# How web3 changed open source - Ivan Bjelajac | Attic42

- Channel: [ETH Belgrade Community](https://streameth.org/eth-belgrade-community)
- Date: 2023-10-07
- Duration: 31:20
- Watch: https://streameth.org/watch/yt-0EWByUJngJ8
- YouTube: https://www.youtube.com/watch?v=0EWByUJngJ8

## Transcript

well everyone thank you for coming this will be a very interesting I think presentation because I will try to fit in two presentations into one time slot the second one is by a guy I never met but I want to stand in for him because he didn't get a Visa because he didn't ask and that is uh that is you know a very good way not to get a visa to Serbia if you don't ask but you need it so I will try to squeeze in both lectures at once after my slides I I will have his and I will try to do it as good as I believe he would have done it if he was actually here I'm doing it out of respect to his long useless flight from Thailand to Serbia or whatever so today I wanted to talk to you about how web free changed open source and considering this you know room is the size that it is and it's not like there is a million of us here you know we can always we have chats about certain Concepts I'll try to make pauses if someone wants to ask something just start waving your hands and we can we can discuss it you know even if it is even as I go along you don't don't really have to wait for me to finish in order to ask questions uh the same thing about his presentation that I know not a lot a lot about but we'll try to do my best there too so myself this is me I'm a washed up engineer I build software for life then I build companies for life now I don't really know what I'm I'm actually doing so you know considering that I I I'm very happy that a lot of you came here today it's actually more than I expected I expected when I saw it it's Community stage I thought the turnout is going to be like it actually was in 9 30. so what I wanted to talk to you about is how open source is changed by web free and did it even exist to begin with which I think is maybe even a more important question here how many of you know about this article by Eric Raymond you know Cathedral versus Bazaar yeah so not too many people you know this is for all the guys my like myself this is actually uh one of the first things that I read back back when I was young you know when it comes to ideas behind software engineering I think the main thing is ideas behind you know this is not something that you are reading you know to learn how to code or how to become an actual engineer or whatever this is something you are reading to understand the philosophical and social differences between different types of companies maybe maybe a different phase of building software and this is the the time you know when we used to Pirate stuff here in Serbia but I mean a lot more than we do today and the this work by Eric Raymond actually explained how centralized corporations that we hated at that time like Microsoft build software compared to the new movement that was that was happening which was open source and he basically compared uh these centralized company companies that dealt with long release cycles and with the uh centralized planning and waterfall and everything he compared them to a cathedral where there is a Dogma around how are you doing things uh he compared them to a Bazaar where you basically just come in and start selling stuff and uh basically that was his understanding of what open source is compared to what what the proprietary software is and if you look at these things you know in terms of structures in terms of how do people actually collaborate Etc these things were actually very different but if you look at it from today's standpoint you know maybe not a lot of people talk about Cathedral versus Bazaar but we do talk about proprietary versus open source in relatively the same manner and what I would argue is that this is actually not true you know maybe it was true in 1995 it's definitely not true today when we look at the way that the corporations are building software it is the same thing as open source is doing it there are a lot of tools that were created in the open source movement uh both from the actual tool standpoint and from the standpoint of the philosophy behind the tool and open source brought a lot of these things into corporations when you look at you know pull requests they they came from the open source world but how many companies do you know that actually you know have more than two Engineers that are building a proprietary software that do not use something like a git or another versioning system and who do not uh we have pull requests they do not exist you know if you look at it from a methodology standpoint there is nothing in corporations that is done today that is actually different than the approach of building open source the problem with open source on the other hand is that the only way for you to monetize open source for a long time has been to actually open a company so from that standpoint we can also say that the open source also changed in a lot of ways and there is no real open source when it comes to the way how are you developing software the main the main thing that we have actually between proprietary software and open source software is that proprietary software has a pretty bad concept of security by obscurity you know in terms of your software is secure because no one knows what the source code code is actually doing which we can argue that is just bad security practice in general and the second thing that we have you know in close all sorts of the development is that the software itself is a mode and it becomes much harder for someone to copy you because they they have a lot more to do in order to copy you than just to adhere by a certain license that you chose so when we are looking at the at this you know it's uh pretty much Blended in there is a difference between a security model models in terms of your access to the actual source code and there is a difference in different business models behind them but these business models are not really that much different in open source world and in proprietary software world so um I would say that from this perspective proprietary or an open source software are pretty much the same thing today in many ways outside at least from the Spectrum standpoint you know there is no cold war between proprietary and open source anymore there is no East and West everything is pretty much Blended in with a couple of different flavors and we learned this in many cases in a kind of a hard way you know when you're looking at open source software uh there are four usual ways how do you actually discuss open source and how it is actually monetized and of course like everything you know people decided to categorize and standardize them and we started you know usually in open source with Professional Services Professional Services are a good way to go you have no motor on Professional Services basically just your brand if you are a big contributor to an open source project maybe people will pay you more or hire you more often but at the end of the day you are limited by man hours in this case and this is why you know open source was looking for other ways to monetize and the first things that happened were hosting Solutions pretty much maybe you didn't want to deal with the infrastructure so you just want to Outsource the dealing with the infrastructure to someone else in and you know you're going to pay money for that and that becomes a part of that open source project in both cases you know Professional Services or in hosted Solutions you are paying either a freelancer and the contractor or you are paying a company you know so in that way there is no open source without companies or you know trusted individuals and these two models proved to be relatively scalable the main thing there is a relatively you know it was hard to defend against the hosted Solutions because it's easy to learn how to solve something and it's also uh not you know if it's analyzing software everyone can basically learn how to build stuff on on top of it so they were hard to defend he had no mode and they were relatively scalable so open source turn into open core with projects like redhead Etc mainly and what happened was that we started having software that has paid components uh there are a couple of the people here that worked with me that we were trying to do a proof of concept the proof of value project for uh for polygon you know before it was even merged with Mantic and we tried to um basically Fork one uh interesting uh Decks that that was called and is still called Loop ring and we tried to do a friendly Fork of looping and we basically wanted to show that polygon compared to ethereum works much faster and that you can run the same software on it and this is another point that I have there you know they these guys basically told us uh it's free for Youth Of course to you know Fork looping we can get to know each other this is very nice if you want uh to use blueprint though you should maybe think about buying our order book and orbit or their book was like maybe a hundred or 200k I don't remember anymore so we've said you know let's uh just uh build another order book and put it in and see what happens we just need to use the same order book on ethereum and on polygon and you know show people that this can be done and what happened was that we soon realized that a big part of the pretty much core looping code are actually in the paid proprietary order book so you cannot actually run looping with another order book or understand what the attack actually does unless you have the the order book and I actually wrote to to the guy you know who is the CEO of luprink and told him that you know can we do something about this and this really seems like uh we are the approach you know in building open source projects and he told me that it's not that this is completely normal and that I'm crazy he got a bit mad and considering I currently believe you know and this lecture shows that the open source doesn't really actually exist I think he was actually right so you know sorry for him for you know blaming him for for that that thing that he did but I would say at the end of the day that this is what I'm talking about when we are talking about these Blended Solutions between proprietary and open source because at the end of the day you know all of the projects because they have to have a centralized company behind them and that company has to play pay people and they have to do things they need to have a business model and if they have a business model that business model needs to become defensible and if your code is a part of making that project defensible then you are probably going to try and do it at the end of the day you know open course projects open core projects pretty much usually are trying to give you additional tools that you can pay for but at the end of the day you know both both hosting projects and open core projects are also trying to use code in some way to defend the core project and to make you pay for it in some way so at the end of the day there really is no free beer or free lunch we are talking about mainly mainly projects that are just trying to monetize in the best way possible and proprietary has this also when you look at the way you know that open source open core and proprietary software are monetized you know we can really see that the proprietary fall of the open source in this path you know proprietary saw what open source is doing they said well we are already not giving you any code for free so so we are going to look at the premium models we are going to look at free trial models we are going to get you hooked in other ways because that's what you are doing at this position you are trying to with open core with marketplaces with hosting you are using the free version of the product to hook someone in order to get into your system and to start paying for something else later and the company that is going to benefit from that is the company that is charging the service and then came Bitcoin and ethereum and others too but you know I would say that these were the first two projects uh that showed the open source world that there is suddenly a third sorry not for the fifth way for you to monetize open source and this is through tokens suddenly you know uh these are the first projects and the other projects that came after them at least the legit ones these are the first projects that showed that you can actually accumulate in the crew value and the get value to the open source contributors just by building the software not but by creating a company behind it sure some people created companies sure there is ethereum foundation but at the end of the day ethereum foundation is a foundation and you can get money you know based on based on what you are doing for the ethereum community without directly benefiting a singular company in any way or getting paid by a singular company just if you are an ethereal holder you are already a part of of this journey in some way because uh you know whatever is uh is built on ethereum is supposedly supposed to increase the value of ethereum itself and at the end of the day we are going in that direction that smart contract platforms and other types of blockchain platforms are basically showing that they're currently probably the best way for us to build the next version of the internet this is of course arguable you know it's possible we are going to figure out a better way to have a fall taller than internet than blockchain but that part didn't come yet and this idea you know that was summed up in this fat protocol layer thin application layer thesis and driven by tokens is what got us the type of ethereum that we have today with all its good sides and all its flaws and when we are talking about our ability to build economy to battle fake news to basically play around with data in many ways going forward to uh this is probably the more Superior way to do to do it you know I'm not just talking about you know is this better than Banks or something like that it's actually not important if it's better than Banks the important thing here actually is that it is more integrated than Banks first time in history of building software you can become a developer you can create an open source project monetize it through uh talk some type of token and suddenly also have a built-in payment mechanism and payment processor in the core technology itself you do not have to play around with the using third-party services and this is also going to solve global warming you know because we can create Virtual Worlds we can have wars in Virtual Worlds you know you can uh we can create an infinite number of metaverses where every country can get to be USA so you know from this standpoint because we can create virtual content that it's Limitless we can also expand Limitless and there is no point of us having any types of problems anymore there are a lot of people that also you know were talking about in the last couple of years especially when bitcoin price went down how Bitcoin spends the same amount of electricity as the country of Netherlands I did a couple of lectures when I was talking about that that maybe also means that we do not need Netherlands I'm sorry Netherlands maybe there are people here from from there but you know I believe that we can have a virtual Netherlands on top of ethereum it's probably going to be just as good if someone wants to do a proof of concept I'm very willing for willing to do that so hopefully you know uh we keep it keep Bitcoin and ethereum and Netherlands gets canceled that is the message I'm trying to send here uh considering that now we can have you know Virtual Worlds we basically do not have to have uh real ones anymore and it's probably going to become very expensive to live in the real world so uh you know hopefully guys you know you can have fun you in Belgrade while it lasts I'm not really sure what that future is going to look like but because it becomes much easier you know to build stuff in the digital world than in the actual world you know this is probably the world is going so ready player one woman might be our future anyway and because we are living in these weird times you know I would say the main play in web 3 or after web free basically becomes at the end of the day who owns the data because when we look at what people are talking about you know personal security today it is actually very funny because the the more technology that we have the more the more ways we actually have for people to exploit the technology uh our previous company managed WP or Devana Technologies was basically sold to GoDaddy in 2016. it was a very good thing I think for myself for a couple of reasons first reason is I got to work with some really really smart people and Go Daddy who used to build the internet from the 80s and you know up even before worldwide web uh and these were some really great people that I could learn a lot from the second thing was I I got quite a bit of money for me back then and I was really happy about that also I would like to happen again and the third thing I think was that when I was a part of GoDaddy I realized something that was really interesting to me and that is that GoDaddy spends I think 25 or 30 million dollars per year back then on a team of about 15 people running cyber security for GoDaddy in a very particular way you know this was not one of the teams that was trying to prevent attacks they were just trying to analyze attacks and they dealt with godaddy's system security system log files that were around 1.3 1.4 terabytes every month and with the 25 to 30 million dollar budget for software plus their salaries they were able to actually go through around one percent of the of that data so in about 99 of the cases GoDaddy had zero idea what actually happened it doesn't mean it was breached it just means they didn't know and from this perspective this is a problem that we have with the fortune 500s today this is why we have more and more cyber attacks which is this which is why the internet itself is becoming uh completely unstable in many ways and we are defending it not not by the core protocol we are defending it by threat party tools most of each and are open source in many cases develop and maintained by three people and in many cases not money cases not monetized because when you have free people you do not have a company and it's very hard for you to actually monetize something out of that the web free also has has a number of problems there you know we have very a very small amount of clients we have a relatively smaller small amount of core contributors and Developers I'm not saying that the current version of every is the solution but if we are talking about of the future and security of internet then we need to realize that all tools on the internet need to become monetized in some way that the internet itself needs to become monetized in other ways and not just in advertising well it's not just but you know I would say mainly in advertising and that we also need to have security and privacy and Trust as part of the core protocol and not not as a third-party vendor solution in general and this is what actually got me hooked to web3 uh I always liked open source because before I realized that it doesn't exist and I also believe that the web free today is doing for the internet uh what basically compute what computer networks uh back then got with the internet fault tolerance systems today are getting on top of the internet with the web free and I think this is an interesting idea that we need to think about and there is a big problem that we are actually lacking standards and hopefully this changes in the future because the main thing uh in the future is going to probably go into the direction where because open source is already I would say the main way that we are building the internet we have more and more corporations Building open source Solutions we could even argue that all open source Solutions are backed by some kind of a company outside of every and tokenized world and there are also quite a lot of them there too so the main uh if you we look at it from this perspective you know it used to be that we could have just read the data and we paid the people to produce the data and we paid for people to build the software and we consumed it and then we we became a part of the the process of creating the data ourselves with web 2 and the data was still owned at the end of the day by the platforms that we are creating it on and if we are thinking about web free in this idea that it is very important who owns the data and if we are looking at what is it currently happening with AI which is showing us that we are basically looking at the future of the internet that is not going to be driven by by what type of software do we have it's going to be driven by what the type of data do we have the access to and how much data we can be actually process if you look at open AI as a project first of all it's not open you have to it's it I think it's really really amazing you know that you call a project open and then you the first thing you do is you say in order to access this you need to to give to us for permission so permission the AI is a great project uh however however you know what permission the AI is uh doing it it's basically taking the data that the platforms gave them or that they stole from you or whatever you want to call it and they are basically packaging it for your use the the software itself that they have or the actual llm models Etc they're not that new yes they did some very Innovative and new things but the models themselves are five plus years old this is not it's not like open AI had a huge breakthrough in many things that they had some breakthroughs but they didn't have huge breakthroughs in many things they basically just threw more money and computer power and would give more access to data to to charge GPT and to their other platforms than anyone else did before and from that standpoint you know ownership of the data was actually the game changer for them and this means the tornation of the data is the future of the internet for us so if we actually want to own the data also or we want the not not to have big corporations owning the data then it suddenly becomes very important for us to have these protocols that actually enable that and considering that finally we have uh you know tools that we we can use to actually monetize projects that are actually open source and open in different ways than just having a company because at the end of the day when you create the company you have to embrace a certain type of behavior and from that standpoint I believe that we are in a better position than ever to actually do something like this and it's not so much about proprietary or open source it's more about gated or open access to data or you know your ability to choose who are you giving your data to so from that standpoint I believe you know we should be thinking about this new version of the cathedral versus Bazaar you know there was this old meme you know back in the early 2000s and you know it basically said on the internet no no one knows you're a dog there is this meme you know in blockchain role that on the internet no one knows you're afraid it's going to be more and more important for many ways because the problem that I just mentioned in terms of GoDaddy having in 2016 2017 when I was there this is the problem that we are going to have in one uh you know one residential block of new Belgrade in 10 years or so because the number of devices we have connected to the internet which are internet users you know you and me we are not really internet users the the real internet users are the the devices that we are connecting to the internet and that number is going to grow exponentially you know iot and 5G and 6G they're basically giving us more bandwidth to deal with the data and to deal with software than ever and to have more connected software than ever and this is going to become a huge security is going forward anyway so considering that the internet itself is pretty much garbage right now and it's relatively overrated you know we probably need to think we'll be thinking about ways to solve it and if you are asking me the best way that we currently have going forward is to think about uh fault tolerance systems as the backbone of this future internet so uh can we can we skip to the next guy's presentation so I do that too and then we have questions well it's not this one I can do this one too unfortunately uh we are going to uh proceed with the next lecture uh okay around 10 minutes so so no slides for me okay yeah we can try to to go like super quickly have a little bit uh yeah it's okay I don't need light so his presentations was about uh is it better to build you know build you know internet um also the presentation is about the difference of building software slow versus moving fast and the the thing that I wanted to tell you there is it is much better to build uh quality good software and secure software fast than it is to build uh very bad and secure software slowly and whenever you have a chance to build it fast and good always build it fast and good rather than slow and bad uh I it took me 22 years of software development to learn this it is an important an important message yes yes that is exactly true it also took me 40 years of life to learn that too yes yes Alan Ford knew it you know back in Yugoslavia and we we lost that knowledge so it's it applies to software too and always ask for a Visa when coming from Serbia if you need a visa yes unless if you don't this will happen and I will speak at the conference instead of you any questions yeah sounds good in that case I'll be outside drinking rakia and drinking coffee and whatever I can get to drink because I'm borderline alcoholic at this point and yes I'm an open open source alcoholic so from that standpoint you know thank you for being here and you if you want to chat about something I will be outside drunk thank you foreign [Applause]