 Hi everyone, this is Dennis from Problab. I try to speed run my tiny presentation. So we as Problab were running the nut hole punching month. So first of all, thanks to everyone who participated. So what were the project's goals? And so net hole punching in general, we want to have full connectivity among all nodes of a Lippie to Pee network despite nets and firewalls. And in general, this specific project, because we wanted to gather information that guide protocol optimizations or also some implementation details. And we got a lot of data for that. The measurement campaign ran from the 1st of December until officially until the 1st of January this year. I analyzed the data that we gathered until the 10th of January. So a little more, we had almost 300 API keys generated that could, in theory, have submitted data. But after some deduplication and so on, I'm pretty sure that with this 154 clients, so we had 154 clients deployed that have hole punched or punched 47,000 peers around the world and contributed 6.25 million data points that we are now analyzing or I have started analyzing. And on the right hand side, you can see where the clients were deployed and, sorry, and the remote peers resided. So we basically probe the whole network, so this is the whole world. So here we have the time on the x-axis, on the left y-axis, the success rate. And each dot is an individual, so the color of each dot is an individual network in which the client was deployed. And so basically what you can see here is that the success rate, if we probe the whole world or all the other peers in the network, we have a hole punching success rate of around 70%. And so I will get to this in a bit. But yeah, so this highly depends on the network conditions. But yeah, so we have around 70%. And the faint red line in the background is the number of data points that were submitted on each day across all the clients that we have deployed, which at peak were 35,000 and yet dropped after I have sent out the email that the whole punching campaign is done. So what were the insights so far? So I'm still working on the final reports. So the success rate is around 70%, but as I said, since we're probing the whole world, it depends on the network of clients as well. However, that's the success rate that it would experience right now if they used the whole punching protocol today to connect to a random peer, other peer in the network. Then some other insights. We expect it quick to be more successful than TCP. We couldn't verify this. So it seems to be that they are similarly successful with regards to hole punching. IPv4, IPv6, IPv6 has a much lower success rate. We're still not sure if this is a measurement detail or actually a problem in the implementation or the protocol. Interestingly, it's not round trip time dependent. So if you know the whole punching protocol a little bit better, we're synchronizing both peers and we thought that the round trip time is a big factor into place a big role in the success rate, which we couldn't verify. So that's also good that we're not aiming to optimize for that. Then peers, as expected, have a worse success rate in VPNs, but also we have interest in another protocol optimization for that lined up. And so this graph at the bottom is quite insightful. So we try the whole punch three times, but we found out that if a whole punch was successful, it was with 97.6% successful with the first attempt with the first try. This means, let's optimize the protocol to change strategy with the second or third attempt to increase the odds with the second and third attempt. So this is also another insight for that. Alyssa, the three, like there's another protocol improvement that we're proposing based on feedback from the foster talk that Max and I gave in Brussels last weekend. Also some implementation details issues kind of thing. So still discussing discussions going on. Around some weird data points that we have which could indicate some bugs or not. So we're still discussing that. And in terms of next step, I'm working on the final report. You can see the gif at the bottom right. I'm already started with that. There are many more graphs and so on, many things to look into and a lot of angles to look into the data. And yeah, feel free to have a read and this is the next step. Thanks a lot.