Rethinking linux.conf.au

Linux.conf.au has long been one of your editor's favorite events anywhere in the world. It typically features one of the most diverse and interesting programs and is hosted in a different city every year. And the whole thing is fueled by that classic Australian energy and humor — even when it is held in New Zealand. With well over a decade of history, LCA seems like a solid and well established event. So it was a surprise to run across a discussion suggesting that there might be no LCA in 2014. LCA, it seems, has found itself needing to rethink how the conference is organized and run.

Linux Australia council member James Polley started the discussion with a post on the current status of LCA 2014:

James noted that LCA was once the only major community conference in Australia; now, instead, there are several. Perhaps, he surmised, there is no longer a need for LCA? Or, perhaps, it is time to move from volunteer organizers to a professionally-managed event? Or, perhaps, it's time to take a break and see if any interest develops for 2015?

Participants in the discussion raised a lot of concerns that the conference has simply gotten too big and complex. Potential organizers, they say, are being put off by the sheer time commitment required. Some past organizers (such as Russell Stuart, Brisbane 2011) disagreed, saying that the actual time required is not as much as it seems. But there is no denying the fact that LCA organizers tend to look awfully tired and haggard even at the beginning of the event and thoroughly fried by the end. Putting together a conference like LCA is a lot of work.

So it is natural to think about ways to reduce that work. Perhaps LCA should go back to being a smaller event? There were proposals to reduce the number of talk tracks, eliminate various social events, and even to drop the 1-2 days of miniconfs that precede the conference itself. LCA did not originally include miniconfs; they were first added by the Brisbane team in 2002. But the miniconfs have since become an integral part of the conference. Their contents are not under the control of the program committee, with the result that the areas covered — and the quality — vary widely. But the best miniconf talks tend to be quite good indeed, and the miniconfs serve as an important entry path for speakers trying to get into LCA proper. It would be a shame to lose them.

Another idea that came up was to settle down and have the conference in the same city every year. That, in your editor's opinion, would risk repeating the story of the Ottawa Linux Symposium. There is a long list of reasons for that once-dominant conference's decline, but one of them was certainly the organizers' unwillingness to move the event to new locations. Even a city as nice as Ottawa gets a little tiresome after several years in a row. A new location every year helps to keep LCA fresh and interesting.

The volunteer organizer model also helps in this regard. LCA has managed to evolve a mechanism where each year's team is given a great deal of freedom in how it runs the conference. Behind the scenes, though, a "ghosts" committee (made up of prior organizers) oversees the effort, provides advice, and sounds the alarm when it sees something in danger of going wrong. The end result has been a conference that is, in some ways, new every year, but which still runs like a smoothly oiled machine.

A shift to a professionally-organized event might take some strain off the volunteer organizers but it would have to be done carefully if it were not to kill the magic that has made LCA such a good event for so many years. That would not be impossible to do; the Linux Plumbers Conference has thrived with a great deal of organizational help from the Linux Foundation. Such a setup requires professionals that are willing to defer to the "amateurs" for most of the important decisions; it can be done, but it's not something that happens by itself.

Donna Benjamin (Melbourne 2008) thinks that workload issues could be addressed, perhaps with the help of professional organizers and a team that is distributed across the country. But, she says, there is another, more difficult problem: the fact that the organizing team must sign up for a lot of criticism from the community.

This sentiment was echoed by a number of other participants in the discussion. In our community, it seems, no good deed goes unpunished; even an event as well run as LCA is going to draw its share of complainers. When a difficult job starts to appear thankless as well, the number of volunteers is certain to decrease. But potential organizers should also heed the words of Andrew Ruthven (Wellington 2010):

Finally, one could also argue that most conferences have a limited lifetime. Linux Expo and LinuxWorld are long gone. Even the much-respected Linux-Kongress, arguably the first Linux conference, was last held in 2010. LCA, having started as the Conference of Australian Linux Users in 1999, has certainly had a long run. Perhaps LCA, too, is reaching the end of its life span?

Your editor does not believe that to be the case. We are not witnessing a conference heading into senescence; instead, it is a middle-age crisis at worst. There is too much that is valuable and unique worldwide in LCA, and the people who attend the conference every year clearly appreciate it. LCA can be seen as a sort of free software project that, after years of success, needs to reevaluate its processes and governance. Once that task is done, LCA is likely to be stronger and more vital than ever.

For 2014, the deadline for bids has been extended for a few weeks, so there is still a chance for interested groups to sign up for a chance to host the event. There is talk of putting together a distributed team that, most likely, would propose to return LCA to Sydney. One expects that somebody will step up to the plate and make the event happen; who knows, perhaps 2014 will be the year that LCA finally is held in Broome.

---

(Log in to post comments)