I do think that thinking about this model for opens responding eyes extremely interesting and so those are just a few models that's like kind of where I wanted to end this discussion it's just a survey landscape like what we're working on but I wanted to leave some time for people to ask questions just doesn't like it either their own open source projects or about open source finding each other leave and to get a sense of how people are thinking about this problem in this room so all been up for questions to have it thank you all [Applause] Oh quite interested in the UNESCO project and I wonder how he's started in the the coin project how to see how people contribute to the project of Bitcoin how do they did it do you miss well especially yeah so so you know so obligated and really just decided that let's try to get more introduces law like go ahead and there were a few different smaller bounties that you but I think it was mostly to do with front end work just like cleaning up parts of the TX please changed he mostly sort of came up with a psyche on his own after trying land exodus finding that most of them just didn't have very good UX but the reason that I think we were help Hayden is because we already have a network of developers who want to look at open source software and then at the same time we were only getting funding because people are scared about Hayden's project they just don't really have a nice way to donate so I think those were sort of the two things too that got us involved amuse clock we obviously it's an amazing project I think it's actually one of the best taxes out there but those resources I mean the sorts of things that we do in his case Oh I like to recommend the best way to it's a sample from the developer community if you have a massive project how to get information if your terminal is not here what is the best way to do that and how we can make it is this simple oh how we can put out the community thank you yeah that's a very good question by the way that we handle here today on our community so I think though that would be a good opportunity to dive more into this question like in particular I think there is a lot of opportunity to get people in either D so the cryptocurrency community through get go engage but there's also ways to get its web to developers generally engaged so for example like Oktoberfest is happening right now actually so theater repo actually everyone here at grocery throw ship it up like hacked Hubbard fest labels on issues that anything people can get involved in I think that's really the most important step is just like a lot of maintainer x' that go like I'll just like maintain this forever and all like I will bother an agent contributors all to those to postpone requests which is sometimes necessary but I think that by engaging those people who are frequently you're able to get those people eventually becoming familiar enough with the repo to become Romain Taylor's search ladies urgent as people I think is a really good first step in terms of the funding you're always welcome to open up a great grant I make my grants so it's actually like open to any ones to use but I can also recommend open effective I know been deflected it is for fiat currency mostly it's mostly us-based but they're expanding globally and I think they do a good job also targeting a different pool of donors to their sort of grants pool than we do and I would recommend talking to them as well - any last questions awesome thanks so much [Applause] you