 Hey everyone, my name is Justin. I'm J Flory 7 and this is So this workshop presentation is mostly a quick dive into the community operations team And we're going to try to teach some of the metrics tools that we have available in the Fedora stack So we'll generally start with a quick introduction to com ops and then we'll dive into some of the metrics And we'll talk about some of our as we move closer to the end We'll start doing more of our our hack session topics that we have planned at the end So the metric session should be about 90 minutes after this But just as a and then the latter part will also be like a hack session for some of our tickets that we have to So as a quick introduction for comm ops, what is what is community operations? What is what is the comm ops team? So comm ops first came up around November 2015 and it was part of this idea that the Fedora project is huge And there's so many different corners of the project that it's almost impossible to keep up with all of it at once There's so many different things happening in the development side and the design team with the ambassadors It's almost impossible for any one person to know what's happening in Fedora at once So part of the vision for the community operations team was to bring heat and light to these different areas of the Fedora project So it's kind of if you can think of as marketing as kind of this outreach Community operations is maybe a little bit more like in reach So some of the things that we help work on is like the community blog Which is the platform that we use to help deliver Fedora news for Fedora contributors across the project You can go there and you can see different blog posts that have been written by different contributors across the project We also work with a lot of the metrics and data tools I think this year we also had some we've had some GSOC or Google summer of code slots in the past to work on some of these tools As well as some other projects as well But we kind of do like a mix there's two different sides to the community operations team one of which is more of working with the the quant or the quantitative tools like the metric stack like Fed message data gripper and some of the others and Sachin is more informed on those things than me And then on the other side, it's also a little bit more non-technical So we look at some of the other things going on in the community One of the current tasks that we're working on that we have our plan for a little bit later We have this ticket for a Fedora appreciation day, which does kind of take a It was partially motivated by what Ubuntu has they have a Ubuntu Community appreciation day, which is a really simple idea that one day a year Everyone's encouraged to say thank you to each other in the in the contributor community and thank people for their work in the project If you ever heard of the site happiness packets, it's kind of a similar idea of just Taking some time to say thank you to all the people who are putting their time into the project So towards the end we'll go a little bit more into that But now I think you want to talk a little more about the All right So one of the major part of comm ops is what we do we do a lot of metrics So when you say it's metrics we you know it can be even metrics. It can be badges metrics It can be no reports of annual reports of various teams and a lot of stuff like that. So we have Something cool we have something called a comm ops toolbox. So this is this is a part of the wiki where we have a list of awesome tools It's just cool. Yeah, so this is our toolbox. So we have things like we have the We have IRC bar dot py that that pulls data about Messages from IRC. We have a annual gripper tool which gives you yearly statistics There is data gripper that we use to pull data from from from Http server Fat messages is our central messaging hub that that you know holds all the messages And and this is a tool I was working on for my previous GSOC. So what this tool does is It'll take you know a list of usernames and it can pull data about it and about about that person And you can generate statistics like the ones you saw during the keynote presentation Then there is stats cache which is a tool that that is built to you know have data cached locally So you can have faster transactions And then this is one awesome tool was built by Remy So what it does is it can take meeting logs and it can build, you know beautiful word clouds Yeah, so these are the some tools we use Okay, so our metrics 101 So if you if you're interested in building any any metrics tools for Fedora or if you want to generate some people's data You you will want to start with data gripper So yeah, so this is data gripper So this is this is the one tool we have which is used to you know pull data about people So if you so the link is apps.fedoraproject.org slash data gripper So what you can do is you can you have a you know specific set of get requests You can make and you'll get JSON responses or Http responses based on what you asked for so let's say I want And this is a quick clarification for the video to this is all pulling message for our data from Fed message If you're not familiar with Fed message this is like the federated message bus of Everything happening in the project you can almost think of this is almost like a raw fire hose of data of Everything happening in fedora all at once so this is things like someone got a badge someone made a commit to a Peugeot repo someone submitted a koji build Someone wrote a blog post for the fedora magazine or submitted something to the fedora planet There's all these different kinds of things that are emitted onto this bus of things that are happening in the fedora project One thing I think is really cool just to see on free note There's a channel fedora-fed message and you can just see everything coming out all at once it's By itself, it's impossible to make sense of because it's just so much data You have just lines flying across your screen, but once you can use a tool like yeah, actually like this So by itself, this is impossible to make sense of but you can kind of get this idea of these are all things that are happening in Fedora right now And this is like the live feed like you're seeing fedora move and happen right now in front of you And it's when you start using tools like data gripper that you can take this this raw fire hose and start filtering it down into meaningful impactful conclusions and you can start forming it to Actually work with it and start manipulating it to ways you want to work with it No, so this is all tied to fedora user accounts or for the fedora account system So like if you go back to the channel really quick That might I think that would like there's also ansible playbook runs We'll also go in here and you can kind of see like most of these if it's not a bot There might be like a fast user name that will come up like Escalter started a new build on the for testing copper So like it is mostly tied around like username based data with like with the fedora account system But there's also other things too like you can see a bunch of Tascotron results from QA. They're coming out right now, too The answer to the question So a bit more into it. So what fetmes is does is it has a I know a central message hub that listens to all the Messages so we have a lot of web apps, right? So we have a girl. We have Copper we have we have like a lot of git repose So what it does is when someone does some kind of an activity in that web app It pushed out a message to fed message So we do a subscribe to it and we can pull messages from fed message All right, so our data gripper is this tool that you can use to actually, you know get historic messages from this So what do you can do is so this is the syntax? Alright, so this is the basics in tax for pulling data for user So what it does is you can go to apps that fit our project dot org slash data gripper Slash raw so you you want the raw messages so you go to this URL and you can give your fast username there. All right, so Can anyone of you give your fast username I can I can show you sorry. I just refreshed it It's it's loading. It should be back up in a while So this is how it works. So if you were wondering how where I got this data from there is a awesome reference tab that you have in data This is a 500 This might require a restart of data Yeah So I Wanted to show you the API calls. So these are like a few things you can do. So you have a user You have a user tab which you can use to get data about the user And if and that will give you like all the data the user has it will give you all his meeting logs It will give you IRC messages It will it'll do a lot of stuff So if you if you want to fine-tune if you want to go deeper into it so you can do something like hey, I want only the Meeting stuff so you can do a category is equal to meat bar and it will give you messages only about the No meat bar messages that were generated for that user. So this is like I want my data and I want data that no only Only you know subscribes to the meat bar. So if I was in a meeting it will give my give me all the data Which was coming from the meeting board and this also has you know a really cool thing called Delta So a Delta is of something you can use you know to fine-tune the result even further So if I have you know if I want data for one week it would be the timestamp in epoch time so it would be 24 Into 60 that would that would give you for one day and This would you know you can just you can just do the math I think those are not caught by data grip data gripper, but you have a channel for it So it gives you it's Fedora Fedora minus. I know see it gives you Messages like here. This is down and this swap values are low I don't think this goes to the data gripper though. You cannot search for this because it's it's a different data All right. So this is a sample response you will get from any message. So you'll have Arguments it will show how many pages are there in your response you have rows per page how many messages are there on one page and You'll have things like raw messages raw messages are actually the stuff that you know you will require so category There'll be certificate. That's that's not required And it will it will have a signature it will have a timestamp of when this happened And and that is something called topic that comes from FET message. So a FET message has this you know Really good naming conventions. So you'll have this this is going to be This part is going to be same for almost all the messages So it is like federal project at org and you have this is on production so if you if it's on staging it will it'll be dot stg and You'll have stuff like if it's on IRC you will have org that federal project dot broad dot IRC and you'll have Whatever comes after that. So if someone gives me, you know karma, all right So this is the delta. I was talking about you can filter using start time and end time So if you if you specify a start date, you you can get you know, you can get data from that date And if you specify an end date, you will have like, you know, a time range where you can get data So this is what we use for getting data about an even so let's say I want to gather data about flock 2017 what I will do is I'll go to the badges project. So this is Justin's profile on badges Yeah, so we have this bad. So this batch is given to all the people who are in flock So what we'll be doing is we'll be scraping this this website and we'll pull all the usernames from there So let's see that there are these people who have claimed the bad so far So we'll be pulling all this data and we'll be taking this and writing a script putting it in data gripper So it'll be like I'm gonna be pulling data from data gripper for each of these users and I'll be generating graphs out of it You can there's a QR code on your So if you if you scan the QR QR code, you'll get this badge. So using these data, I can you know, maybe pull data Let's let's say I was I just pull my data from yesterday or probably for five days and I can figure out what I was doing So let's say there were Pagore commits and there were no I was an IRC giving karma to someone so Those people would have done something, right? So you can gather a lot of data from that. So I have a few sample reports All right, so What I did is I pulled data of all the commops members for one year and So this was pulled by one of the tools I made I'll get back to that in a while. So let's say I want data of Justin I'll go to his folder and Okay, that's this Jason data right now so that Won't make much sense right so this is going to be an endless list of what he did and it's it's going to be a List of all the you know things that are mentioned in Jason. So I want to prettify this I want I want to make sense out of this. So what do I do? I write scripts for this So I have I have the weekly data of commops and I have generated So this is the tool I was talking about this is the tool I built for my GSOC so what it does is it it takes messages from data gripper and You know you can build graphs using it. So how it works is it's a Python script So you basically go ahead and enter the username you want so it's going to be an I ask in it So let's say it's definitely. I'm sorry. It's a it's a fast account name So it's going to be and just for quick reference. This is in pager dot IO slash gsoc dash stats It's also in the commops toolbox that we showed earlier if you're trying to find this takes arguments of user name So this user is going to be this is the files username and I can specify the mode So if it's more it can be Jason it can be PNG it can be It can it can be like most of the image things you use So I'll just go with SVG right now and I'll grab this data for two weeks So what it does is it initializes a background process for pulling data from data gripper It's slow because So this was built from the Jason I showed you earlier So, you know this takes takes up all the messages it it passes it into graphs and you get to see what he was doing So let's say he was he he has he was working there were two thousand two hundred and eight messages on wiki He he was tagging the packages. I mean there were he tagged three thirty six packages This is for last year Yeah, so he had one seventy eight messages on github Mailman track Yeah, so he was working on for good for quite some time he he was an ISE doing something he was probably giving karma to people and This is a film and so you can even filter out things like if you don't want wiki you can just click on it and it'll go off Yeah, those are yeah So our topics are basically the entire string is so it's gonna be all good for the project at all dot something So that's a topic. So if you want to go into specific categories like IRC dot karma, you can also do that So it's gonna be a subcategory of the categories It's like IRC is the parent and it has a lot of styles like IRC karma like IRC meeting and stuff like that I think I also have that's for that. I mean if I if I had Yeah, so this is I went into the figure stuff and here you can see how What he was working on in figure? So he was he opened up new project. So that was like 22 person pull requests There were issues. He was working on something on issues. So it might be creating issues it might be editing it and This is good. So this is maybe he made good pushes comments and And there was one pull request. I guess Yeah, so this is part of So he had 109 messages out of that He had like 60% where something to deal with issues in this in this example It used raw percentages, but you'd still be able to get the raw numbers out of the data gripper for that They would have any questions so far or anything they want answered before we keep going And there's something you can you can also generate markdown reports So let's say if you want to push your reports to github you can you can generate a markdown report So it would look something like this. You can basically See what he was doing. It's on wiki page Okay, that's a lot. Yeah, so this is the specific things he was doing. So he was like commented on a ticket he edited the priority and He tagged tickets. So these are like specific things he was doing. So this is the graph we show I have there is a visualization of this There's also something called the gauze report. I Don't have it right now, but So gauze report is something really cool. It's like a live visualization So if I if I pull data for one year and if I run the gauze visualization It would show you it would show you a really cool graph of what's happening and how much messages were there each week and stuff So I could have shown that but let's see if that works We actually do have a Three-part series on gauze Let me see if I can just pull these up and I'll show them on my laptop So gauze is it's a visualization for some of the development history. It uses a Version control system and we'll animate it when to act like animations like video. Yeah, here we go So you might have seen something like this before so gauze is actually taking All of the activity I guess in this case it's for the fedora calendaring system And this is going through all of 2015 and Let me see if we can put these links up to the community blog There's actually a tag. It's community blog dot fedora project org slash tag slash gauze Actually, yeah, yep that one This is we have this a three-part series about Michael. I contributed now what? And this is this was something that we did about a year ago Which was trying to go deeper into why people contribute to the fedora project and it was using gauze as kind of the tool to make some of those conclusions With the data that it was pulling from Could you go and play one of those two that we have well? I can't remember what we did in this one in part one. You know which one was this for is that all all three of These did different topics in the Fed message bus I just can't remember which ones are which so this was I think was this oh Is this all the Fed message for 20? So yeah, this is all of fedora in 2015 for everything IRC Pajur track This is kind of like a cool way to get a visualization out of it and And in this case the series that's on the com blog actually does try to make sense of what all this is it's not just a I mean it's kind of cool to watch anyways, but There actually are some ways that you can use this to try to do I mean you can kind of see a visualization to like all these hot pockets of where all this attention and where a lot of Activity is in fedora and you can also kind of see too There's a lot of areas that probably aren't going to be well represented there like for example the ambassadors are like the Development side of fedora is obviously getting you can see some of the hot pockets with like the track and for sure repos but on the other side you also would probably see like a lot of either if they're smaller areas it might be like a event report But there's things that aren't well represented in the fed message bus as well And that's something that is we're trying to think more about and try to find ways to get that Get ways to represent that activity into the fed message bus But it is a little more challenging because it is more soft soft kind of activity It's not something you can track with get get push pull request meeting end meeting start So it is kind of a nice way to see that too In this case part three and this was from part three. This was the aggregate Which Yeah, so if you look at all the blog posts here You can see activity from fedora elections the infrastructure get repos kernel regression testing IRC karma and the fedora calendar as well which is also nice if you're if you have your own project that's in pleasure and If you want to build visualizations like this you can actually create these so the tool that's used to make these is actually called Fed message to gorse Can you pull that that repo up really quick too? And this is a tool that you can actually use to take data from fed message Manipulate it with data gripper if I'm right and then you can actually Yeah, I guess it's on on Ralph's github We should probably move that to sure now that I think about it But this is a tool you can use you just use image magic fed message, of course some of the fedora infrastructure packaging for fed message You can choose just like you were we were doing earlier with data gripper You can see that you just run as a Python script You use the amount of days that you want to use and then you can also filter it down with some of the different topics or Categories in pleasure as well. Yeah, so there are two things you can do with this You can either use this as a standalone tool to generate data from fed message or you can integrate it your own program So let's say I wrote my scripts and my scripts was initially just PNG's and SVG's So I wanted to extend that so I use this so I use this as a Python wrapper and I integrated this code into my code base So this code is available and this also has a Python API So you can just go ahead and include that in your your own code base and you you can generate beautiful gorse graphs out of it So I'll give you a quick walkthrough of what I've been doing So for this tool I actually took some time to you know add a help menu to it So what you can do is you can do a lot of stuff with this So the thing is this requires a mandatory username. So that's kind of obvious So if you want to you know pull data of a specific user, let's say I want to pull data of Justin I'll just specify minus you or minus minus user and I'll give the first username and Let's say if you want to pull all the data for a category So you will be mentioning minus you all so that's like all the usernames and there's something called category You can use that is like IRC or Paguer or Which were category you want to pull data from and then there is start date and end date which you can use that that's gonna directly go to fat message and gonna pull data for only that amount of time and There's gonna be a group statistic So what this is going to be helpful is let's say if you want to generate a report for the Fedora Council meeting You can just go ahead and run this tool and and gather your data for an year And you can you know generate graphs out of it and for the report you can say hey We did this we had so many messages in Paguer and we did some so we closed out so many tickets and stuff and Then there's a logging mode where you can dump the JSON So if you if you want to write any of this specific metric tools You can go ahead and do that using the log The log method and mode is the type of output So let's say if you you want it in SVG So you'll just type minus M SVG and if you want it in PNG you will just do minus M and PNG and It's gonna be there are gonna be weeks So you can specify minus minus weeks or minus W So if you want to pull data for like three weeks or four weeks can just go ahead and specific Hey, I want data for only four weeks. So this is gonna give you data for four weeks And there's also this mode called interactive mode. So if you run this tool with minus i it's gonna pop up asking like hey What's the username you want to pull data for so I'll just type in my username as like Is there a specific length for this you want data? So it's like you can Specify number of weeks so you can specify a date So you don't have to like script it so it will it will do all the hard work for you It will prompt you to enter stuff and yeah This is a really bad time for data All right, so I Talked about all the nice things data gripper can do but there are some limitations of data gripper So the first thing is it's not very reliable as you can see right now So you will need some kind of a local stash that you will know you will need to store all the data So let's say I want to generate this data, you know, like maybe after two years or maybe after three years I'll have to pull all the data again. So We came up with so that in Fedora Infra repo we have something called Stats cache. Yeah, so this is a tool that works with Data gripper. So this is gonna pull data from data gripper and it's gonna store in local database So so this is gonna be a one-time process. So if you pull data once you can pull data from 2013 to 16 And you can store it locally So the next time you want to you know generate data for a user You don't have to go to data gripper and pull data You can just directly communicate with stats cache and you can know like a build dashboards out of this So this is I don't think this is like quite Not maintained. So, uh, no becks found out this really cool tool called Percival So that is by bit earth your team. So I was going through that so that that looks like quite interesting Yeah, so this is the tool I'm talking about. So what this does is this integrates with a lot of services So we have we have an instance of hyper kitty running. We have get we have a lot of git repose. So instead of, you know doing Get requests for all the things what we can do is we can build a fet message wrapper around this tool and This this run basically runs on the elk stack. So it's elastic stack elastic search lock stash and kibana So elastic search is is the place where you dump all the data Lock stash is the one that gratifies the data and kibana is the dashboard. So what I have Set this up. I'll give you a quick demo. All right. So this is a sample code for getting data from hyper kitty So hyper kitty is is the our front end for the mailing list so what you can do is you can specify the URL of the hyper kitty here So it's like I'm pulling data for federal join and then you can You know mention where you want to archive it. So I'm gonna store it in some place called archives And then you can you can specify a daytime here from where you want the messages So it's gonna be like from message in report at fetch. I'm gonna fetch all the data from From our archives from 2014. So when I do that It's gonna store everything in something called inbox format. So this is this is gonna be archives of You know all the males from 2014 to 2017 and Then there's something called Mailbox plug-in. Yeah, so what this does is it it kind of passes whatever is in mailbox and it will give you data out of it so This this is actually a json format. So this is gonna print Let's say the raw json. So if you want stuff from that Let's let's see what are what are the stuff in there? One thing that I think this is worth Clarifying to about we're showing you all these like these really cool visualizations and data and that's all there And it's there for you to work with and access, but then it's kind of like so what it's here You can see it. What's the point of having all this data? What can you actually do with it? What can you use it for there's kind of this one of the things with commops is one of our Kind of the the motivations that we have is this idea of storytelling and When you look at a project like fedora, you have all of these different corners that are all doing their own thing But at the same time we're still part of this bigger project and when you can see and understand what's happening in your community In your project it makes it a lot different when you're thinking like I don't really know what's happening over here I don't know what's happening in this place when you can take this data Make sense out of it and make interpretations make conclusions about Like what we did there was a talk at last year's flock that went into like it was like a analysis of the lifespan of a contributor and in this case we actually saw that in the year in 2016 of The people that entered the fedora project a Majority of the people that entered fedora only stayed in the project for three months the majority of that Some of them would continue contributing others were active for for less than three months But a majority of them were active in the community for three months and by using these tools We were able to see that we would never have been able to know or understand like oh Hey people started dropping off like they attended less meetings And then there was this this drop off towards the end where they stopped contributing to the project And in this case if you have your own project in fedora or you have something that's in the fed message bus You can actually use this data to try to make interpretations and conclusions about things happening Right in your own in your own part of the project, too And when you can take this and actually make sensible interpretations and conclusions It lets you the project maintainer or part of a community team Actually have a better understanding of what's happening and find ways you can improve on how you engage with people in your project like for example Like the design team if you have some of these data or some of these tools You can better understand maybe weak points and maybe your onboarding process Or maybe there's a point where you see people get started like with fedora badges and then they drop off after a point Well, what was that blocker when you know that that's the point where he will drop off Then you can dive deeper and you can understand like oh this is what we might need to look at a little bit more I didn't realize people would make one badge and then they would drop off Was it there they didn't feel like they were really contributing or was there something that was a disconnect there With this data with these statistics That's what kind of gives you the base to go off of when you know what the when you have an idea of where that Cut off or what problem there is then you can actually start making a plan you can make you have an idea Of what's happening and you can try to fix a problem Yeah, so Yeah, so the question is is mirror data available in the fed message bus for you to view and work with So, okay, so to get the mirror data into the fed message bus Yeah, so if you want a mirror data, I think you would you would want to talk with the infra people because I don't think Normal people have access to the server. So as No mad diem said that there are no trackers that we use to identify users. So all we have is specific set of IP addresses So obviously for obvious because we cannot give out IP addresses of people and that would be problem So if it's a public IP then, you know, that would be if it fall into the Yeah, I definitely think there would be ways because I'm assuming this is like mirror manager, right? Is the software for I Mean, I think there would definitely be a way to get fed message integration or some kind of support for collecting data And a mirror manager like Sachin mentioned we would have to be a little bit careful about what kind of data we're collecting But I think there would be that a lot of these are these are all libraries All these applications in fedora infrastructure are using the fed message libraries to To send messages to emit messages onto the bus So I know it would be possible to probably get I don't know the mirror manager project that well But I definitely think it would be possible at least get some try to get some kind of meaningful data out of that I think that would actually be probably a really valuable place to try to get some data to understand more about our users instead of just our Contributor community to Specifically, I don't know how to how to go forward on that But I know that the possibility for integrating it and working with fed message is definitely there So this is one part of the Perseval tool So this is using mailbox and hyper kitty. So what it does is it pulls all the data from the archives And this is a one-time process, right? So if you pull the data once it's gonna be archived in your local stash And you won't have to ever connect to internet again to gather statistics And this is a good thing because you know, it's you're saving bandwidth and you don't put stress on service like I did on data griffin All right, so these are the stuff you can grab from The mailbox so it's going to be unix from and there's going to be a content type It's not really worth to get the statistics on you have from to subject You have the body or if it's a reply it's in reply to and then you have the date So I I just grab the from a field from the messages and it looks something like this So this is just plain text. This is I'm just printing this in Python So what I can do is I can I can go ahead and I know store this somewhere. So I Will put that in stats cache. I mean, sorry, I'll put that in lock stash and So a lock stash is this tool like No stats cache that I mentioned so it's what it's going to do is it's going to gather all the data and it's going to enrich the data So let's say I have from and and let's say I have sent five messages So, you know, this tool possible tool is not very intelligent to understand these are same So what it does is it puts into lock stash and lock stash is gonna Categorize stuff together. Let's say I have like messages from Eskimo the Fedora project So then there are five messages. So that's gonna club it together and it's gonna enrich the data Yeah, so this is the tool that actually, you know, you can use to build dashboards from what data we have So I have added that data in into a kibana and there's this tab called visualize So what you will be ideally doing is you will be adding a new visualization after so I have added all the mailbox data into this So I can choose what graphs I want. So let's say I want to do a pie chart And then I'll click on mailbox. So that's the data. I added So the question was we have all these tools. We have all this data But what's kind of the next step for where we're going to go with the project for making this and do consistent actionable data that we can make sense out of and try to use to make real plans and actions in Fedora Yes, the thing is All this while we have been using, you know, a lot of custom metrics like we had a lot of Python scripts to generate data So that is something we cannot, you know, actively maintain because it's going to be either lost in some Fedora repo or it's It's just not, you know, reusable. So what we are going to do is we are going to take all the fitnesses data and we're going to put it into The stack we are going to use Percival on it. So the long-term plan is, you know, to write a plugin for Percival That's going to gather all the data. So what you can do is Percival has this tool called No, right now Percival have support for Asgbord They can gather data from Git, GitHub, there's HyperKT, Jira and stuff like that So maybe in around five to six months, we'll have a Fed message over there and Yeah, this is all So you can use these things, you know, to pull data. So let's say I have I need to pull data for Let's say Asgbord. So I can just write scripts for this I can write a plugin for it and I can put all this data into dashboard and I can, you know, build dashboards on the fly and Let's say I want Fed message. We don't have, you know, specific things for Fed message yet So I'm, you know, I'm planning to contribute to this project and write a Fed message plugin for this So you could do something like Percival space Fed message and then you can specify a category and it's gonna Get all the data for that category and you can build whatever you want on Kibana. Like as you can see here, I Mean, I think the database is not refreshed and it requires quite a lot of internet So what it does is you can go here. You can select what you want So let's say this is like inbox data and this kind of request a refresh But I don't have the bandwidth to do that right now So this is, you know This has everything in mailbox from this is Fedorazine archives from 2014 to 17 So this has all the data. So if you want, you know, let's say data from 2016 to 17 You don't have to write a special script for it. So you're just gonna go here select the dates and there's gonna build a graph So we'll have this you know dashboard I don't probably in our infrastructure and anyone can go there and build dashboards and share with the community And just to kind of add on top of that it's kind of like this like so what? You can see again in this tool specifically It has a lot of support for all these different hooks and back ends that we are already using in Fedora And it is kind of this big idea that we have all this data and we don't want to ask the ambassadors We don't want to go to the ambassador and be like hey if you want to have a fancy chart and graph You can come write your own Metric stack for or metro your own dashboard for us that we don't want to make it hard for other people to get This information so part of this is go like why these tools like cabana and you might know other ones like Grafana as well What they are really great at doing is making sense of what all this data is really doing in a really easy to understand Format like you can see you can have all these different kinds of charts You can have like a pie chart vertical bar graphs Gages and there's all these different ways that you can visualize this data and turn it into something that and you Can point anybody to look at and be like oh I understand this this makes sense This isn't a giant JSON blob file that I can't read So that's why like this is part of like the the next step of how we're going to get there and do this because It's really taking it from this point where we have all this data and we're writing all these manual reports And it's really hard to kind of bootstrap people into this and it's really confusing This is kind of this platform where we can start pointing people to and be like hey Here's your dashboard for comm ops or for a specific project It's really kind of the the next step for how we're going to actually make it easy for anyone in fedora to make sense of this data Because it is like a lot of this right now is like Writing your own Python plugins and scripts to pull this data out and we know like not some people Hey, if you're all if you're all about that We would love to have you come by comm ops and spend some time with us But if you're not We want to try to make that easier for other people in the project to understand and that's ultimately what the goal of Kibana of Visualization of dashboards. That's what the big picture of this is So this is kind of how we're trying to make sense of all this and make it easy to share with the rest of the fedora community So that way you don't have to be the Python wizard to make sense of what's happening in your own sub project if that's not We want your feedback, but we don't want to make you write Python code We recently build, you know a graph which would show ambassador locations by activity So that's something we recently did so we have people to you know work on Python and you know to build the fat message plug-in so maybe next year Matthew will not I know have to write these scripts and run it for 16 hours You can just go to this dashboard and he can he can pull it out of this Just definitely an action item that we definitely do have on our plate is to start conversations with people in the door infrastructure to start Basically the same as we're doing now. I Definitely recognize the need that we need to have cross collaboration and pollination on these ideas so I definitely that's probably one of our biggest action items right now is to start those conversations and try to Get some support from other people in the food in the project community to help us build those tools So we do have some we kind of do have some group of people in cops working on that But I would like to try to cross pollinate it out to infrastructure So I'd say that's definitely an action item that we are it's it's on the table They answer the question There's stuff like goals. I think that's also taken care of by Fedora hubs So if I'm not wrong, that's the feature. You also have let's say we have something like I Need to get the badge for making thousand wiki edits And so I we can have something like how many have made and how many to go so it's like you are 50% the goal so that I Think it's already there on hubs. I think sign is here So you have a progress thing like you have you are like so many messages away from this badge I saw that in the mock-ups So just to get that clarification too so for the grimoire or for the cabana kind of visualizations it's more of focusing on the Community-wide Fedora project holistic view where there is some support already in Fedora hubs at a more Individualistic contributor level like the example was I'm this many messages away from getting Like a mailman badge or I'm this many commits away from getting a Pajor badge That would be more like the individualistic level, which is what support or it's in progress for hubs This is more of like at the holistic big picture view for the entire Fedora project So does anyone have any other questions or ideas or feedback on metrics or anything that they you came here You were like I really was curious about this and you didn't hear it answered or you're still wondering about all right So I think from this point on we're kind of going to go and switch over to the hack session Since there's a decent amount of people in this room. I think it'd be really great I don't know how many of you have I mentioned it at the beginning of this session about this Fedora appreciation day Ticket that we've had for a while So I think that might be one thing that we work on from here on out and A few other things I just need to pull up We had some we haven't either pad with some of these tickets highlighted But I think from here it'll be a little bit more