Good afternoon and welcome to this talk. I'm Erika Zini, Debian developer. I'm the one that realized the Debian Contributor site,

which is about making sure that all Debian contributions are acknowledged, no matter what they are. And Debian Contributor came as a way to fix the old situation we had in Debian,

where the only, I begin, is that OK? To fix the old situation we have in Debian, where the only kind of visible type of contribution was mostly being a package maintainer.

And this is the kind of page that you would have for your contributions in Debian in the past. A list of packages, bugs that are open, versions in several distributions, and so on. It's pretty nice you see the work people are doing, but you only see package maintenance.

And all the people doing other things, like translations, Debian has a vast group of translators, for example, reporting bugs, porting to different architectures, were just not visible. And some of them still are not visible.

And the people doing those were unsung heroes. Nobody saw their name, occasionally on mailing lists, but that's it. And the people, including Debian people, including the people that were not maintaining packages, they thought that Debian was only about package maintenance.

And the process to join Debian was called the new maintainer process, even. So it was all around package maintenance, but that was not reflecting the reality of the project. So at some point, we approved the diversity statement,

which was a first step in acknowledging diversity in Debian. So the Debian project welcomes and encourages participation by everyone, no matter how you identify yourself or how others perceive you, we welcome you. We welcome contributions from everyone,

as long as they interact constructively with our community. And while much of the work for our project is technical in nature, we value and encourage contributions from those with expertise in other areas and welcome them in our community. So this was voted as something we all agree with,

or well, mostly all agree with, like an important document in Debian that we agreed to follow, in principle. But this documented the effective situation, but actually seeing non-package maintenance contributions still

did not happen. There was a blog post in my blog where we decided that we have the need to make other kind of contributions visible.

We tried to make a little list of what other kinds of contributions are and where some of those information could be found. But still, there was no actual technical, central way of showing these things.

All of that because Debian is complex. This is a simple infographic about Debian, which in a corner is get involved, which you could blah, blah, blah, and then blah, blah, blah, blah, and so on.

That kind of gives a bit of a visual idea and a very simplified visual idea of contributing.

But different parts of the project track contributions in different ways, some by email, some by login name, some by GPG key. It is not really feasible to map all of Debian because it keeps changing. And people set up new systems.

And we have Debian maintainers, people that commit on Aliot, people that report bug, people that edit the Wiki, translators, people that participate in making these discussions. We have people maintaining packages, people reviewing somebody else's packages,

people who organize conferences, people that go to conferences, people that run Debian booth at events, people that do press releases, system administrator, people that develop web applications for Debian. So yeah, all of those should somehow get credited.

I missed a slide. So it is a social problem of self-perception. Therefore, we needed a technical solution.

And this is what I'm going to present today. The main idea I had was to introduce a new title in Debian, which was the title of Debian Contributor. And there is no process to become a Debian Contributor.

When you contribute to Debian, you automatically are a Debian Contributor. And when you are a Debian Contributor, you can have your name on a list. I'll now get the list.

There's a list of Debian Contributors, a list of all the 1,750 people, as much as we can tell, that contributed in 2015.

Actually, that should be changed in the last one year, maybe. So we have credit for anyone that does things that actually contributes to Debian. You can also choose not to have your name on a list.

There is some privacy management in this. So you can opt in and out of the list. By default, all contributions done with the Debian org email address or all of Debian org email address are public.

And things done with the private email address are sort of mostly opt in. And there's a way to manage to get out of the list if somebody doesn't like it. And now, we finally have a place

where we can thank and acknowledge the contributions. So I feel a bit better in Debian because it was unacceptable that in the past, people would contribute to Debian and not even get a thank you. So one of the important things is that it kind of

automatically expires. So the list that's presented in the main page is of people that did things this year. And you can still get credited for the past years. You can go back in time. And so when you do things in Debian, you get thanked even though you stopped.

So you can say, well, I did things back then. That was nice. You can then you have a page with lists everything one has contributed to.

So I have uploaded packages since 2007. Actually, 2006. I should do some merging of these kind of things. Messy at the moment. But I've also been an application manager

to help people get into Debian. I commit to repositories and get Debian org. I talk to people on the backtracking system. I've done some editing in the wiki.

I developed the nm.debian.org website. I send patches to the Debian system administrator and so on. People can be thanked for way more than developing packages. And so this actually starts to change the image that we

have of contributions in Debian. And then I have the login button.

In order to get this done, there was the need to do identity management. So say that I am also this email address.

I am also this GPG key.

And I'm not exactly sure what is that key. I'll need to look into that.

Oh, well. Oh, that's a bug in the site that I fixed yesterday. There's an audit log of all the changes that happens because there's a robot that tries to be smart and do some auto-association. Oh, yeah, that's my other key.

And you can change your full name to whatever you want. And then because people have all sort of things in full names, and they have different full names in different parts of Debian, and one here can choose how to be credited.

And then I can manage visibility of contributions from all other identifiers that I have. So if I don't want to be credited with my work address and I only want to be credited with my private address, I can have some control on that.

And so in order to get this implemented, we had to also introduce identity management. That did not exist in Debian. In the past in Debian, there was really no way to say that email is actually me.

And this opens a lot of things that can be done in Debian in the future where new services can interface with this and deal with people instead of email address, GPG keys, and so on.

Identity management has been quite tricky because you have you read this before. Falsehood programmers believe about names.

Basically, all of these things are not actually true. So for example, people's name change.

We have several people in Debian that have legally changed name and even non-legally changed name or are in the process of changing names. And then we need to collect data.

And the main design principle on collecting data is don't expect me to do it because Debian is fast

and it's impossible to go and find everything. So I designed the site in a way that allows every team to submit contribution data to the site. So every Debian developer can go here.

Well, every person can go here and see the list of data sources that are currently sending data to Debian contributors. And every Debian developer can create a new one.

It has a name, a description, a URL, an authentication token, and some other things. And basically, you configure the authentication token and you include it in the post request when you send the data and you're done.

So everyone can send data. Every team has a possibility of sending data. The general design principle for this was when in doubt, relax the technical requirements. So the system does not support gaps in contribution.

There's only first seen and last seen. It is OK to have a granularity of a month even. It is OK to have some lag and send contributions in a cron job run like once a week.

It's OK if some contributions are missing as long as when a person keeps contributing, eventually they will show up. So I tried to make it easy for all teams in Debian to actually participate in this because harvesting, yeah,

cannot really be centralized. So teams need to do it. And then there's some auto matching so that all your user names are mapped to Debian user names and things like that.

The technical details in the system is that there are users that are people. There are many identifiers related to a user. So a person has two email addresses, three GPG keys,

one or two login names, and so on. And data sources, each data source has several contribution types. So for example, in package maintenance, one can maintain a package or one

can sponsor the package of somebody else. And there are contributions which link identifiers and contribution type. So that email address has done this contribution time from that date to that date. That's a data model. It's reasonably simple.

And data submission is documented here. It's basically posting a JSON that

looks like this to the site. So you have an identifier, which is a login name with a value Enrico, that in this identifier, login Enrico,

that for the data source that we are submitting,

has done talking from 1st of January 2010 to 16 November 2013. And that's a URL that can identify those kind of contributions in more detail. That's all that is required. Actually, that's more than what is required,

because you can even post without dates. And the system remembers the first date it has seen and uses the submission post as the last date. So you can have a data source that just has a list of people who are currently doing things. And you just post the list of people, and that is enough.

So the idea is that it should really be as simple as possible.

Now, it's not about Debian maintainers. It's about Debian contributors. Actually, an official Debian developer is not necessarily a Debian contributor, in case they actually stop doing things. Debian contributor is actually a serious thing in Debian now.

People talk about Debian contributors, rather than talking about Debian developers, unless they want to refer to people who have voting rights. What was called the new maintainer process for getting into Debian? Now, it's called the new member process.

It was a clever hack that allowed to maintain the same acronym, so that documentation didn't need to be changed very much. And when somebody wants to go to the Debian conference and ask for sponsorships, the Debian organizer

asks for explanation of what one has done in Debian, or a URL to contributors Debian. So it starts to be like a curriculum of people in Debian.

And a really nice way to start contributing to Debian is to contribute to Debian contributors and make it nicer. And then one is contributed for contributing to Debian contributors.

Future development in the site after the Debian conference. So now we see that there are people that start and end contributing to Debian. And so we can tell them welcome and goodbye.

So recent development is having a list of people that just started contributing to Debian. Ideally, there is a group of people in Debian that are considering to start sending email

to people who show up doing things and saying welcome. If you need anything, get in touch. I'm a real person. Just so that we make sure that people are not left alone doing things. And they can ask.

They have a point where they can ask for help. And we are starting to consider using this to track people who stopped contributing to Debian. So here's another list of people who are listed as official Debian developers but who don't seem to have been Debian

contributors since 2010. And then there's the Debian missing in action team can send an email to these people and saying, are you still around? If you're not, would you mind closing the account in a nice way so that we don't keep possible

attack vectors of password guests into Debian? And we can still compute the quorum when we run elections. So after changing the way Debian perceived itself

with the Debian contributors, the site has started to be actually a resource to work with membership. And a bit in advance because there is a few of us.

So I guess question time could make more sense than lengthy talking. I would say question time.

Question. I found it really nice that the DevCon organizers were also thinking of making or starting to make a resource for this. So it's a kind of contribution that people often, I guess, haven't thought in Debian where people are actually

doing physical work in order to make. Yeah. There's work ongoing to have people who spoke at DevCon, people who volunteered for the DevCon video team or organization team,

and potentially even DevCon Fattendees show up in the list. For DevCon Fattendees, we need to check because some people need to really opt in and not show up by default in that list. So we'll see a bit if the system is good enough

to make that happen. But yeah, there's been talking also like people running Debian boots in conferences like this to show. You mentioned that it's possible to exclude some activities,

for example, if you were talking on your own or talking for your company. Yeah. What do you think would make sense to, would it be possible to have some multiple virtual identities so that you can have a list of contributions on your personal side and a list of contributions you can hit for company A and a list of company B?

It is possible. Any multiple accounts, right? Yeah. You can have multiple Alioth accounts and manage them as different identities and log in in different ways to the site.

What we miss is a bit of a guide on how to do it. For example, if you want to keep things separate, do not put, I mean, you can make one email for work and one email for other stuff. But then if you have a GPG key with both emails,

things will automatically be associated with one of the two or the contributions done with that GPG key can only show up in one of these two accounts. So it would be a good idea to have two GPG keys also.

And that sort of a thing should also the tips when signing the change log of a package, you can set this environment variable to choose which of the two identities you are using. We miss a bit of a guide like that, which would be a nice thing to have.

At this DEPCONF, I even dropped the idea of we could have people having two Debian developers, I mean, being twice Debian developers with different identities. And I would have nothing against it. And people reasoned, oh, yeah, yeah, that makes total sense.

I was afraid that people would go like, ah, that's nonsense. Instead, they were like, oh, yeah, yeah, yeah, sure. So yeah, that would require two GPG keys and a bit of reasonably careful data management. And if one doesn't want contributions to mix up,

the problem is that once an action has been done, there's really no way to revert it. I mean, when you sign and upload and upload, that's it. That will show up in 1,000 places.

There's an email sent to a publicly-occapped mailing list about that GPG key, that email address, signing that upload. So you do it once. It will get credited, and it will be pretty hard to change. So that requires a little bit of, yeah,

precision and attention. How did you get all these? I mean, I think all the things you're now doing is a project that I submitted today, and how did you get all these to do that?

I mean, I'm not social media at all. I guess all the things are not really related to each other. So how do you convince and let it speak to you? Partly, I've said that if one doesn't show up, they're not Debian contributors.

And dear teams, do you really want that people do work in your team and not show up as Debian contributors? That would be really bad. So in a way, having this list is a motivation factor. For some of the important things,

like the backtracking system, I actually went and wrote the data mining script myself, and then I tried to tell, now you maintain it. I don't want to know about it. And it still runs as a cron job as my user in some machine, and I'm upset by that. And that should change.

And sometimes, another one thing that works really well is sometimes I get an email saying, oh, I do lots of work in that team, and I don't show up in your site. And I'm like, yeah, of course, because there's no data source for that team, right one? That's the documentation. And they're like, oh, OK, I'll do that.

And after a while, they're like, oh, now I show up. I'm so happy. So that also works. But there's not as many teams as I would like actually showing up. There is a wish list of the source code

and the information for contributing are listed at the bottom of every page in the site. Wink, wink. And here is a list of data sources that we know we would like to have and some information on

how one could help making it happen. So currently, we still do not track the localization teams, because it's really hard to go and find

who did some change because of the way things are done in those teams. But I don't know those teams, so it's hard for me to figure it out. But yeah, oh, and we need the Debian project secretary

to send information about when people vote, because that's also counted as a contribution. And that's easy. This should be really easy. So hopefully, yeah.

What? No, no, the list is public. So actually, the list is not just public. There is a crypto thing. So there's a list published together with some crypto

token that you can use to verify that your vote has been counted or something like that, or at least received. I don't remember how's this thing. Yeah. Well, the big missing thing is localization at the moment.

This needs to be updated. The backtracking system now works really well. I'm very proud of the work I've done. Spent like two days on it, and now it's good.

Yeah, more questions. People usually ask me about badges and achievements,

and I don't want to add any to this site. I would like this to stay objective and not tell people, I'll give you a badge if you do this, because I would really like this to be only about tracking

who's currently contribute to Debian and not start saying that person contributes to Debian more than that other person or that sort of things. If somebody wanted to build a badge system in Debian, then they could build on top of this,

because this is the first place in Debian that actually thinks about people and not about email addresses. And so one can somehow interface with this site, get the user, the list of persons and ways of tracking them, and then build badges

if they feel like. But yeah, this I would like to just stay as a list of Debian contributors. OK.

There are, oh, sorry. The question was, at which point one person shows up

in the site? And this is up to the various teams in Debian that send data. They can decide what is a contribution. But I'll list the teams that we have and what it takes. So to be credited by Debian system administration

is one single commit in their Git repository. Debian security tracker, I have no idea. distro tracker, I also have no idea. Perl packaging group, one patch committed in their version

control system. Backtracking system, at least five nails in the backtracking system. This is just one commit, just one commit.

One blog post published in the official blog of Debian. Having participated to the Debian conference once, although they haven't sent data yet.

Yes. Well, for Debian contributors, everyone that, OK,

at which point, what practical actual action

you need to do to show up here? No, that was not that. Are you asking what was the threshold so that the contribution would be accepted? Yes. If you write a patch, you can just submit it. And if it works and doesn't do any, it's probably accepted.

So basically, pretty much no threshold at all. If you do a wiki patch, any of them in wiki, you will show up there. So basically, at least for me, I spent years and years trying to figure out which package I want to package and maintain, and never found the right package for me. And this changed the whole, like, this opened my eyes

and yes, there. You don't need to tell people who's doing what you're doing. So for example, about people that we can consider new contributors, we can have a look

to give an example of entry. And this person has sent one package to be evaluated by Debian mentors.

Again, package maintainer. Maintained and did one, edited the wiki for one day. It could have been just once.

Lots of these people are maintaining packages. Has been sending mails to the backtracking system

from July to August. So I guess opened the bug in July and kept mailing on it. After five mails, one shows up in the system. And the limit of five mails is because there is spam that sends mail there, so we want to automatically filter out spam mails.

Can I get an example of people doing anything that is not package maintenance? I guess there's still a huge amount of packages in Debian that we made there.

I guess. Oh well, package maintenance, package maintenance,

wiki editing. Yeah, edited the wiki for a couple of days. Not continuously, like maybe even just twice.

So yeah, the idea of this is that there's no official process. You do things in Debian. And for most of the things that one can do in Debian, there is no need of getting some official membership first. The only thing one needs to become official Debian developer

is to vote and have an email address at Debian.org. But for pretty much anything else, it's a just do it thing. And as soon as somebody just does anything,

one tends to show up here. As long as things are done in a team that has set up a script to send the data to the site. And if you don't know what to do in Debian,

since last week, we have the welcome team. And there is an IRC channel, hash Debian welcome

on IRC.debian.org, where any question can be asked. Questions will not be replied in detail, but at least one can get the pointer of what to do next.

So that's another new thing in Debian. So if somebody would like to get on that list but doesn't know exactly how, then first step is going on that IRC channel and then asking. And somebody may ask, so what would you like to do?

What would you like to learn? And try to give some direction. Does that answer your question?

I guess if there's no more questions, we can end it here. Thank you very much for.