So I’m back at my hotel room very late after the second day of CLS, and it’s been a really great event. It’s the first time I’ve been here, but others have told me it’s grown a lot from last year, and there have been a lot of really useful sessions. It’s been great to talk to other people who work with communities about some issues I’ve been trying to figure out. I would definitely recommend CLS highly to anyone who wrangles a community, there’s just a ton of valuable stuff going on here.
I kicked off day #1 at a session on how to be wrong gracefully, which is something that comes up all the time in the Fedora community. One of the most interesting points that came up was a discussion about how the different forums for feedback can work differently, and some can be better than others in this regard – it’s very hard to carry on a coherent conversation between multiple parties on Twitter and make sure they’re all involved with all the replies, for instance, which is much easier on a forum.
After that I went to a licensing roundtable hosted by Gervase Markham, which became a pretty freewheeling Q&A / debate session with Gervase as the one-eyed man in the kingdom of the blind. (Notes weren’t taken for reasons of prudence, in case anything sensitive came up). Gervase answered a question I’ve always had about the GPL: apparently if a company modifies and uses a GPL’ed codebase internally, the corporation is taken as a single legal person to whom that codebase has been distributed, and is not considered to have distributed it to the legal persons of its workers – so if you’re an employee at such a corporation, you aren’t personally considered a recipient of the code, and hence you don’t have the right to access the source or redistribute it yourself. There were some other pretty interesting points about the possible consequences if Google were to lose the Google vs. Oracle Java / Android lawsuit (all header files would be subject to copyright!), and the patent license grant clause of the GPLv3.
Next was a session proposed by Evan Hamilton on ‘what we all do every day’, which turned into a discussion on how different the roles played by people called ‘community managers’ can be. One really useful thing that came up was Jeff Potts’ conception and categorization of the various different things he does, which he expanded into a blog post due to overwhelming popular demand.
Jane Wells and Chelsea Otakan ran a session on providing recognition and incentives to community members. This is something we’ve been thinking about for the Fedora QA community so it was really interesting to me. There were quite a few people in the session and it seemed like we were all struggling with the same problems – handing out rewards to community members can be divisive, can result in people starting to game the system to win the rewards (especially if the rewards have substantial value), can turn into a major complex management project, and can even be counter-productive – Stormy Peters has a great take on this called would you do it again for free?, which was brought up in the session. In the end we seemed to agree on a few things. One was that the best kind of rewards are contextual – things that are only valuable to people who are actually invested in the project. The classic example is a project t-shirt: if you don’t really care about Fedora, you probably don’t want a Fedora t-shirt, and the fact that it doesn’t have any real significant intrinsic value lessens the negative effects rewards can have. Also, often the best reward you can give is simply to say thanks to people, in a genuine and personal way: it provides a great ‘warm fuzzy’ effect and avoids all the negative consequences that can be associated with more significant and tangible rewards.
Finished off day #1 with a session on gender issues led by Louis Suárez-Potts which unfortunately doesn’t seem to have been noted, but which was really helpful – I’ve been looking for ways to be more proactive in bringing more women into Fedora QA, and I got referred to Christie Koehler, who is working on organizing events to try and help women get started in F/OSS projects; I’d love to get Fedora and QA involved in that as target projects. (Are there any Fedora women in Portland who’d like to help out with that? Talk to me!)
I kicked off day #2 by doing a somewhat modified version of my presentation on principles and lessons learned from doing Fedora QA, It boots, ship it! – CLS notes here – trying to tailor it to the CLS style (which is more based on roundtables and group discussions than presentations) and to an audience which was kind of split between people interested in the general community management principles and people who were interested in learning more about the specifics of QA. It was a struggle to resolve that split at times and I felt like I let some of the people who came along down a bit, but it seemed like some people found it valuable, which was great, and it was nice of people to show up!
Next came a session on relaying user feedback, mediated by someone who seems to be referred to variously as Todd Gage or Todd Guse from Slide, neither of whom seem to exist on teh interwebz. This turned into a really interesting bridge-building exercise between those of us who work in communities that are open by default, like F/OSS communities, and people from corporations which have always had very closed and separate relationships with individual customers. Something I hadn’t really known before is that often the drive to keep feedback private in those communities comes from the ‘user’ – the customer – rather than the provider, and the corporate provider might actually want to open up their feedback processes more, but struggle to convince their customers to buy into that.
Koray Löker ran a great session on the challenges of including non-native English speakers in communities, which was a neat example that a session doesn’t have to be big to be productive – it was mostly a three-way discussion between Koray, myself and Cedric Thomas of OW2, but it definitely helped us all to formulate some basic principles on another topic that’s very relevant directly to my own work – we currently don’t do a lot actively within Fedora QA to try and make our events and activities available in multiple languages, for instance, and we may be able to do a better job of making our English discussions and documentation accessible. I know I often write unnecessarily florid English, which isn’t helping non-native speakers much!
I finished off at a session run by Alex Lefebvre (who I’m also having trouble finding) on turning users into active contributors, which turned out unusually. Monty had a strident take on the issue – he believes that a large majority of members of most communities will always be users who don’t contribute back, and it’s better to focus your efforts on enabling the minority of users who do contribute to contribute more effectively, and growing your user base in order to make that minority larger in absolute terms, rather than trying to make it larger in relative terms compared to the majority of users who do not contribute. MJ Petroni (sorry, that’s the best link I could find), the principal at causeit, disagreed, we got stuck on the disagreement for a long time, and the debate was about the most heated of the whole conference that I saw. In the end it was Cedric who managed to unpick that MJ was particularly unhappy with Monty’s choice of the term ‘leeches’ to describe the users who don’t contribute actively, whereas to Monty this was an unimportant label, and that helped to get past the roadblock – we thought it was interesting that a lot of the stuff we’d discussed right before in the non-native English speaker session had come up immediately afterwards, with Monty (a developer type for whom English is a non-native language) and MJ (for whom English is his native language, and terminology is very important) stuck on opposite sides of something of a linguistic and cultural divide!
Outside of the conference sessions, I had a great time chatting with Sumana, Jono (he’s famous, he doesn’t need a link), Dan Allen (come to FUDCon, Dan!) and way too many other people to list out. I also think I’m finally going to wave the white flag, and try to sort out a system where I can use twitter, identica, and google+ in some sort of organized federated way, strictly for Fedora communication purposes. It’ll probably involve Bitlbee, because let’s face it, Bitlbee is freaking amazing. Watch this space!