On Wed, Mar 19, 2008 at 7:45 PM, Gervase Markham <[EMAIL PROTECTED]> wrote: > Frederik Ramm wrote: > > Those who make the decision have no authority anyway. So why not > > record the rationale without the decision, as it is anybody's free > > choice to follow them or not. > > Then welcome, everyone, to the world's first anarcho-collaborativist > project. > > It's an unusual model you are suggesting, and not one which (as far as I > know) has worked anywhere else. Still, I guess we could try it. After > all, it's not like this project is important or anything. > > _Every_ project has authority. It's just a question of whether that > authority is recognised and controlled, or unrecognised and random. If > you say "there is no authority", then in fact authority rests with those > who control the servers, or the root passwords, or those who have more > time to make the database conform to the way they think things should be > (e.g. the Cyprus situation where, so far, we have abdicated authority). > This is generally not a good way to allocate authority, because the only > way the authority can be exercised is "Do what I say, or I'll take my > ball and go home". > > In a standard open source software project, authority is (roughly) a > meritocracy. Authority flows to those people who contribute and who > demonstrate themselves competent. This model works quite well. It seems > like an obvious model for OSM to adopt, and it's the one I _thought_ we > were adopting. But hey, maybe I'm wrong. Maybe the authority being > exercised by the OSMF in working on the licensing problem is an > illusion. Maybe a project newbie should have just as much weight in > making decisions as those who've been mapping for years.
This isn't a software project Interestingly much of the software development is run as you suggest. It tends to be "owned" by people who look after it, and obviously we have the guys with admin access to the actual servers who do their best to keep everything running smoothly and will therefore act as your authority when it comes to those systems. But once you get past all the software and the basic data model, there really is no "authority". Anarcho-collaborativist would be a pretty good description I think. I doubt it's the world first though. Granted there are certain pieces of software/presentation which will provide controlling/driving influence on that, but as anyone is free to generate their own replacements for these, this will in some regards be diluted as OSM becomes larger. However, all the notable drivers have been themselves heavily influenced by the data actually put in. So yes, obviously people who contribute more, or at least people who change things more, are going to have a bigger influence. And yes, the server guys could install technical measures of control, not that they have or necessarily would, but like all opensource there's always the possibility of a fork (even if at present it would be incomplete in the history department). And also there will be some people who are more influencial than others. A new user is unlikely to have as easy a time trying to generate a new tagging scheme and getting people to use it than an experienced mapper would. This is largely all normal social stuff. So when a group of 5-15 people on the wiki all vote for a feature and "approve" it, it generally means absolutely nothing. Tagging lives and dies by it's usefulness, by how much it's used, and by whether anyone actually producing an end product decides to use it. So if there's any authority it isn't in the approval process. How this whole approach can work in the face of continual targeted vandalism/conflict has yet to be fully explored. I can't say I have the answers to that, but so far these kinds of problems have been suprisingly rare. Wikipedia's answer is a load of rules, largely overseen by a ruling group with all the authority, and a pile of technical locking/blocking measures against particular users/ips/articles... and if you go down that route it's a big snake pit of politcial problems. Dave _______________________________________________ talk mailing list [email protected] http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
