Re: [WikiEN-l] Eschatology and Wikipedia - Liquid threads

Maybe it would be better to try to define what the features we need for better communication are. I'm not quite awake yet, but here are a few thoughts: - Ability to follow conversations without having to jump around across multiple talk pages. - Ability to be notified of conversations that are interesting, for broad values of "interesting". (Conversations about watchlisted pages, in particular categories, or part of particular wikiprojects for a start.) - A single page "newsfeed" of comments that combines conversations we are following and topics we've defined as interesting, as well as optionally RC feeds. - Ability to comment directly from a particular change, which could supplement or replace the edit summary. This also gets a lot of conversations off the traditional user talk page, because much of what we bring up on a user's talk page is about some edit they made. Keeping this with the change in question makes it easier for others to join the conversation. - Complete replacement for threadmode talk pages, possibly to the point where the "talk" tab gets replaced with a "discuss" tab. - Ability to tie in conversations elsewhere seemlessly without duplication. For example, a conversation may start over a particular page, but the issue being discussed is really a particular infobox. A "cc" feature used judiciously (and possibly one that's editable) would help this. This would be most helpful if you could CC another article's discussion, a wikiproject, and/or particular individuals. - Whatever solution needs to eliminate archives and hatboxes as we know them now. If conversations are to "archive" it should be straightforward to restart them, and doing so shouldn't be considered a taboo. Liquidthreads is a start, but it seems to me like the mission was "figure out how to bolt forums onto an existing talk page", without enough consideration as to what problems need to be solved. It's an interesting hack, and it has let us experiment about what works and what doesn't, but without change it's not going anywhere. It's been my experience that talk pages melt down around 15 or so participants if the topic is contentious. A little below that and you can have a productive discussion, too much below that and you aren't getting anywhere. I think many of our best collaborations are small teams of 2-4 dedicated people, with a few more that come and go. Discussions between collaborators tend to fit into several categories: - One on one discussions between editors for mentoring, collaboration, or simply to find out "WTF?" - Asking "is this ok?" before making a change. (This is fine, but some of these would be better to just be bold.) - Asking "what was wrong with that?" after making a change that was reverted. - Small collaborations. - "Local" uncontroversial applications of policy and process. (Local, as in local to an article). The typical AFD falls into this one, with 2-20 participants. - Controversial applications of policy and process. Highly contentious AFDs, and virtually all requests for advanced permissions. Reason tends to go out the window on these, and the signal to noise can be really bad. - Global applications of policy and process. (Mass deletion, or controversial mass-changes.) - Dispute resolution and arbitration processes. - Project-wide discussions. Policy changes and the like. We need to have effective communication for all of these that can preserve a high signal to noise ratio, and that can scale well all the way from one on one to project-wide discussions. Right now, we're fairly strong at small discussions, but we break down horribly on big ones. It's also too easy to disrupt a large-scale discussion, and since we rely on consensus to make decisions, a little disruption can sometimes delay reaching consensus indefinantely, and can do so without crossing the thin line into what is perceived as misconduct. Also worth considering are social problems. Disruptive editing, personal attacks, privacy violations, and the like, although not technical problems, do have to be considered when designing discussion systems, because they are social issues that can be partly solved through good implementation. On Sun, Dec 26, 2010 at 7:37 AM, David Gerard <dgerard(a)gmail.com> wrote: -- Faith is about what you really truly believe in, not about what you are taught to believe.