RIPE 89.
Plenary Session ‑ 2pm.
Main hall.
28 October 2024
Opening plenary:
MIRJAM KUHNE: Welcome everybody. Hello, welcome. Please take your seats, we have a minute or so, half a minute, please come in for the welcome. We have quite a packed agenda for the opening plenary so I'd like to start on time. Its great to see so many of you joining. Come on. It's not quite two o'clock yet, I will wait a few more seconds for people to take their seats.
Okay, well let's get this party started. Welcome everybody and my name is Mirjam Kuhne, I am the RIPE chair, for those of you who don't know me, at this point usually also introduce Niall O'Reilly the Vice‑Chair, he is not here at the tonight, you will have an opportunity to talk to him later in the week, I am sure he is online and watching me to make sure I don't make any mistakes
> welcome to Prague. Just start with some statistics here. We have more than 500 people registered, 600 people registered, 485 checked in from 54 countries, I think it's amazing to see such a wide range of countries presented here. Lots of people from Germany, close by, Netherlands, always on the top five. And of course also a good number from peoplemm from them Czechmm represe republic,n I amt sure there will be more people coming in during the week
I am also pleased to see over a hundred newcomers here and we had a good newcomers session and please don't hesitate to come with any feedback to me or anybody else from the RIPE NCC or RIPE team.
We also have a number of local hubs again, this is really starting a bit of a tradition, we have done that two times, it means that some people get together in those countries this time it's Romania, Poland and Yemen and IP stead of sitting alone at home watching us, they actually get together and participating in the RIPE meeting as a community in their countries which is great and you can see more information about the local hubs online. They are not going to be here all week all the time, some of them are only two or three days, we will get some more feedback from them later on.
This is actually the fourth time in Prague for the RIPE meeting, we talk a little bit about the history, the first time was RIPE 14 and in 1993 and that little table, you don't have to read it, it's actually in the minutes, you can find it all online. That's a view of theEMP B IP servers at the time. Anybody remember what that was? Oh there are two people there, three, wow! That's fantastic. What did that stand for? I just remember the junk it was.
AUDIENCE SPEAKER: It was the European multi‑protocol backbone and that translates to an antiEbone thing that was sort of, I think it was funded by Germany. It was crap.
MIRJAM KUHNE: All right, thanks for that history lesson, introducing a CIDR BGP 4 European networks and proposals related to IPV7 which didn't quite make it in the end I suppose. Then we came back here in 2001 for RIPE 14, there was a discussion of global policy development, we need that or is it more better regional, there were a number of V6 address space presentations and then we had ‑‑ address policy presentations and then we had summaries of working groups at the end of the week and so in addition to the number of working groups, we still have today, we also had those three working groups, I don't know if you can read it, there was a Test Traffic Working Group, there was the CentreDNR Working Group and Deliveries Working Group. Yes, we don't have those any more, at least not officially. And we came back again in 2010 for RIPE 16, then at the time test traffic measurements working group became the MAT Working Group and also at the time we had discussions on what to do with the last of IPv4 and the picture you see there, you can find it online, is the one of the BGP key signing parties, you might still recognise some of those individuals who were active at the time.
So that was just also I'd like to dive into, I am glad to be back here and thanks Prague for the local host to have us, we will get to that.
As always, we have a RIPE Code of Conduct and you can read the full Code of Conduct online, you should have ticked the box when you registered, if there's anything you would like to report, anything you feel uncomfortable with, we have a Code of Conduct team, we added two more members to the team, you will see the pictures here and also on the walls outside and they have a room there, a dedicated room out here if you would like to meet with them, you can use that.
Then we have number of working group chairs who are all here this week and are responsible for the agendas and the content of the working groups later in the week. Unfortunately we are missing one there on the top and that's actually Eric Bais who unfortunately passed away shortly after the RIPE meeting in Krakow last time and I for one will certainly miss him during the week, I would like to ask us for a minute of silence to remember Eric at this point.
(Minute's silence)
Right. Thank you for that moment. We also have a book in memoriam book outside in the registration desk if you would like to share some thoughts or memories and that will and hadded over to the family, not everybody had the opportunity after the last RIPE meeting to share some memories.
There will also be an opportunity in the Address Policy Working Group on Wednesday to come together and share our memories of Eric and remember him.
Right. So, moving on. The RIPE Programme Committee is responsible for the plenary sessions, you will hear from some of the Programme Committee members later on. There will be elections this time, two seats will be up for elections and we are having quite a number of elections governance related activities going on during the week so I added a slide and there's a link to the website to find more information, we'll have elections for the NRO council and the address supporting organisation, address council, no? Yes, advisory council, no? Address council. AC! There are elections going on during the week and you will hear a presentation about the details later on, the Programme Committee has two seats open. There's also you seen maybe from the RIPE list a RIPE nominated committee started its work so there's a chair appointed and the chair is now asking for volunteers for the NomCom to then go on for selections the next RIPE chair team.
We'll have a number of working groups looking for or selecting new working group chairs, I think there are already a number of candidates and working groups will do the selections during the week, you will have that happening and lastly on the list is a slightly different things, it's will go related to our governance an that's the questionnaire that was sent out by the ASO AC on the principles of creating an RIR known as the ICP‑2 document that was discussed here earlier and that will come up during the community plenary later in the week.
Right. So here's the meeting plan, you all see that online. We don't have ‑‑ I keep looking at my badge because but that was the olden days when we had a big booklet as a badge, we don't that that any more, there's one small change, the diversity and inclusion session are the Tuesday now, swapped it around on request, and instead of that we have two BoF sessions on Thursday. The rest is pretty much as usual. Today and tomorrow are plenary sessions, Friday, plenary sessions and Wednesday, Thursday, a little bit of Tuesday also working group sessions. So they can all wrapped up in plenaries.
Today also in the afternoon we'll have a quite interesting session of the Best Current Operational Practices Task Force, two main topics, one is continuation of the initiative to document operational practices in times of war and natural catastrophes and there's update on that and some progress, and the other topic is an idea to possibly document how we do, how we build consensus in the RIPE community and we can talk about more about that, we always heavily lean on the IETF, it's something we do differently and discuss if you need to document that.
And then the two BoFs on Thursday are ‑‑ one is actually an open platform or panel discussion with some representatives of the local network operators groups and see how we can maybe work together a bit more, what does RIPE need from the NOGs, what does NOGs need from RIPE, how we can continue to collaborate in the future and cross pollinate each other and the other BoFs is a continuation of the task force that was closed, the task force that looked at recommendations for open resolver operators and DNS resolver operators and how et recommendations will be implemented in practice, I think that will be also interesting, they are both in parallel.
The diversity session is tomorrow afternoon and they are always different and exciting presentations and don't focus on one specific community but how we can, as a RIPE community, can become more open and attractive to new participants.
Last but not least of course the RIPE meeting wouldn't be RIPE meeting without socials and with a lot of space and time to be able to engage with each other and socialise with each other. As usual there's a welcome reception today after the session at the end here in the hotel. Then there's a networking event party tomorrow night and then we have one on Thursday, the dinner this time in the National Museum, and Wednesday you are on your own to either rest, don't want to see anybody or arrange your own social.
And all of the information you can find on the website. And you should all have received information from the RIPE NCC and also your Meetecho link you can follow the presentations and comments and chat and question and answers online as well.
And last but not least, I will try to keep it short because you are having a lot of content, I would like to thank all the sponsors andthe hosts. We have cz.nic of our local host and sponsor of the meeting and RIPE NCC is a big sponsor and organiser of the meeting and lots of NCC staff helping out and AWS and a number of silver sponsors, IXO nd Verisign and IP for global by Hilco and response sponsor internet net and Cesnet connectivity sponsor, so big applause.
Next up I would like to ask Ondrej Philipe to come up and say a few words as local host of the meeting, the fourth time we are here in the Czech Republic.
ONDREJ FILIP: Thank you very much, the role of the representative is to welcome people, host to you great a company ‑‑ I usually because you know, you can go outside and there's some installation and you know we are a great company but more importantly, you know, for a few decades, I have been studying who important phenomenon of Czechs and Czech language, in Krakow he told you something how we translate to Polish, now you have the table, you can translate many words in Czech, so today I will talk a little bit about Czech culture, a special feature those real people have but before that, I have to start with an apology. I know inviting so many people to a Congress centre which close attached to a shopping mall is weird, it sometimes happens. I have an explanation for that actually.
You know, we have multiple layers of significant days in this country. Level one, important days. Usually that means nothing, they are usually nothing happens, people do not remember what those mean, you know, usually it's some tragedy happened maybe with the exception of international womens day but other than that, nobody remembers anything.
That's level one.
Level two, not much better, because we don't work on those days. These are public holidays, but we don't close shops. For example, 17th November, that was the end of communist regime in this country, what is the best way to celebrate? You go to a capitalist shopping mall and buy everything you can and that's how we celebrate.
(APPLAUSE.)
That's bad for people who are working there but they get some extra money for that, so at least something.
Then we have level three, you know what it's like today. Another example, for example, was assassination of one of our rulers, he was killed and roughly a thousand years and because he was lately called patron of this country, we celebrated, I know it's weird celebrating assassination which is more than a thousand years old but that's how it is. Many people complaint so some dude was killed and I cannot buy my bread and groceries. That's how it is. That's level three.
Level four doesn't exist. When I typed into the chat that shops are closed, I got like 20 replies, what about pubs? Don't be afraid, you know, there's no further escalation, pubs are not closed on any occasion. That's fine.
So now what happened the important today? I will tell you in a while but first of all what I want to explain here, Czechs are weird people, they are really fascinating but non‑existing objects, things, persons, countries but one of them is called Janet Zimmerman, imagery person we created and projected all the biases about us into the present so he was for example inventor, he was a musician, he created several theatre plays, he did almost everything in this country. Well the small problem is he is a majority but that's fine. On the other hand when there was a contest by Czech TV, the greatest Czech, you know, he was almost going to win it. So then the Czech TV consulted with BBC because this format was licensed by BBC and BBC denied, they said they cannot, that imaginary person cannot compete so unfortunately he didn't win but we all know he is a hero, that's the level of fascination we have.
Again, they had to change the rules of the contest. Which I am really, you know, kind of upset because in the British version of this, which was called hundred greatest Britons, King Arthur was competing and he is also not... but anyway. So there's the fascination of non‑existent objects. So what happened today. It was are you havely a little bit more than a hundred years ago when something like that was founded. People who are old enough or paid attention in the lessons of history know this was called Czechoslovakia and it consisted of four parts basically, Carpatia, a lot of names but then Slovakia and then three or four lesser Bohemia, Moravia and Silesia. Also king of Bohemia which was kind of before, which was the country we used to live before, we became part of Austrian empire which by creation of this was destroyed, we had a flag which was roughly the same as Polish, we created this flag so we had this flag with three colours and country called Czechoslovakia, the history went on, the country was called republic of Czechoslovakia, there was the second world war, many things changed but after the war we lost part of it but it was still okay, there was still Slovak and we still call it Czechoslovakia, it was called the socialist republic and Czech Slovak federal republic and in 1992 it was called Czech and Slovak federal republic, what happened in 1992, it was kind of a problem, what happened was this. We lost this part and it was, you know, I know... although pro protected the flag, you know, we promised not to but we still kept, it was really tough to pretend this is Czechoslovakia, in 1993 this country called the Czech Republic was founded but you know after 30 years, we still celebrate, after yeah, after 30 years of nonexistence of that country celebrates the formation of which is really weird but we do that.
So welcome to Czech Republic. We are really fascinated by non‑existent things but today is one real thing, the hospitality of the country, I hope you enjoy it as well and I hope you enjoy some of the social events we decided to sponsor for you to so welcome to Prague and enjoy the meeting.
(APPLAUSE.)
MIRJAM KUHNE: Thank you for the history lessons, that was really interesting, Ondrej. Next up we have Hans Petter Holen with an update from the RIPE NCC and details about this meeting.
HANS PETTER HOLLEN: Thank you very much. I am Hans Petter Holen, the managing director of the RIPE NCC, your favourite regional internet registry. Also the network co‑ordination centre, and the secretariat for the community, that's why we do these meetings.
So as you should know by now but I'll repeat it in case you don't know, the RIPE NCC was formed by the community back in 1992 as a project and later on as a membership organisation.
And we have staff and an elected Executive Board that oversees me and the staff and we provide services and implements to policies that are set by the community, the community is open to all and it's for coordinating technical matters and other matters between you who operate or care about internet in our service region and the whole world.
So logistics: I am here to talk about Meetecho, if you want to participate in a meeting digitally, you can use Meetecho for that, you need to register for that and you can book the unique login link you will get by email and you will need it for polls and asking questions in the Q and A unless you go to the microphone, but to get in the queue it's useful to have that link so you can raise your hand and get priority in the queue. There's also a live stream page, so if you want to watch this without logging in and without the ability to participate, then you can do that.
And all the sessions will be recorded and you can then watch them again and again and again when you come home to really focus on the great technical content that we have here.
So, I already said there is an audio and video queue, so you can click on those items to get in there, especially if you are a remote participant, that's your way, there is a Q and A session, if you want to ask questions, please enter them there, there's a chat where you can chat with people that are in the chat group but please don't ask the questions there because the chairs won't pay attention to the chat group.
And there is also stenography that makes sure that you can read what I think I'm saying and sometimes what I should have been saying and sometimes you start laughing and I have no idea because they make a joke on my behalf.
Miriam already mentioned that the badge contains useful information, please wear it at all times so that we know that you are actually have paid for this meeting because even though we have sponsors and the RIPE NCC, we are dependent on your contributions. You can also use this to connect to the wifi.
We do have an official photographer here. If you do not want your picture taken, you can get a yellow lanyard and then the photographer will do his best to avoid taking pictures of you, and I would ask everybody else also to bear that in mind when you take pictures of each other when somebody has a yellow laniard, please don't take a picture of them and at least don't publish it on the internet.
The RIPE NCC, we have a meeting organisation team, we have a tech team here at the back of the room, we have a support desk which is straight out and straight ahead. If you need help with anything. And you can also make appointments to do register checks where we will review your registration data with you, please speak to my colleagues in any area of interest you have with our services.
You can see here on the agenda we have speakers in a lot of the sessions, Emile will talk about the MAT Working Group, we have Angela and Marco presenting in Address Policy and Vesna and Anand in DNS and myself and Hisham in the services working group and on Thursday, Dave, Edward and Tim in routing and Rob will give the most exciting during the whole week which is the technical presentation in the closing plenary.
Then please participate in our feedback sessions. Please help us collect feedback regarding on training and certification, Pedro Antonella and Fallon are here and they would love to talk to you so we can improve our services and that's what I had. Thank you.
(APPLAUSE.)
.
MIRJAM KUHNE: Thank you Hans Petter. All right, next in line, we are kind of making this quick today is Ulka Athele from the RIPE NCC to go through the process of the elections for the NRO AC, I am not going to try to say what it stands for.
ULKA ATHELE: Thank you, Miriam. Good afternoon everybody. My name is Ulka Athele, I am a part of the communications team at the RIPE NCC and we are organising the Number Resource Organisation Number Council which sorts the address supporting organisation, address council election at RIPE 89. The NRO NC consists of 15 representatives from the communities, we have three representatives from the RIPE community, Herve Clement, appointed by the RIPE NCC Executive Board; Constanz Berger who was elected last year and at RIPE 89 there's one seat occupied by Sander Steffan, his term would have ended in December this year but he stepped down shortly before the end of his term in April ahead of the RIPE NCC Executive Board election. Thank you Sander for having served on the NRO NC.
(APPLAUSE.)
We will be filling a seat for the three year term starting in January ending in December 2027. We have two nominees. Jordi and Andrei, both familiar faces in the community. You can find all the information about the nominees including candidate videos, their bios and statements of motivation on the RIPE NCC website.
In order to be an eligible to vote, you must have registered for RIPE 89. You need to have selected the option to vote in the NRO NC election, there's an additional eligibility criteria, you are only eligible to vote in this election if you have attended one out of the previous eight RIPE meetings which means any RIPE meeting from RIPE 81 to RIPE 88. And finally, you need to have checked in for RIPE 89 even if you are attending online ahead of the deadline. If you forgotten to register to vote, if you are on site, please go to the registration desk and ask them to register you to vote. If you are online, go to your personal options page using your unique link in your email, select the option to vote while you are on a that page, please check yourself in.
You can check whether you have been checked in if you see a green tick mark next to your name, this is what it looks like for me on the mobile view. If you are on site, you should have been checked in when you picked up your badge, please just double check your name in case, if you don't see the green tick mark next to your name, please go to the desk and ask to be checked in, if you are online, you need to check yourself in.
If you are not eligible to vote, we will send you an email, if you think you should be allowed to vote, there's a form you can ask for a review and there's a deadline to ask for the review.
The deadline to register to vote and to check in is tomorrow at 2 U TC plus one, that's Tuesday, 29th OK, voting will open tomorrow at 5:00 UTC plus one and voting will close on Friday 1st November at 9:00.
How to vote, I have 30 seconds.
You will receive two emails from assembly voting. Please make your you can receive emails from assembly voting, one will contain your personalised voting link, click on vote now, the other email contains your unique pin code, copy your pin code from the second email, paste it into the platform, if you have a very over‑enthusiastic password manager, you might need to copy both codes, I just need a few more seconds. Go to the ballot, select the candidate of your choice, you can choose to abstain, if you abstain your vote does not count. Click on continue, submit your vote, you can also check your vote and a reminder of all the deadlines again and if you have any questions or you need any help, please contact us on nominations [at] ripe [dot] net. Thank you.
(APPLAUSE.)
MIRJAM KUHNE: Thanks so much Ulka for walking us through that process, it's an important election and I hope many of you will participate and now I would like to hand over to Max Stucci and Valerie Aurora, they are both going to present to you some information about and I am handing over the plenary to the PC to chair and from then on on, enjoy.
MASSIMILIANO STUCCI: Hello, good afternoon everybody, I am here as the Programme Committee chair to introduce you to what we do. And show you what we have done in the last six months. So first of all, what do we do. We are a team of 12 people, you will see the pictures in a moment. And we do four main things, we compile the programme for the plenary so today, tomorrow and Friday. We prepare the BoFs and the tutorials, if you liked something or didn't like something this morning, part of it is our fault.
We prepare the agenda so we define when talks are going to happen, there's a magic recipe that Jan was sitting here in the front has and we all don't know what he is doing withtI am joking, there is some preparation that's needed there. Then, it's magic. And then this is the part that we concentrated a bit more time this time, it's work with the submitters to adjust or improve their presentations, I will talk about it in a moment and then last you see here us, we are chairing the plenary sessions. So who are we? You have seen already our pictures, these are slightly different, but it's still the 12 of us which comprise a chair, a couple of vice chairs, Jan and Brian, and we have a local host representative which you saw already on stage. We have a representative from MENOG and Jan is also acting as the representative from SE.
The rest of the committee is elected and there are elections. So what are the stats? We see a little bit of fluctuation every time on the type of submissions, and you can see that we are always, at least for plenary talks and lightning talks, we are always roughly at 50% of acceptance rate, so we are going to keep our work as it is because we see that we have a good number of submissions all the time, but what has changed from RIPE 88? First of all, we decided to start, when we reject a talk, to include the reasons why we rejecting it. Because we felt we needed to do it. We felt we owed is to whoever spent time trying to submit a talk, and there is a secondary reason here, that is if and this can help for whoever would like to submit for RIPE 90, which will be in a great location, so please do. If you submit after the first deadline and we reject your talk and explain why we have rejected it, you still have time to readapt it for think about the comments we gave you and maybe resubmit and we will take that into consideration.
We also have a new document, RIPE‑829 and we have you have date submission guidelines, the RIPE Programme Committee charter, there are some small changes which include the removed reference to the ENOG representative, a better representative for roles inside the PC and potential financial support availability, there are other smaller changes but I didn't think to include them here.
Then we have updated the submission guidelines and here we have Valerie to tell you about them because she did all the work so I didn't feel it was my intent to present them so.
VALERIE AURORA: Thank you. I led the work with the rest of the Programme Committee to update the submission guidelines, we noticed a trend in submissions that was a little different than recently. The main purpose of the updates is to answer this question: Will using AI make your submission more likely to be accepted? And the answer is generally no with some exceptions.
So slight details. AI is very useful as spell check, grammar check, many use it successfully to requite a draft, it doesn't help to write the final text, it tends to be long and boring but both of these things are possible IP issues, talk to a lawyer if you need to, and graphic design, it is always difficult to read. I would be happy to chat more if you would like to know more. Thanks.
MASSIMILIANO STUCCI: Thank you, Valerie. So a little mention about lightning talks. We have a couple of slots open for Friday, if you have any idea, talk to us or go and submit. Please include some slides even if they are basic, just give us an idea of how you would like to present not only what you would like to present.
Then last, it was already mentioned, we have elections for the Programme Committee. Two seats are up for election, there's Moin and Antonio Prado who are going to step down and you can nominate yourself until Tuesday, tomorrow, and voting takes place until Thursday.
And the results will be announced during the closing plenary. Would you like to know more? Talk to the PC committee members and check out the RIPE charter which is now RIPE‑829 and e‑mail us if you want any other information. With this, we might have any questions? But if we don't, then and we don't have time because we were five seconds from the end and Jan stopped the counter, now we have a great presentation which will be unfortunately remote, the second time in a row that Geoff is going to present remotely. Not remotely, with he have a video of his presentation and he will be online to take any questions, we have a presentation from Geoff Huston about the failures of the internet to network security. I guess in a moment you will see the video and it will start and thank you very much.
(APPLAUSE.)
GEOFF HUSTON: Good morning everyone, I am Geoff Huston, I am with APNIC, I am so sorry I am not with you today, it seems that me and my bike parted company which would normally be fine because I didn't get just get on the stationary bike, I was ejected from a bike travelling at some speed and the result of bumping and grinding and so on caused some real damage to my poor body and when I said to the doctor I think I am travelling next week, he goes ah, you are joking! So... here I am. There you are. The good news is I am feeling a little bit better every day, healing, but unfortunately I am not able to give this talk in person. So you are going to have to put up with virtual me. But hopefully the topic is kind of interesting and provocative in some ways an gets you thinking about some of these issues so let's look at it.
I have been doing a lot of work in measurement over the years and part of the reason is because nothing seems to happen the way we thought. You know, some of these things, and you think a lot of the areas of security would be a no‑brainer, surely a trustable and trusted communications infrastructure is better than not. You know, security is one of the things we rely upon in terms of interactions at a distance where there's no other Merc Nim to mechanism to understand about authenticity and the internet is plagued by all forms of ‑‑ plagued by all forms of abuse and fraud and crime, somehow the basic elements of security seem to be resistant. I am like why is it taking almost 30 years for DNSSEC to got nowhere? 10% of the main labs aside, 30 years for one third of users to do validation, that's a huge amount of time, if it's taken that long to get that far, how long will it take for a useful level of deployment that actually means you can count on the DNS and rely on authenticity of the answers you get from it? That's a hard question. Really hard.
What about routing security? It's looking much the same. This is a long effort of 30 years to try figure out the right framework for doing routing security, and yet we are still a long, long way away from saying this is good. When you launch a packet out there on to the network with an IP address, how do you know it gets to the right destination? The answer is you don't. Surely that's not right. Surely we should be doing more about this.
Now, what's failing us is, is it you and I are lousy at our jobs, we should retire and go and do something else? Or is there something else out there that despite our efforts getting common adoption of infrastructure is just being resisted, and I suspect that the real answer is yes, these are instances of market failure because nobody is in charge of the internet, no one is telling us what we do. It's all a case of responding, each of us individually, to the pressures we see of a common market place that everybody is trying to optimise themselves and supply goods and services at a price that consumers want.
And somehow the signals in that market are sometimes perverted and what you would think is common interest actually doesn't work, that folk don't understand or don't get that signal. So, is this area where security is just failing to gain traction in the market and yes, of course it is by definition, and let's all go off next door. But no, not so fast. Perhaps we could flesh out the concept a little bit and try understand why we have to push so hard to get folk to do routing security and DNSSEC, even V6 is the effort of a massive multi‑‑decade push, why does it need a push, surely, surely it should just happen. So let's talk about some terms, what are the implications and maybe a few thoughts how we can improve that picture.
What do I mean by infrastructure? I am talking about the infrastructure of our common framework that defines the internet on names, addressing and routing, and security is really about the uniqueness and the authenticity of use, that this is my address, it's not yours, you can check the authenticity so we talk about integrity of those recommends and security is about how to protect that. What's market failure, invtesipaedia says its inefficiencies, irrational behaviour don't lead to outcomes for everybody which is kind of what I'm saying and the public good is not excludable, my use of it doesn't stop your using it and non‑rival Russ, I can't use all and neither can you, we can all use it to go, it's a public good.
So what's the issue? Well there's no one in charge of the internet as I said, it's largely a deregulated environment, totally dominated by market pressures and a certain amount of regulatory blinking down on the edges but nevertheless in general, there's no forcing functions to ensure that everybody adopts a particular mechanism or technology. We talk about security, nobody is forcing you to do DNSSEC, it's your call. Whether you generate ROAs and so on, your call. And unless it's the common adoption of these things, it is actually not because everybody was being told who must do this, it's because we all see some perception of individual benefit in doing individually, creating a collective common good that everybody does it. Market forces, not regulatory impost.
The problem is that that kind of picture sort of assume some degree of homogeneity in the market, everybody reaction acts to the pressure and perceives it in a similar way. If we see it differently, then that's much harder to sustain. If only a few week perceive that the benefit is worth the cost, only a few folk are going to do. All of the markets on the internet do tend to be heterogenous, they tend to differ in perception. Why isn't everybody everybody on V6? There's a different benefit of cost of adoption across the internet and that leads to variance in outcomes because the different individual perceptions and so what happens here is there's no clear immediate and common judgment of the benefit of doing anything. The internet is not a vastly uniform place, it's actually a place that has a fair amount of common perceptions of difference.
So the next part of this is there's lots of moving parts in a network. There's the upper level of service and content. There's the level of infrastructure in terms of hosts applications and services. And then at the lower levels, there's kind of the ISP and moving packets around. And they all see the world through different lenses and if you look at the ISP level and go well, if I do RPKI route origin validation, improved the picture for my customers, surely, can I charge them more, no. Similarly if you are a DNS hosting provider, can I charge more for DNSSEC, in general when folk have tried that, the answer is customers don't adopt it, they say no and go where else. The experience was the only way DNS was not large amounts of uptake was by discount, which by creating a reduced cost to create the incentive.
So it's very hard to monetise investments in security individually, and that becomes the disincentive. We see this in the data, when we look at the proportion of routing objects not covered by a ROA and you could see a few years ago, it was 50% of V4. This stuff is going very, very slowly.
How long is it going to take? I don't know, but it is sometime and the net result across four years of looking at this, looking at the proportion of users who cannot get to a destination if it's invalid in terms of a route origination validation. Invalid. It's been about 20% now for some years, we are not just making head way in that particular area of security. So all of that effort with ROAs and do the right stuff isn't actually making much traction on the larger problem, we are not making progress here. With DNSSEC, the amount of validation, the inverse, the proportion of users, 90% ten years ago. Even today, 70% of users will happily go somewhere if it's badly signed, don't care, because the recursive resolver infrastructure doesn't care.
The picture isn't good, and even if you look at what we do inside the regional internet registries, do we sign our registry? Do we actually say here's the answer and here's why you can trust it? Here's the authenticity of that answer in terms of the digital signature?
None of us do. Oh, it's too hard. And the perception of benefit is so low, it doesn't out weigh the incremental cost of doing it, you take the easy way out.
Now, what does this mean by designing an application? The lack of infrastructure security everywhere, the picture is either with DNS queries and application, I can't trust the answer. Will I generate a packet and send it somewhere, I don't know if the packet reached the intended destination, I can't trust the answer. The fact that you sent a packet to somewhere doesn't mean the answer is good. These are problems for the application, and the safest course with partial adoption is to assume there's no adoption. If it's only partially there, you don't know which parts are secure and which parts aren't, so you have to assume none of them is secured, none of it is reliable because there's no way you can do it any other way, yes, you just can't count on universal availability. And so what do you have to do? You have to do it yourself. You have to make sure that the application itself is making its own efforts to ensure the validity of the destination, whatever you are trying to do, and do that without relying on infrastructure.
That's most visible in the uptake of TLS. This is a report looking at the availability of the use of HTTPS in the web from a cloud based perspective, that's not 30%, it's not 70%, it's 97% and it happened without a massive effort. The application you have got to do, a little bit of prompting from the browseers an it's ubiquitous, application level solutions are getting a far better benefit of deployment than infrastructure.
So when an application is using TLSes which they do these days, what value do they place on infrastructure security when I am alreadying TLS as DNSSEC adding to the picture? No.. how about RPKI? I don't care which address I got to, if we can prove its authenticity by TLS, I will trust the answer no matter what, what's the value of RPKI? No. So applications don't value infrastructural levels of security, particularly when they are only partially adopted because they can't rely on it.
I think it's part of the bigger picture. What we are seeing on the internet is a shift upwards of value. In the 1990s, when things were if you will, scarce, the network was actually rationing the scarce resource of comms, that you couldn't afford to replicate your content on servers all over the planet, you placed them at one point on the network and used the network to bring users to you because you couldn't afford to replicate.
Then that created a value proposition where the network was the arbiter, the rational, the scarce resource and actually all the value of investment was inside moving things around inside the network. The value transformers moved up the stack. The inexorable pressure from Moore's law to a computing cheaper process and cheaper storage, oddly enough the use of more capable dignal processes was creating cheaper comes, we are talking about... for the same fiber cost as what used to be hundreds of megabits, we are seeing a world of abundance down in the network layers, at the infrastructure levels, complete abundance which has pushed down in value in terms of the commodities and the true value used the upper levels of protocol stack, and so now applications are taking on many of the roles that we have thought in the '90s should have been done at the common infrastructure level. The applications going, I have got all the money, I don't need a common infrastructure solution, I will just build my own. And so the lower layers were commoditised and stripped of intrinsic value. And oddly enough, it's created a weird relationship within the stack, that all of a sudden hosts and networks have some level of distrust with applications. And so what we are finding for example in the Apple ecosystem, Apple produces Apple private relay, a system that completely obscures everything that the device is doing in terms of application behaviour, obscures that from all the lower layers of the network.
They can see nothing.
And so what we are now seeing is applications kind of using techniques such as QUIC and MASQUE to wrap up their service interaction, to wrap up their activities, that actually increase the distrust and suspicion, no one can see what's going on across elements of the stack and what that means is that if I don't trust the network, then if it does these routing security DNSSEC and so on, if I don't trust the infrastructure, I place no value on whether it does infrastructure security or not.
If I place no value on it, I am not going to fund it, and that leaves the open question, well who is going to be left to fund it? If I believe the network is untrustable, it's hard to then say well I have done some measures to make my part of the network trustable, please pay me. It's well, sorry, no.
So when you say well what about more money for DNSSEC? The real issue is well who is going to pay for it? Where would the funding be directed even if you could find some money and who would make use of that? And the real issue is, even if such infrastructure funding might exist, it's very hard to identify where and why you would place effort when applications, the service level, is kind of going nah, I am going to see it doesn't work properly and have my own mechanism to ensure the authenticity of the service of the application. And a similar picture exists with BGP and routing, the exact same set of questions, why does anyone care any more and the answer is well they don't and because of that, we are providing our own solutions, the solution from the content distribution network is actually to not routee. Once you have deployed CDN with a few thousand points of presence, every single access network is adjacent to one of your points of presence. There's no routing anymore and so when you say to a very large CDN, what's the value placed on routing security? And the answer is I care more about the authenticity of the advertisements I see from my directly attached clients. But I really don't care about the rest of it. I don't care about the security of the routing system because I don't route. I don't move traffic through large amounts of third party networks, it's me and the adjacent access network all the time. And so infrastructure tends to be marginalised and, quite frankly, there's underinvestment because there's no point.
Now I am painting a pretty extreme picture here, maybe there's another way of looking at this. And that is can applications do it better? How can you improve the assurance of the authenticity of the services? And certainly there are points. TLS has its problems, glaring problems. Can we use something else to actually pin down TLS because that whole network of third party certificate authorities has caused us problems in the past and will no doubt cause us problems in the future.
Are there are other approaches that make things better? Well, interesting. Because you see we are constantly evolving on the internet and partly this is the inexorable pressure of Moore's Law. That computing processing storage is getting ever cheaper per unit and what that means is we are actually making a network infrastructure that is more capable at the application level. We no longer have to do one thing at the bottom and have everyone trust it. Common infrastructure security mechanisms reflect I think a picture of networking and it's a few decades old.
And, quite frankly, the technology shift is lift everything up and actually allow end points and applications and services solve the problems directly. When we think of securities being an ISP problem, a market failure at the bottom level and it requires regulation, then maybe that creates regressive imposts, these aren't helpful anymore. So when the FCC proudly pronounces a few months ago in May we are going to start looking at infrastructure routing security, maybe that's a picture that's a few decades old, maybe that's just living in a past that isn't relevant any more. We don't do it at that level. Applications are now capable enough and have enough resources through abundant computing and processing and storage that's been solved in a different level.
And it's kind of always be aware of regress I have impost that reflect the past that the markets have long since moved as past, the realities of the economics of networking have moved on.
So what should we do? Persist? What we have been doing for 30 years has become a habit, let's keep on doing it. What do we need to do to make everybody doing ROAs, get everything happening where it's being failing for the past 30 years? I think pressing on is almost like anti‑gravity, it's pushing deeper and harder into something that's structurally flawed that doesn't reflect the realities of the situation in which we live.
Because realistically now, the money and the control quality, the control of authenticity, the control of the user experience has moved up the stack into the application space. How can we help these application transactions create the necessary assurance to give the users of the authenticity of a value in transaction? How do we create that level of assurance without asking the infrastructure to go and do things it has been failing to do for decades already? I will leave you with that challenge. I don't know what the answer is. But nevertheless, I think it's the question of our time. If we can answer that question, I think we find ourselves working in a more harmonious relationship with the true economics of the internet itself.
Hopefully I have left some time to answer questions here that you might have about this, thank you very much.
(APPLAUSE.)
JAN ZORZ: A question online? Geoff is online.
GEOFF HUSTON: Hi.
JAN ZORZ: We can hear you.
GEOFF HUSTON: You stole all of my photons... bring them back, it's dark here.
JAN ZORZ: Okay, please, Rudiger was in the actual queue.
RUDIGER VOLK: Hi, I am seeing one point where the infrastructure actually counts and and I think that got lost in your view of things. In the good old days, people were saying that really important things about the internet were first, universal connectivity and providing secondly permissionless innovation. Well okay, the universal connectivity I wonder is kind of really depending on the infrastructure. Kind of yes having the end points verify each other's identity is fine but if the connectivity does not help to actually transport the packets from point A to B and back, kind of well okay, nothing happens, and well your explanation that only the connectivity from the CDN to the end user counts, of course means there is no universal connectivity.
GEOFF HUSTON: Yes, that's very true. 90% of all traffic comes from somewhere down the road, the local data centre to you, oddly enough to make the network faster and cheaper, and cheap is the big leverage here because of Moore's Law, it's cheaper to replicate content to deliver it everywhere in the background and on demand push it through the last mile, it's been easier, simpler and cheaper to change the entire architecture of the internet than it's been to actually make the infrastructure do things that quite frankly we can't get them to do at this scale so yes, there has been a massive shift and the reason why is quite simple: Follow the money. And today's network is certainly amazingly cheap and amazingly large, but at the same time it's not really a network any more, it's just a last mile delivery system. Thanks.
JAN ZORZ: I am closing queue, in the interests of time, we are way behind the schedule, please be briefer than the previous...
JEN LINKOVA: . There is no power and devices are dead, anyway, Geoff, I kind of agree with you to but not completely to like 66.7% because as far as I can remember we traditionally define security as... and integrity and... and we cannot necessarily solve the application level, installation needs to be at a level if destination is not available... my second last point is do I understand you correctly that to promote security and make it be efficient, we need more attacks on infrastructure.
GEOFF HUSTON: Okay. The other way is let the applications do it themselves. Thanks.
JAN ZORZ: Daniel was in the Meetecho queue, do you want to go?
DANIEL KARRENBERG: My name is Daniel Karrenberg, almost retiring. I think that the flow, the way you present things Geoff is that the security, and I am putting it more to the point, the security on the network level has to make sure that the network level keeps functioning, if the network doesn't function, then all the application layers of security that you have is not useful, and I don't buy your argument that CDNs are everything, because I can think of quite a few applications that actually require a little bit more wide area of security. So, my conclusion from what I hear from you is let's focus on the security that keeps the network running at our level, as ISPs and as packet pushers, we should look at the security that keeps the network running and not breaking it and let's not try to solve the application layer security and all that kind of stuff. So I think your perspective is quite useful but your conclusions I don't really agree with.
GEOFF HUSTON: The problem we have, Daniel, is that a number of these issues that happen at our level, whether it's IPv6, wheter it is security in the DNS or security in the internet, it's now taking us more than three decades and the uptake is still lack lustre, slow, it seems we are pushing the rock up the hill all the time and you kind of wonder if it is better for everyone, security is one of these things that everybody needs and we all acknowledge that, why does it feel like we are pushing against the hill or against gravity, why isn't this being enthusiastically adopted and it's trying to find those reasons why there's resistance to a good thing that I'm trying to search for here, and my view is the money has moved out of the picture and doing things on the cheap is always hard but you know, there might be other reasons as well but always pushing is the problem. Why is this so difficult?
DANIEL KARRENBERG: Because the network still works.
JAN ZORZ: We need to speed up.
[TOM]: Tom from Cloudflare. Thank you very much for being up this late, I am well aware it's past midnight at this point, thank you, I object to a couple of things, first of all I do think the application can do things independently from the network, the best example of this was a Crypto hijack that happened in Korea last year which included TLS and everything, due to the RPKI invalid hijack, they were capable of siphoning quite a few Bitcoin away from people so the application is still very much independent from the network and that's why we need to secure the network, I think an important point to make as well is that yes, from a purely numbers perspective of how many prefixes are assigned, we are not making the progress we'd like to see but if we look at the traffic we are doing, we see significant uptake, and I think that's a very important component here, infrastructure is hard and we should applaud the work we have been doing so far and should get better at doing what we are doing but it's important to measure the right thing and just absolute numbers is probably not the right thing to measure here.
GEOFF HUSTON: In the interests of time, yes.
SPEAKER: I will keep it brief, the reason we should keep pushing for security is resiliency is the king of the game here and you can't have it unless you try to enforce it in the entire stack and trying to focus on one area I think is a big fallacy that will lead to disaster, thank you.
GEOFF HUSTON: Thank you.
JAN ZORZ: Well thank you, Geoff, very much. Get well soon!
(APPLAUSE.)
So where's Fredy, did you ever ask yourself what can possibly go wrong with FTTH? ? Let's ask Freddy. Please. The stage is yours.
FREDY KUENZLER: Thank you, it's awkward to talk after a rock star because usually when you share a stage with a rock star, you are before the rock star, no. ... bear with me please, don't don't throw any tomatoes or axes. So we talk about FTTH today. By the way I am Fredy Kuenzler Keunzler and I think my first RIPE was 54 so I'm an old man.
Swiss telecom market quick reminder, we have in Switzerland we have three major players, thermobile players and they have the largest market share on broadband so this is probably like anywhere else in Europe, just to give you a little context, quick word about Init7, we are almost 25 now and we are in the... a good place to visit. We operate some infrastructure FTTH infrastructure along with our backbone once, AS 3030, if you add up the number it becomes 7, the lower the number the faster the network and yes, we operate as of today about 320 FTTH pops, we are buying layer one fiber from the incumbent utility, this is the marketing slide that's supposed to show by my marketing team, we got also a number of medals.
What are we talking about? We are talking about a few people or a few institutions in Switzerland, so quick crash course in Swiss politics. As we are a democracy, we have executive, this is called Bundesroute in German or federal council, we have legislative with two chambers, the national Parliament and of course there are some courts as well.
The one we are talking is the highest one, the other one is federal administrative court. These two are important of course, there's also some administration, the federal department of environment, transport and energy and communications is caring about our industry, DETEC and current bounds route, this is his name and if you are reminded to some speciality, Rossity is actually a food.
We also have an OFCOM, a federal office of communications, that's the Bernard mice Sen and of course we know each other but not from the industry but from soccer which is nice.
Then we have the federal communications Commission, this is ComCom. The current president is one of us, it's Christian Martin who used to be at Cisco and Google. The legal basis of that is the telecommunications law. If you are already bored, this is important to understand the context. I am sorry about that.
We have a Competition Commission, this is ComCo, don't mix it up with ComCom. So it's not my fault, in Germany it's actually... (Speaks German) it's a bit easier to distinguish. The current president is Laura Baudenbecher and she's very much in this antitrust law.
So please distinguish between those two. This was the cost and now we are talking about some telecommunications history. You all know the history, about 25 years ago or a bit earlier in the late 1990s, most of the incumbents, they became kind of independent, the PTTs, the post telephone telegraph, they were split into post services and telecommunications, and I think there are some youngsters which probably never heard of that so this is where we are coming from.
And at the time a lot of competing telcos tried to enter the market to provide better or cheaper or both services in telephony and the internet.
Some failed. So this was just a rant. Some of you might remember those logos and some are still around. I yes, a lot of promise were made, a lot of big money fantasies and some went bankrupt. At the time, because there was an only incumbent which had a hundred percent market share at the time in any telecommunication, especially telephony service, they need to be regulated to at least grant some halfway fair market situation for the new entrants. This happened with telecommunications law in Germany, it's called (Speaks German) in Switzerland it's a bit awkward name but there we go. And at the time these unbundling these lines to the corporate line subscriber, they make useful for other services than just telephony, and a lot of alternative telcos installed these DSLAM and DSL aggregation boxes to provide services that were fine up to about 25 megabits and in Switzerland the unbundling peak was reached around 2012 with 300 K subscribers, which were really unbundled to put that into context at the time it was probably about 8, 10% of all the households.
Later on, DSL evolved and the G.Fast came up and etc, etc, you know that they required higher bandwidth required that there was not a technical unbundling possible. So we had ‑‑ the requirement was one box per copper bundle, so the unbundling was kind of by the technical development no longer possible.
So, some came up with so‑called bitstream excess which is, in Germany it's called Layer 2 BSA, I stole one of these nice pictures off Deutsch Telecom, sorry for that, it was basically one box with a VLAN provider A and provider B and a third VLAN for provider C to access their customers directly.
But in Switzerland, we unfortunately didn't, we never got bitstream because it required a law change and it failed in 2007 that we could get bitstream access, it was actually meant as a temporary measure only but it wasn't really introduced.
Copper U LL remained regulated but became more and more useless because only of the bandwidth, only about 25 megabits and that can't compete these days any more. So the 300,000 subscribers which were really unbundled, they declined over the year and there are only a few thousand left as of today.
But there was light at the end of the tunnel, we got fiber, we got FTTH installed around 2006, mostly in urban areas. So for example in the city of Zurich, there was also a lot of public money invested. By the utility of Zurich, they started to build a lot of fiber, as of today we have the city of Zurich has around 280,000 FTTH connections built, we are well above two million as of today. If you put that in context, it's really the fiber activated the fiber, the homes connected and not just like in germs knee where they always say we have a lot of fiber build but it's actually just passing the building and there's no real connection made.
So to make that note; every builder, every FTTH builder invents the wheel from scratch. OFCOM invited to round table and take out some standardisation which was written in the couple of years after 2008 and they were like a standard, a number of scream etc which makes it really conveniently states to provision services.
And the most important thing probably is the 4‑fiber per household, each OTO box have four fibres and they could get for to do all services an it's open access layer one so everybody can access that, it's a point‑to‑point network topology.
2019 there was another revision of the telecommunications law and that's the problem or the discrepancy in the Swiss Parliament you have, on one hand you have the idea of a free open market. In telecommunications, on the other hand there is always a Switzerland still owns 51% of the incumbent, they always want to get this dividend of currently about 580 million Swiss Franks or maybe 600 and something million euro every year and no one wants to touch that money flowing into the federal fund.
So that the technology agnostic regulation was actually proposed but has not been adopted by the national Parliament.
And when regulation failed again, the incumbent thought yes that's a good moment to monopolise the network. So they announced another 1.5 million FTTH to households until 2025 but they more or less silently changed the network topology from point to point to multi topology and got rid of the point to point point by granting the open access to everyone.
Later that was actually considered as a legal obligation by I will come to that in a minute.
So, what happened. It's the so‑called Glasfaserstreit and if you, it's the fiber dispute, if you Google or the hashtag, you find a lot of stuff I wrote in the past couple of years.
That kept me busy for a long time. Because after, about three weeks after the announcement of Swisscom doing point to multi‑‑point, ComCo came on the stage and started a preliminary investigation which is now named as an Netzbaustrategie network strategy and with the number. Swisscomis actually quite a good client of ComCo because there were several cartel antitrust cases, for example they also tried to margin squeeze broadband, and this is not new to you guys because it happened all over Europe. So, about half a year later we filed a complaint when we noticed that our customer couldn't connect to our own infrastructure any more but we had to buy Layer 2 services from the incumbent and then we asked, we filed and we asked ComCo to investigate, but what we didn't know at the time, the investigation was already running so we probably just gave a bit more spin to the wheel.
Because only three months later there was a precautionary measure of ComCo happening so Swisscom could actually, it was forbidden for them to continue with the point to multi‑point network infrastructure to roll out this further because if they feared that Swisscom could actually, why this procedure of ComCo was going on, they could actually do a fait accompli and that was forbidden. But of course Swisscom appealed against that, to the federal administrative court, as I said before, and there was then later on a hearing and this was quite an experience, me sitting in the middle with my lawyer, just me and beside there was ComCo and on the other side Swisscom and those two fighting, so that was quite an experience.
So, but as a matter of fact, the incumbent ignored the order of ComCo, they just continued to build the fiber, as they wanted a point to multi‑point structure. In total we don't know exactly the number but this is my assumption based on a lot of facts, there were about 750,000 plots or households built the wrong way, the illegal way.
And then later in 2022, there was our great day, and I have to confess we opened the bottle, it was a good moment for us and for the market and for the telecom market as a general, we won, and there were about 220 pages, it was 100% victory for ComCo and us. That was in a sense remarkable because in this decision of the court it says they did it on purpose and it's all their own fault if they have to fix it now.
So they stopped two markets, some ready‑for‑market FTTH which gave some problems also by the end cost, they wanted ‑‑ they were happy to get fiber and the day after they couldn't get fiber any more, that was a bit sad but, yeah. It wasn't their fault.
The decision was about one year later confirmed by the federal court and while this precautionary measure were finally decided, the main proceedings is still going on. We got a decision earlier this year, it was announced in April 2024, we won again, another 200 pages, so I am now really used to all this paperwork. And they got a fine of 18.4 million what they have to pay, but of course they appealed again and it's now again at the federal administrative court, pending.
So this is what I already said, around 750,000 FTTH are illegal, otherwise they cannot market them any more. The so‑called feed or clean up project is the name, this is on the way as of today there are some 300,000 plus which still needs to be converted, a lot of which are already fixed.
So what we can say is point to multi‑point is history as of today and forever because I mean as you will know, fiber infrastructure will last for decades, and so we won this case for the sake of the Swiss economy and end users.
So that gives some opportunities for everybody, and of course we try to use these opportunities, most of the other service providers do XGS‑PON on today based on point‑to‑point, they still have splitters but the splitters, they are installed in the central offices so point to multi‑point topology works on point‑to‑point but not the other way around.
Ourselves, we and some other service providers, we do more internet stuff and not XGS‑PON despite that XGS‑PON is certainly a lot cheaper by one subscriber, and we also use this latest Cisco gear. It's used more more like data centre things but we are quite happy with it. There was, I think two years back, a presentation giving some insights about our gear but what we, our experience, what we made.
And new opportunities, we have now a new product called Ethernet7, it's actually based on VDNP XLAN which is normally used in data centres, we use it all over Switzerland, it carries internet service, it's available for quite a while from the incumbent but there was never competition and here you see the effect of the so‑called Glasfaserstreit of the fiber dispute, we achieved, we can compete them now in a productive field which was without competition and we can sell it a lot cheaper and better because as of today, ten megabits, a plot with ten megabits is still around 310 Swiss francs in rural areas, and then we thought well, we could do something cool for the community on our infrastructure, and it's called Exchange 7, this is probably the, it's still very experimental, we just installed this IXP and tried to manner and tried to set up an internet exchange (IXP manager) you can use your home to get connected, I said before more than two million OTOs are capable to connect and are certainly available also in most Swiss data centres to connect with this new exchange, we are not sure yet if it's really going to work out and if we keep it on the long run, but it's an experiment we can do on this infrastructure that I think, well some people like it. We also got the IXP allocation from RIPE NCC and we are happy about that so. If you like the idea and want to connect, talk to us or send an email.
Then there was also, if you want to further reading about the topic fiber optic dispute, we wrote quite a long article about 50 minutes to read, it's available in French, Italian, English and German, sorry no Czech but yeah...
That's the story with all the details. We also did a fiber map. I think it's also kind of cool where you... oops, what happened? Something happened. Can you read it. Okay. Well you can read it. Yeah, you can use that QR code if you want to try it out, you can zoom and see every house where fiber 7 or our product are available which buildings have already connected and then that's me and that's it.
Thank you.
(APPLAUSE.)
MASSIMILIANO STUCCI: I didn't want to put more pressure on you, I just stood up because I saw you were towards the end of the presentation. We don't have any questions? I didn't see any online. Okay. We have Daniel. Before the question, just so you know, to connect to what was presented earlier about the Czech holidays, in Switzerland every Sunday is level 3.
FREDY: Actually where I am, I am living in a new place for four months now and so this is a tower with 50 floors and I am on floor 12 and I take the elevator, step out and step in and then I am in a 24‑7migro, I can even shop at 2 o'clock in the morning but there's no alcohol. There's no alcohol.
MASSIMILIANO STUCCI: Because they cannot check IDs. Yes. Daniel.
DANIEL KARRENBERG: Yes, did you get any bad press when basically the incumbent customers couldn't get the fiber connected.
FREDY KUENZLER: Of course, of course, especially in the French speaking part, there was, the rumours spread, there's some silly provider in the German speaking something, these idiots, they keep you from connecting to the fiber, yes, and so that's a little sacrifice we had to bring.
DANIEL KARRENBERG: Okay, little question, how much revenue or turnover if you missed because of this?
FREDY KUENZLER: Well we were going to sue them because there's, I think a ticket on number 836 or whatever, about a thousand, it says please sue them, it doesn't say please sue them but you can understand it like them. Yes, a couple of millions probably over a couple of years, yeah. But I think in general the total cost of the incumbent is probably about billion but they will never confirm it of course.
MASSIMILIANO STUCCI: I don't think we have any questions online. Thank you very much.
(APPLAUSE.)
Now, before we close the session, two things: One is, please rate the talks to let the Programme Committee know if you like them or not. If there's any comment you have about them, so you help us decide for the next meetings and Jan has a message for you.
JAN ZORZ: Yes, with my hat of the NomCom chair, thank you Ondrej and the board for appointing me. Yes, that. So please volunteer for the nomination committee, we have to choose the RIPE chair and Vice Chair for the next five year term and if you have attended at least three out of five RIPE meetings, you are eligible to volunteer for the NomCom, there will be a selection that's completely random and I think after this meeting, we will announce who is in the NomCom so please I send and e‑mail to the community mailing list with all instructions how to apply, there is also link somewhere here, please, please, we need as many good volunteers as possible. Thank you.
MASSIMILIANO STUCCI: Enjoy the rest of your meeting.
(APPLAUSE.)
(COFFEE BREAK)
