 Hi, okay then. So my background, I probably joined as the least programmary person on the program because I used to be a programmer, but now I'm sort of a management type. So my projects didn't involve like a huge amount of writing deep level code. But when we started to, on the first day, sort of let's see what happened with Molly, like trying to get Lotus running by following the documentation, I realized that we ran into a whole bunch of problems. And when I, and so I made my project being trying to figure out how to get the documentation or the build targets, the whole pipeline all the way through to making it so that someone can get to make a deal in Lotus simply by following the instructions without having to go off to Google or anywhere else. Some of these fixes are actually relatively straightforward and just documentation updates or clarifications, but some of them are actually there's problems with the build targets, there's problems with things being targets not being built at all or being missing. So there's various installation pathways, but they all have issues. So W get very simply linked to, there was a sort of copy and paste use this version linked to a very, very, very old target that we shouldn't be using anymore. The app image was 50% of downloads until it stopped until it was disabled in November because there was a problem with it being built and the snap target needed a small correction and the homebrew target didn't work until Russell fixed it. So none of these targets could do the main thing which is getting someone to the point of being able to type Lotus client deal and have it work without extra steps or Googling. Only thing that worked really was building from source. Also, I was looking at the docs.file.io documentation not the Lotus.file.io and that's like a completely separate set of documentation that is similar and has a different set of instructions and a different set of problems. So like as the beginning, the issues were things like broken SIM links, missing steps, mismatched docs, missing and broken CI targets, stale links and a broken install script. This leads to some other questions which is how did it end up being as broken as this? Which pathways are the important ones for new users? How do we know that these pathways are succeeding for users, not only just the targets work correctly but they're the ones that are popular and how will we know if they break again? And so if you wanna track those things in aggregate, you really want to understand what's happening in stats all the way through from someone landing on Filecoin.io or even having someone at Southwest say, hey, you should check this out all the way through to them actually connecting to the network. So then I started going off and looking for stats. So I went looking for Google Analytics, took me a while to track down how to get access to Google Analytics. I actually only got access to the GA for docs.Filecoin.io yesterday, thanks to Jeff Everard, then download stats for the different artifacts. So you have to look at GitHub downloads, also GitHub clones and then Homebrew and Snap. Well, they both have, if you wanna look at the stats of how many people are using those things, you have to go to different places. Although we could actually track on the documentation pages, we could put in links to possibly to track outbound links to those things and see whether things are doing stuff like that. But actually the hardest part is, is this leading to people actually connecting to the network. So I sort of spoke to a few people, Steph gave me some help. We're like trying to look at whether we can get some stuff from Sentinel or the various dashboards and my impression was that we have a huge amount of data on miners and the things going on there, but we actually don't track in any reasonable way how to get to just someone's just made this process work and they're connected to the network for the first time and they've tried to do a deal in Lotus. I think I found the answer to that, which is that in all of the various documentation paths, the thing that people try and connect to is APIChain.Love and we direct people to that for their first experimentation. So even if they go on to do other things later, if we can look at either the gateway stats or the daemon stats for API.Chain.Love, I can probably get to a good first approximation of people successfully following this documentation and becoming a real live user on the network. The only problem with that is that I still haven't managed to get access to that data yet. I'm hoping that I will get it today or tomorrow. So it's been a really good project for exploring the different parts of protocol labs and different parts of follow kind and figuring out who's in Pathfinder and some of the stuff that we were alluding to yesterday, but I haven't got to the end of this, but I'm also not at the end of this. I am good to get this finished. So my next steps are once I have all this, getting to a point where I can track end to end what's going on through this, then get, oh, one of the other things is that there's currently no Google analytics on Lotus.FileCon.io, and I should talk to, I wanna say pronounce it incorrectly, Mayank or Mayank? Yeah, that's absolutely correct. Okay, I've realized that you've been looking at the Lotus.FileCon.io, we probably should have spoken earlier because we should try and get GA back on that. And then we can track the activity via the various pathways and onwards into actual engagement when they get there. And then I want to deploy the fixes in the documentation so that we can see the effects of what we've done. Then after that, we measure what happens and see whether we've made a difference. We continue to iterate on it. We try and apply the same techniques to some of the other clients and tools and apps that we have, and maybe even build a dashboard to track conversion through this funnel, ongoing, do it for different platforms. And I'd actually like to do some, and maybe this has happened already, I don't know, some actually observed usability testing of not only the documentation of the CLI tools and see if we took developers who've never used our tools before and followed them through, we could probably figure out a bunch of ways in which we could unblock their onboarding. That's it, you can go and look at my work in progress and I'll continue to update it. And I wanna thank Lex Luthor, Corey Schwartz, Ognots, Steph, Florian, Russell, Molly, Lindsay, Jeff Everard and everybody in the cohort because I had a great time, thank you.