andrewsomething@ubuntu:~$

Just another Ubuntu weblog

Reflections on being an Ubuntu Universe Contributor

with 7 comments

Last week I made the big leap from “contributor” to “developer.” At the most recent MOTU-Council meeting, my application to become a Master of the Universe was approved!

Now, I want to take a minute and think about the Ubuntu Universe Contributor (aka Ubuntu Contributing Developers) program, how we can better utilize it, and ultimately encourage more people to get involved in Ubuntu development. But first of all, what is the UUC program anyway?

On the wiki page, Ubuntu Contributing Developers and their roles are described as:

  • Are members of the universe-contributors team in Launchpad
  • Are collectively responsible for the maintenance of most of the packages in Ubuntu (the universe and multiverse components)
  • Merge new versions from Debian, work on bugfixes and new packages
  • Continue with sponsored uploads
  • Participate in technical discussions with other Ubuntu developers, providing ideas and feedback

That’s pretty vague. A lot of people who aren’t UUCs do merges, work on bug fixes, and get uploads sponsored. The qualifications for joining are even more so. You have to meet the requirements to be an Ubuntu Member and submit an application to the the MOTU-Council.

We need to be more clear on both the goals of the team and what it takes to join. But this becomes a bit of a sticky issue. We shouldn’t be placing hard metrics on the requirements for joining. Joining should be about the quality not quantity of uploads and connecting contributors to the development community not quizzing them on technical aspects. Becoming an UUC doesn’t give you upload rights, but it does make you an Ubuntu Member. So while it is a development centered path to Ubuntu Membership, its really a community designation not a technical one.

So how about who is an UUC? Looking at the UUC members list, you see that there are 16 individuals on the team. Including Chow Loong Jin (hyperair), who just joined, congrats! Of those, nine of us have gone on to make MOTU. So there are only seven proper Ubuntu Universe Contributors. So, it seems that becoming a UUC is a good step to becoming a MOTU, but still not all that many people are using the program. In fact, more people have become MOTUs this year so far than have become UUCs.

According to Ubuntu Top Uploaders list, 395 people uploaded at least one package in Jaunty (although it’s not a completely accurate count as some people have more than one listing due to using more than one email address). Of those, only about 100 people uploaded more than 10 packages. The top 25 all have over a hundred uploads. The level of involement is amazing, but it is also extremely top heavy. We need to shift that curve down.

For me, the main goals of the UUC program should be to draw reoccurring contributors further into the community and acknowledging them for the work they have already done. It’s about both retaining contributors and increasing their involvement. The more connections someone has with a project the more likely they will continue to contribute and increase their level of commitment. Someone who gets a few patches sponsored into Ubuntu but doesen’t get involved in the community may well switch distributions when they have an issue and take their contributions with them. Someone who becomes engaed in the community will stick around and help solve the issue.

We also need to be making it easier for contributors to navigate the sometime bureaucratic processes involved with getting patches sponsored into Ubuntu. Dan Chen has already talked about some of the plans coming out of UDS Barcelona, including a new Launchpad group to take on this issue, ubuntu-reviewers.

As every one who went to UDS is now home, I expect we’ll hear alot more about the discussions that went on there. I’m looking forward to it, and hope to find ways that I can help make the plans made there a reality.

So, if you’ve read this far, what are you’re thoughts? Have you contributed to Ubuntu? What was the most annoying part? Have you considered joining the Ubuntu Universe Contributor team? What’s stopped you?

Written by andrewsomething

June 1, 2009 at 1:15 pm

Posted in Ubuntu

Tagged with , ,

7 Responses

Subscribe to comments with RSS.

  1. Congratulations!

    Zac

    June 1, 2009 at 2:05 pm

  2. So when are you going to join Debian?

    Mark

    June 1, 2009 at 3:22 pm

    • That’s definitely something I’m aiming for, but I suppose I’m still a bit intimidated by the process.

      Honestly, as lame as it is, one of the things holding me back from starting the process is not having my key signed by a DD.

      Luckily for me, DebConf10 will be held in my home town. I’m hoping to find a way to contribute to the planning. It will be a great way to get involved with the Debian community beyond maintaining my packages and pushing patches back.

      andrewsomething

      June 1, 2009 at 3:38 pm

  3. I don’t really get the point of it. I intend to apply to MOTU by the end of the year, but I don’t feel like it’s necessary to join another team to just…keep doing what I already do. I can triage bugs, fix bugs, write patches, and generate debdiffs. I’d like to be able to upload on my own instead of annoying others to do it for me all the time, but MOTU, not UUC, would get me that, so…why bother with UUC?

    Mackenzie

    June 1, 2009 at 4:26 pm

    • Well that’s part of what I’m getting at really. The reasons for the program are somewhat unclear. Someone like you, who’s already an Ubuntu Member and well on the path to being a MOTU, it isn’t really worth it. The main reward to being an UUC is Ubuntu Membership.

      We need to make the goals of the program more clear, not just to contributors but amongst ourselves as well. Where I see the benefit is with drawing new people into the developer community. You are clearly already a part of it.

      ~400 people had uploads to Jaunty, and there are only 129 members of ubuntu-dev. UUC should be a way to acknowledge the contributions of those who aren’t as clearly on the path to MOTU and hopefully get them to consider becoming more involved.

      andrewsomething

      June 1, 2009 at 4:46 pm

  4. Do you have good summary statistics as to the size of the Universe developer/contributor pool and the workload at each step? I haven’t seen an effort to summarize what the Universe workflow actually looks like in terms of resources and throughput.

    Trying to visualize the process as a sort of production line process might help you understand how better to organize resources for best throughput efficiency. It could be the UUC program is actually working more efficiently than other steps in the process. In a production line capacity a highly efficient component looks idle because its waiting for other components to send it work. What feeds into the UUC program?

    I don’t think I can offer you any insights from examining Fedora’s contributor stats, as the workflow is completely different. But I will say that I think your instinct about having a top heavy distribution of contributions is something you should worry about. But you have to be careful how you read the top contributor information.

    How many of the top contributors are Canonical employees or otherwise paid to work on Ubuntu? From a contribution perspective you should try to separate paid contributors from volunteers. The volunteer contributors are going to have a very different carrying capacity than the paid timers. Try to build up histograms of the contribution distribution from paid timers versus volunteers at different points in the process. Don’t try to be perfect about the separation, but a best effort will help you when interpreting what your volunteer workforce is actually doing and where you need to help them.

    What you want to see is a much small mean/medium workload in the volunteers compared to paid timers but a much longer long tail at every point in the process. Whether the full timers dominate the raw contribution numbers isn’t really the biggest concern, not initially. The more important question is
    the volunteer contribution distribution significantly shaped differently at different points? That will help you find potential bottlenecks. bottlenecks which squeeze the distribution cutting off the long tail or pushing workload uphill.

    Once you get a feel for the volunteer contributions you can really start to have a discussion about how you want to optimize that distribution at each step in the process..and develop expectations on the level of time commitment you expect volunteers to be able to put in on average. And then experiment with optimizing workflow based on that expectation.

    -jef

    jef spaleta

    June 1, 2009 at 4:55 pm

  5. Congrats, Andrew!

    Peng

    June 1, 2009 at 8:08 pm


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: