Brainstorm reconsidered?

Mon 10 November 2008

Jef Spaleta, Fedora contributor and member of the Fedora Board, offers a dissenting view on the value of websites like Dell Ideastorm for community based projects:

It makes some sense when you have dedicated engineering resources to spend and are looking for ideas to spend it on. But if you want to grow new volunteer involvement, I don't think the Idea Storm implementations we are seeing make sense for that. The popularity of an idea simply is not enough. We have to have a mechanism which helps individuals turn personal interest in to personal action...instead of encouraging them to wait for someone else to take their idea and run with it.

Ouch. He's not the first person to recognize there's something wrong with Brainstorm. My personal favorite is #11730: "Make developers pay attention to ideas on this site". The rhetoric of Brainstorm, inherited from Ideastorm, is that users submit ideas and developers implement them. The disconnect between this and reality is causing frustration. The Dell Ideastorm revolves around identifying items for Dell to spend money on to make their product more valuable to customers; Dell managers have a profit motive to pay close attention and allocate resources to make things happen.

In contrast, Canonical has a commitment to making Ubuntu available free of charge, which affects how the profit motive is expressed. In Canonical's focus to make development sustainable, their paying customers are OEMs like Dell, not end users. Under those constraints, they should reasonably focus on the set of items that both users and OEMs want (and might possibly pay for). Automated hardware change recognition and Deviant Art contests don't seem to quite fit that mold.

The rest of Ubuntu contribution is basically the fantastic result of an appropriate social contract. Volunteers seek to improve Ubuntu for their needs and those like them, and contribute that back to everyone. It's a classic stone soup story; the soup tastes great and everyone gets more than they gave. Such people don't need a website to figure out what they want to fix, they need guidance on how to fix it.

Jef's alternative suggestion is to use the votes to find clusters of people interested in a common set of goals, and give them a communication medium to make their goals reality. That sounds nice but I think it's a lot harder than it sounds to find a group of people who can get things done. MOTU has a series of failed failing focused teams: BitTorrent, Java, Games, Audio. Even the ubuntu-laptop team seems to have fallen apart, with no apparent leadership or goals. I don't think it's a matter of subscribing people who vote for a set of features to a mailing list.

From the outside looking in, I'd say that Debian has had more success with their focused team based efforts. Probably because in many cases, it was quite simple to organize a cluster of people interested and already taking personal action. There's probably a critical mass of core competence needed before team approaches can be successful; recruiting is important, but you can't do it without a group who really knows their stuff and can write down the important parts for those that follow.

Within Ubuntu, the OpenWeek is supposed to function like a cross between a conference and a Freshman Activities Fair at the beginning of the semester, and helps address the need for recruiting informed volunteers. However, I think the IRC format, as is, is not appropriate. The first half of most sessions runs like a file dump from a text file, and the latter half runs like a Q&A. The main purpose is to engage the wider community in the development projects within Ubuntu. The Q&A is really useful in getting people to ask questions and maybe even participate in development, but the IRC format hurts the presentation of information in the first half. If we want Ubuntu to further evolve beyond the command line, I think it might be more appropriate to record a 20 minute video presentation or tutorial as a prerequisite to an IRC Q&A session. Perhaps this is something Loco's can get involved in producing with developers?

Another approach may be to make sure high rated items on Brainstorm see treatment at OpenWeek. I imagine fixing suspend and resume would be a massively popular Q&A. Or a session on the highly popular idea, suggested for Jaunty: faster boot times.

Comments !