WEBVTT 00:00.000 --> 00:10.000 Thank you, Peter. 00:10.000 --> 00:13.000 Thank you for the opportunity. 00:13.000 --> 00:24.000 I have my own consultancy and this research is funded by digital infrastructure insights fund, 00:24.000 --> 00:30.000 which is involved for foundation and salon or mid-yard management future. 00:30.000 --> 00:32.000 So no dirty company money. 00:32.000 --> 00:37.000 So there is a preliminary report that I have been working on. 00:37.000 --> 00:44.000 The qualitative part of it is being done by me and quantitative part is in collaboration with 00:44.000 --> 00:47.000 that IE. 00:47.000 --> 00:55.000 And we were looking at whether public DNS resolvers, whether Google and CloudFlur, 00:55.000 --> 01:02.000 whether ISPs use these public DNS resolvers increasingly. 01:02.000 --> 01:08.000 And instead of using their own and how are users doing? 01:08.000 --> 01:12.000 What are users doing when using the Internet? 01:12.000 --> 01:15.000 Do they use public DNS resolvers or not? 01:15.000 --> 01:25.000 We looked at APNICLA data and it looks like that despite the common conviction that 01:25.000 --> 01:32.000 that the use of public DNS resolvers are increasing and ISPs are increasingly using them. 01:32.000 --> 01:36.000 It looks like that it has been going down and has halved. 01:36.000 --> 01:41.000 So the reports describes that kind of trend. 01:41.000 --> 01:50.000 However, this doesn't mean that ISPs do not use public DNS resolvers and in certain parts of the world 01:50.000 --> 01:54.000 they are used more than others. 01:54.000 --> 02:05.000 And this is the trend that you can see that almost all of the regions you can see that it has actually gone down. 02:05.000 --> 02:14.000 However, there are APNICLA data, there are other sources of data but from the methodology perspective. 02:14.000 --> 02:29.000 Jeff's APNICLA data is actually closer to what to understand how ISPs and network operators and users use public resolvers. 02:29.000 --> 02:38.000 And now I didn't want to, I have done this presentation a few times and I wanted to talk about the open source aspect of it. 02:38.000 --> 02:48.000 So can we actually see because most of these trusted public resolvers are closed source and they are very popular. 02:48.000 --> 02:52.000 So how can we actually stop if we can call it a tragedy? 02:52.000 --> 03:02.000 How can we call it a how can we actually stop it and how can the internet community help with providing a maintaining open source DNS resolvers? 03:02.000 --> 03:07.000 And I don't mean not providing commercial aspects to it. 03:07.000 --> 03:14.000 So also why do we need public DNS resolvers? 03:14.000 --> 03:20.000 We can just say well ISPs are going to provide it or we just mandate the ISPs provided. 03:20.000 --> 03:26.000 Our public DNS resolvers are needed for freedom and decentralization. 03:26.000 --> 03:44.000 And this is a theory that I have and we identified the high percentage of internet users in regions with low internet and press freedom that use public DNS resolvers when they face censorship. 03:44.000 --> 03:57.000 So software in DNS resolvers can provide a more affordable solution so that ISPs can and this work operators can use and it is known to be less costly. 03:57.000 --> 04:03.000 So and that's a correlation that you can see in Western Africa they're just all in the report. 04:03.000 --> 04:19.000 I'm going to have can go on. So five minutes. Okay, great. So how can we promote the use of open source and free public DNS resolvers so that we can get so that we can have more trusted providers. 04:19.000 --> 04:28.000 We do have public DNS resolvers, many public DNS resolvers, but how can we actually help them to scale. 04:28.000 --> 04:38.000 And so one hypothesis is that we had in the research was that perhaps ISPs don't use open source DNS resolvers or don't have trust. 04:38.000 --> 04:49.000 But in based on our research and the interview is there was no such thing they didn't really say that they don't want to use open source public resolvers. 04:49.000 --> 05:01.000 So there was there was trust among the interviews that some of them actually said that we purposefully use open source public resolver. 05:01.000 --> 05:30.000 But there are there are other reasons that public DNS resolvers are lacking one is legal liability unfortunately there is a surge and increasing demand of the government to block domain names and this usually happens at the address at the DNS level and regulatory frameworks are out there and even like and this is not an authoritarian regime discussion anymore. 05:30.000 --> 05:45.000 This this is happening worldwide and scaling not for profit and open source this DNS resolver is hard and this work operators legacy decision to use closed source has. 05:46.000 --> 06:03.000 So they do not make that transition to use open source and some network operators had hesitation that open source software in DNS resolvers might be less secure but that is not that was not the reason and we need to debunk that myth as well. 06:03.000 --> 06:08.000 So in the research either so how do we debunk that myth. 06:08.000 --> 06:16.000 So Bert Hoover in 2009 he wrote this amazing blog about rights. 06:16.000 --> 06:29.000 Many good blogs but we need to debunk the myth that open source DNS resolvers are less trustworthy or insecure because among also the legal. 06:29.000 --> 06:35.000 And the policy crowd there is this impression that oh if you want something to be here it should be closer. 06:35.000 --> 06:44.000 And also creating best practices on how to run DNS resolvers so that we we reduce the transaction cost of providing them. 06:44.000 --> 07:05.000 So that as at right task force on DNS resolver best practices we came up with a bunch of recommendations one included like how do you use open source software but it also mentioned like privacy security and also other best practices. 07:05.000 --> 07:10.000 I mean the chair of that group is here you can ask him questions later. 07:10.000 --> 07:20.000 And also help with legal compliance if you don't know already I come from a policy background I have not operated the DNS in my life. 07:21.000 --> 07:49.000 So I try to understand the technicalities but I think that creating a legal policy track that can help public DNS resolvers to provide compliance advice to and to also like help them to say no to some of the blocking request and help them to win courts win courts and like globally would be great and that's one of my ambitions next time. 07:49.000 --> 07:59.000 So I'm hoping that will happen and that's it if you want to talk to me about this I'm around here and thank you. 07:59.000 --> 08:21.000 Well I believe in the global well where does the ambition to create a collective action kind of global legal compliance that can give advice to DNS resolvers come from. 08:21.000 --> 08:41.000 So I believe in a collective action and I believe that some of the the result of the research we can see that there are the DNS resolver sometimes they don't provide their services because they say too costly because it like the regulatory regime has put pressure on that. 08:41.000 --> 08:52.000 So I think that one thing that we are missing here we need a room I know it's a scary but full of lawyers that are here to help not to tell you you can't do it. 08:52.000 --> 09:03.000 Right and my name is right out inside work I can and I used to work for and I used to be part of the security stability at my sweet community. 09:03.000 --> 09:24.000 And within I can and that you now is doing a project where they do a taxonomy of open source software and it's related to the work you presented here where you discuss regulatory frameworks and stuff and European Union type in heroes as I don't speak for either of these organizations. 09:24.000 --> 09:34.000 But next week that's DNS or which happens to be in Atlanta on Thursday and Friday and weren't from Marie from from Google. 09:34.000 --> 09:45.000 That's exactly the presentation of just that's very much aligned with you much with you just presented so get in the room together and exchange in for if you don't already know. 09:45.000 --> 09:49.000 Great yeah I'm a member of the NS work and what thank you that was great. 09:49.000 --> 10:15.000 So how did you classify it in a server as a proprietary because there's not many of them did upload so it's usually open source in it just a nice box that would be a batch so I think that probably needs more detail was actually proprietary and also explaining to the users. 10:15.000 --> 10:23.000 They talk to you that they're really using open source anyway. 10:23.000 --> 10:27.000 Yeah exactly so one of the so for example public. 10:27.000 --> 10:36.000 Google the so I do can I do categorize open source versus commercial. 10:36.000 --> 10:53.000 So one thing is that the DNS resolver says itself that I use closed source so Google public DNS says I use closed or yeah and so and Google public DNS has been used it's increasingly used in especially in Africa. 10:53.000 --> 11:00.000 But point taken I think I wish to consider that in the future. 11:23.000 --> 11:51.000 I think that's a really great idea why I have open source developers have not provided like these templates so that you can easily set up your own DNS resolver using. 11:51.000 --> 12:15.000 So I think that's a really good suggestion I have now I've seen that some so for example in the best practices we discuss how to set up and how to how to run it but we have I have not seen a template but I have to research on it but I have not seen a template on that but I think that's a good suggestion I have to go and do some research and see. 12:15.000 --> 12:22.000 That's part of the DNS to anything like that. 12:22.000 --> 12:31.000 But you see that your configuration is simple for you yeah so I think thank you so much that was very. 12:31.000 --> 12:34.000 We have one question of matrix. 12:34.000 --> 12:56.000 I have I work on encrypted DNS but from the policy and advocacy side of things but so I don't really know the answer. 13:04.000 --> 13:14.000 Thank you.