On Wed, Jul 1, 2009 at 8:32 PM, Jimmy Wales jwales@wikia-inc.com wrote:
I'd be interested to see your positive, assume-good-faith list of suggestions.
One of my favorite suggestions, from Erik, is that we use IdeaTorrent ( http://brainstorm.ubuntu.com/ ) in order to provide a single place for users to engage in very important discussions about all manner of issues relating to the community. Right now there is simply no way for our widely disparate community of users, who have expertise in every area imaginable and whose collective input is extremely valuable, to come together and have a conversation. It takes someone such as yourself to champion an idea to the community and present it in its best form and make sure that the best of the arguments from both sides are heard from users.
As the projects have grown and as they have become more centrally managed in a top down fashion it has become increasingly difficult for ideas to percolate from the bottom up. How can a user with a great idea on one wiki present it and be sure that users from the other wikis and the WMF see it? Likewise, how can the Foundation ask questions of the *entire* community? Neither users or the Foundation have a voice that can reach everyone (fundraising and the like are an exception). There isn't a plausible conduit through which we can present and receive ideas and those ideas are considered on an equal basis with all other ideas and then refined and improved by the will of the community and ultimately implemented (by a volunteer or the WMF).
Regarding the software, I think it's great to hold a conversation on wikitech-l about the best way to replace ParserFunctions. Of course, the way we got ParserFunctions was through a conversation on wikitech-l which entertained a few ideas but ultimately did not have the wider goals of the community in mind due to the narrow scope of the discussions. Usability and encyclopedia writing were not concerns, CPU cycles was. The justification was, and continues to be, well, there is obviously a problem here. Therefore we, the code writers, have free license to develop a new solution, ask our friends in IRC if it looks nice, and then put it on the live sites. It's not even clear how you could extract a consensus from wikitech-l if it were there.
If you take fully consulting the community consensus seriously then there is a very different design model that then leads to development. In this method we have a plausible way of asking a large number of *editors and users* what is wrong with the software. You have to get many of the people who actually edit the encyclopedia a lot and have something to say about what's wrong with it and what's right with it in the same place fully engaged with each other. Right now we do not have tools that facilitate this. Article talk pages are simply not it. Meta is not it - the people aren't there. The mailing lists aren't it - the people aren't here. We represent a tiny minority of the community and a minority of the total number of people who would, if they were afforded the opportunity, have an opinion worth hearing. If you look at the number of people engaged in any conversation which will have a serious impact on all of the projects and then compare that number to any measure of active editors and contributors you will see that it is shockingly small. I encourage the WMF to make that ratio as large as possible, and I suggest that the larger you make it the more we will all benefit.
I get the feeling that many people look at full consultation as a lot of really hard work. I think that's wrong - we are supposed to be leveraging the power of communities. The WMF has the power to enable a community to come together and form a consensus by bringing their attention all to the same place. I think that until something like that happens full consultation is more of a dream that many people aren't even trying to realize and changes will continue to be made to the software and otherwise which aren't really in the right direction. For example, it's not clear to everyone that Wikipedia even needs a programming language. I don't know if it does or not. There are a lot of things to take into consideration, such as usability, readability of the main article namespace, duplication of content, ease of more sophisticated editing, and issues you or I might not even think of! Adding a programming language is not a magic bullet to these issues. It could in fact be that templates and the way we work with article content in the first place needs to be entirely rethought. This is not a conversation that should be limited to wikitech-l. In fact, editors might have a more useful opinion. But in the current system their opinions won't be sought out as the decision to do it was entirely top down.
Lastly, I do not consider a wide distaste for the look of ParserFunctions to be a sanction for a new programming language. ParserFunctions was added because a few users decided to abuse templates in order to get if-like functionality, but it was not done in consultation with the community. Thus, at the very least, I suggest we go back to the community and ask them, from the perspective of just having vanilla templates, how we can improve the software to make Wikipedia and the other projects better. And I suggest that we solicit feedback from as many people as we possibly can - fully consult the community consensus!