This week Jono Bacon announced that Ubuntu Developer Summits will become a series of online events. Having thought about it a few days, I’m now ready to input my own opinion to the discussion.
In the announcement Jono lists openness, transparency and accessibility as the major goals of the Ubuntu Developer Summits (UDS). The decision to move to an online event is supposed to improve these.
In this article I will explain why I don’t think it will, and why the new format looks just another Canonical team sprint. I’ll also cover some of my concerns over the accessibility and equality of the new format and important things I think the online events will lack, but shouldn’t. I will also discuss some of my opinions on how this changes the nature of UDS and the meaningfulness to flavors as well as how this change affects the Canonical-community relations.
Openness, transparency and accessibility
I value openness and transparency very much. However, I’m not sure if the new online format will increase them in any way. Canonical is already making many decisions that affect the community in-house, behind closed doors. Unfortunately, these decisions and their results don’t always roll out as expected.
Until now the community has had at least some voice in the discussion, or at least have been able to react to changes, if they have been discussed in UDS. It’s a pessimistic view, but I can’t see why Canonical employees would start discussing and working more closely now with the community on decisions that have already been made. Now that the event is online, it’s much easier to have more discreet, even hidden meetings about things where community input would be largely unwanted or at least ignored à la the in-house pre-UDS sessions.
While it’s evident that a physical event has limitations when it comes to accessibility, so does an online event. One of the main reasons that made the Canonical-community communication so effective in UDS was that everybody got together, in the same place and at the same time. This means everybody participating in person is (ideally) always available and 100% focused on the sessions.
Now that the event takes place online, people will not be as committed to participating a session in every slot. Even when they are participating, the session might not get their full attention, because there are more distractions when working online.
When people aren’t participating in a session, they most probably won’t be available, which means there won’t be any substitution to the off-session discussions. This is sad, because this used to be one of the most powerful features of UDS for officially recognized flavors (such as Xubuntu) and I believe for numerous other teams as well. Even if people did have these off-session discussions, they are just as useful as contacting anybody via IRC or email any time, so there’s no real benefit of the centralized time. As a matter of fact, people might be even busier than usually during the online events, making it even harder to get a hold of them.
Concerns of equality
One of my worries is the level of openness UDS can create while being an online event. There are a few issues which are raised from the community already, but I want to repeat.
The platform itself is one of the major concerns for the community: many of us aren’t comfortable using a proprietary platform. While it’s an easy choice for Canonical to deploy, it’s also a weird move from a major open source project.
Due to the technical limitations of Hangouts, there can only be 10 people who are videoconferencing at any time in a given session. While there are some similar limitations even in a physical event, I think the online event makes the thresold to join those 10 people higher for a few reasons. First, not all of us are comfortable with videoconferencing for various reasons. Some of the people in the community don’t want to be on the spotlight or are too afraid to do that with total strangers, others might not want to be streamed live.
While 10 is more slots than the amount of chairs we had for important people in UDS Raring, these slots are now assigned by the session leader to people in order of importance that they think the specific people will have on the meeting. My concern is that this might make it harder for some people to be given slots, including those that are not generally well known in the community or people who are known to strongly disagree with the session leaders. We will have to see how this works out in the forthcoming online UDS, but before it’s over, I’ll keep my reservations.
Naturally there are other ways to communicate as well, namely IRC and Etherpad. To be fair, the communication via those mediums worked well enough in UDS Raring, if the session leader simply wanted (think: remembered) to use them. Taking this into account, I don’t really think that taking the event online really makes it any more accessible for anybody. On the other hand, it still doesn’t offer a feasible alternative for taking part physically or being one of the participants in the videoconference.
As written by Jono in the announcement, Canonical wants to make the event as accessible for anybody with a decent internet connection. However, based on the facts above, the equality is ironically mostly virtual. The format change might bring accessibility and some real equality between participants, but in its entirety, I think it’s a step backwards. Those who participated in person got unbeatable benefits and even if they paid for their flights and accommodation themselves, a fair substitution to their money.
Letting the social aspect go
As I mentioned before, off-session discussions with a variety of people were one of the best features of UDS. Those who have attend UDS know that these discussions go on even after the actual conference for the day has ended. Making new contacts and having interesting and important discussions went on all the way to the last drinks of the evening. It’s silly to expect people to take this much time off of their normal schedules and try to socialize online in any way that resembles the socializing in a physical event. This feature is gone.
Losing the social aspect of UDS doesn’t only impact productivity in during the event. One of the unwanted side-effects is less and worse human relations between Canonical and community teams. Meeting people face-to-face and getting to know them behind their IRC nicks and email addresses makes communication in the future smoother. When these personal relations do not exist, cooperation will be much harder and slower in the future, including ever-so-important dispute resolution.
Ultimately, I believe these changes will lead to lower work morale and loyalty towards Canonical in the community. Whether it happens in critical parts for business from the Canonical point of view is a different story.
For many of the volunteers in the community contributing is also about fun. There aren’t many better things than seeing an old friend after a long while, chatting about something you both are passionate about and then have a drink. As mentioned, the new online format is taking lot of this fun away from the community. One of the things that I predict to happen is an influx of people to other conferences like DebConf, especially amongst those people for whom Ubuntu is only one part of their open source ecosystem. It’s impossible to foresee the actual implications for UDS and Canonical, but ultimately one participant lost is one opportunity lost.
Business before community?
As several people have said, this change along with a rolling release are probably good decisions for Canonical as a business, but not Ubuntu as a community. Making UDS an online event will not only cut quite a lot of financial slack, it also lets them focus on the things that make sense to them financially. Ultimately Canonical gets more control over which matters are discussed, which opinions are given voice and who communicates about them in their own events.
In addition to losing the social aspect and Ubuntu being the thing that made the people come together, there are possibly some caveats to Ubuntu development as well. In the worst case scenario, Canonical will lose a lot of contributors from the community, and I can’t see how this wouldn’t affect Canonical’s business too.
The last cautionary sign for the openness, transparency and equality is the lacking and too late communication about changed plans for UDS. Because of this, the first online UDS arrives too early and unexpected for the majority of the community. The fact that majority of the sessions are going to be revolve around the controversial discussion about moving to a rolling release makes the situation even less bearable for the community. The chosen way of communication backs up the vision that these changes are purely business-oriented. Unfortunately, it isn’t very fair for the community.
Furthermore, this is not the first time the communication from Canonical has been lacking. Numerous people have expressed their concerns and distrust in the past about Canonical’s way to communicate about changes in the infrastructure. So far Canonical has avoided the worst damage, but the atmosphere is starting to get tighter. It’s more likely than ever that people will part ways with Ubuntu in a way or another. If people leave now, it’s going to be really hard to do significant damage management and regain trust from the (former) community members.
Impact on flavors
Due to the various reasons mentioned before and becoming a way more Canonical and Ubuntu OS-centric, UDS is becoming borderline irrelevant and useless for flavors. From what I’ve read in the last few days, I believe this is the unanimous response from the flavor teams. This is worrisome for the flavors and the community at large because it is possible that the taken direction excludes the community from decision making that concerns its own infrastructure as well.
It is too early to say what kind of actions several flavors will take and what their position will be, not least because the uncertainty about a rolling release model and its implications. However, there is one thing that will most probably happen: since the flavors are slowly and involuntarily sliding off of UDS, it will be less and less meaningful to organize their sessions parallel to the “main” virtual UDS. Once this happens, the future of the flavors is even more full of justified fear, uncertainty and doubt. The only thing Canonical can do to reduce this effect is start communicating and discussing about the changes that affect the community truly openly.
What’s in the future for us?
The future for UDS is still partly out of sight. In the announcement Jono somewhat vaguely says that after two online events Canonical will “review the success of the next two online events” and “assess whether to continue the online format”. Whether it means Canonical will either get back to the physical format or drop UDS for good if the online format fails remains to be seen.
Until the first online event is over we can’t evaluate how the new format works. However, it is clear that Canonical needs to get the community involved in the event if they wish to keep the community close to them. To be succesful in this, Canonical needs to make sure UDS is equal to all participants. If they can truly integrate the community to the discussion during the first event they will be able to regain some trust and hope in the community to their communication. Ultimately the community can only consider the event succesgull if Canonical is able to achieve all the aforementioned goals as well as the major goals they have set for UDS.
At the end of the day, we have to remember Canonical is a business trying to make money with their product. Keeping this in mind and as an entrepreneur myself, I can empathize with many of their decisions. In my opinion the partly harsh criticism is justified though. Since Canonical has promoted Ubuntu as a community-oriented project from the beginning, it would be only fair to actually involve the community in the decision making and discussions well in advance. If they do not wish to do this, they should clearly communicate this to the community and their users.
I would like to thank the following people for their support, numerous discussions and influence to this article: Elizabeth Krumbach, Scott Kitterman, Micah Gersten and Jussi Kekkonen. Thank you!
This article is part of the article series A day in an open source project.