I believe we should radically change the concept of FUDCon.
(And if you think this post is looking pretty lengthy, the short version is this: ONE EVENT TO RULE THEM ALL.)
I’ve been to a number of conferences in the past few years since joining Fedora. It’s a grab-bag of “types” – conferences like Southeast Linuxfest, SCALE, and LinuxFest Northwest, which tend to be free or almost-free, and tend to have more of a community feel; larger-scale, more commercially oriented conferences such as LinuxCon and OSCON; and conferences that are organized more around a singular group, project, or common interest — FUDCon is certainly an example, but also things like Community Leadership Summit (common interests/problems), and the OpenStack Design Summit & Conference.
In the latter example, particularly with project-focused conferences, the face-to-face time amongst project members is absolutely valuable. It’s the place where contributors can get make decisions, do planning, and generally get things done, in a very high-bandwidth fashion. And I think “planning” is really one of the key attractors. The OpenStack Summit, for example, is held right after release – and is the place to truly trot out ideas, gather around them, make a plan, and start breaking it down into how it will actually get done – over 3 or 4 days. And it is not a place to “show what I did” – it is truly a “where am I going and how is the work going to get done and how does that intersect with other areas of the project” type of event.
I guess I know a thing or two about FUDCon planning; since organizing the FUDCon in Tempe, I’ve been helping out in some way or another with nearly every FUDCon. And thus, I’m going to present the following observations:
- We tend to do a lot of “what I did” or “how this thing works” at FUDCons – and not a lot of planning.
- Hackfests – which gather together specific contributor groups – tend to not always be well-organized, or focused around “let’s finish this thing we are working on.”
- FUDCons are not scheduled at times which are obvious “planning points.” FUDCon Lawrence, for example, will be several months into the release cycle – not an incredibly amazing time for planning around F19.
- We put some focus and effort into the U (users) at FUDCons – which, while valuable, does not require having dozens of contributors present, nor does it make the best usage of the face-to-face time that could be used for actual teamwork.
- 4 FUDCons per year means that, as a worldwide community, we don’t get to get entire teams together.
The latter point is particularly interesting (and has given me a lot of heartburn). While we tend to have more planning and hackfests at the North American and to some extent, the EMEA FUDCons, the extent of teamwork and planning done in APAC and LATAM tend to be gathered around regional ambassador leadership, and folks working on translations in that region. Most of the project teams tend to be distributed globally; people want face to face time with their teams, and we simply can’t haul in everyone from everywhere in our current model.
I’m a true believer in planning and execution. A lot of this probably comes from my work at Intel in strategic marketing — Intel is absolutely relentless in its planning cycle, but the focus on planning and setting goals is what drives innovation forward. It encourages people to think big, and imaginatively; it helps to lay out a roadmap of milestones and tasks to a goalpost in the future.
And I think the model of bringing together a global community at an appropriate point in a release cycle to gather around planning and execution, rather than showing off what we did in the past and maybe working on things we already have in the works, is one that will drive Fedora forward.
What I would like to see is the following:
- One event per year. Starting in North America, and possibly alternating with other regions. Starting in FY14 (that’s March 2013 – Feb. 2014, for those who don’t follow ambassador finances.)
- Get people from other regions to that event. Not “one or two from other regions”; I’m talking about getting engaged contributors with concrete plans and/or demonstrated history of contributions face to face with their teammates. So that that team can get things done, contributors can be part of the planning, take ownership of tasks, and not feel like they’re leaving out a significant portion of their community.
- Have it at an appropriate point in a release cycle, where we, as teams or subprojects or groups or whatever you want to call it, can take advantage of the length of time before us to think about what we can accomplish over the next 2 releases, plan out activities and tasks, etc.
- Perhaps move barcamp to the end, and have pre-scheduled, well-organized, planning/team meetings at the beginning. Yes, I know this is probably giving some of you fits. Here’s why:
- Barcamp sessions tend to be more around “I want to share this cool thing” – which is sometimes an idea, but more often around “learn how to use this thing I already implemented.”
- It would be an awesome time to actually share what teams are planning and have accomplished during their time together.
- Y’all are beat by day three, which I think is part of why hackfests wane a bit on the last day. Oh, did I mention that I think we should move to a longer event? I’ll do that now.
- More days together. Possibly straddling a weekend to reduce the drain on everyone’s “days off work” time, maybe not. But we’re already travelling – and the costs of airfare tend to be higher than the costs of hotel, particularly when hauling in people from all over the world – let’s make the best of the effort spent getting to the event and make it longer.
- Consider sharing this event with other project communities – for multiple reasons:
- Leveraging the buying power of more attendees
- If we’re already planning something – why not let others benefit from some of the planning we’re doing, and offer their community a way to get together in a similar, planning/doing-focused fashion?
- It’s a great way to cross-pollinate between upstream/downstream communities – though we’d probably want to make sure we’re not going to lose focus from participants. (Much like when we have had a FUDCon run parallel to a large-scale more general community conference (that is not focused on planning, but more on how-to’s and usage – where people really want to learn about stuff, but also want to focus on the project in which they contribute.)
- Attract more sponsorships because of a more diverse audience. Money is nice. It pays for food and things.
- Make this event be focused on the “do-ers” – and not the users. I mentioned previously in this post that it does not make the best use of our face-to-face bandwidth, and I’m sticking to that — and moreover, I think that trying to plan a parallel “user track” just winds up taking people away from getting things done. This is not a “we don’t care about the users” statement in any way, so don’t jump down my throat. But I think that mixing up the event tends to leave casual users/potential users/non-contributor users unsure about what to attend, and I haven’t seen any evidence on any large scale that users magically become contributors at a FUDCon. And there is NO REASON IN THE UNIVERSE why we can’t come up with a type of event that costs significantly less to host, requires fewer numbers of contributors to attend, and is geared solely towards users/potential users/potential contributors, and can be made repeatable in many places. The fact that a FUDCon in Pune can draw in a crowd of 500+ shows that there is absolutely interest.
You’ll probably notice that I just used the word “event” a lot, where I might have used the word FUDCon previously. (FUDCon, for those of you who have come this far without wondering what that acronym is, stands for Fedora Users and Developers Conference.)
I envision this to truly be an event of the do-ers – people who do things, get things done. And I’ve mentioned before the funny thing about how the word “do” is right in the middle of the word Fedora. A new type of event – with a renewed focus and purpose – particularly if it becomes more diverse than just us – needs a new name.
And to answer your burning question, because I can reeeeeeeeeeead your miiiiiiiiiiiiiinds: Why, yes! I am aware that this will cost a crapton more money. Bringing in contributors from other regions costs more than if we brought those contributors to a FUDCon in their region – and thus a DoCon, or whatever we might call it, would cost more than the entire 4-FUDCons yearly budget combined.
Is the cost justifiable? I think it definitely is. Will we accomplish more at one worldwide DoCon than we could at 4 FUDCons? I believe we can. Do we have to start thinking about that now? YES.
We are getting to the mid-way point of F18; FUDCon in Lawrence will be mid-through 19. I would expect that we would quite possibly initiate this at the beginning of F20. TWENTY, folks. That is a lot of releases – where we have done truly groundbreaking, innovative work.
We have amazing, talented, engaged contributors in the Fedora Project. And I believe that focusing on the future of Fedora at an event where we have gathered contributors from around the world – planning where we can go and what we can accomplish over the next 2-4 releases, scoping out tasks, executing to plan, and really, dreaming bigger – will lead us through our early 20′s to become greater than ever.