Research Wiki pilot: Difference between revisions
(no "Wikipedians" please, we need people who are not globs of status-seeking-control-imposing-technology-driven shit to be in charge, to get anything done) |
(On the programming and data fetching side auto stub generation is one essential feature to pursue. ) |
||
| (2 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
===On starting the Research Wiki pilot=== | |||
It is clear that a [[policy]] that deals with deletions and such has to be formed. | |||
Research Wiki will have namespaces for [[Product]], [[Company]] and [[Group]]. [[Product Group]]s will be implemented as [[categories]] with nesting. [[Brand]]s may also be implemented as categories, this is yet unclear. | |||
A good plan for the order of adding things is needed. For example documenting (retail)[[channel]]s should be quite quickly brought to as close to 100% coverage as soon as possible since when adding products we will have to begin with a goal of 0,01% coverage. Another reason for this is that the [[Consumerium:intermediate product page]] links to channels where it is sold starting with one or more retail chains where a product is found. This is necessary for providing a view to which products in a certain store have articles on them by utilizing the special:whatlinkshere function. | |||
A simple plan for implementing what was in the old days referred to as [[The Consumerium Exchange]] is presented at [[Simple model for expressing opinions]] which can be implemented with only minimal [[MediaWiki modifications|additions/changes]] to the [[MediaWiki]] software. This plan must be meticulously checked before it is made into an official [[policy]]. It contains a few vulnerabilities to attacks, some of which can be detected and prevented (after or before they happen) and some attacks mostly diluting the reputation of the voting system by utilising [[sock puppet]]s. A more elaborate [[voting]] system may be later on implemented, but this is what we start with. | |||
On the programming and data fetching side [[auto stub generation]] is one essential [[feature]] to pursue. | |||
---- | |||
The [[Research Wiki]] is the easiest element of a [[pilot project]] to start. | The [[Research Wiki]] is the easiest element of a [[pilot project]] to start. | ||
| Line 25: | Line 37: | ||
These pilot wikis are small and don't have a lot of [[wiki management]] problems yet. The above are the minimum required to achieve at least near-convergence. Other structural stuff like [[best cases]], [[worst cases]], [[user stories]], etc., [[trolls]] leave to you to manage. It's probably best to avoid that structure for now - it only starts to matter once there is a very rigorous mission, that fails easily, or, far too many people involved. | These pilot wikis are small and don't have a lot of [[wiki management]] problems yet. The above are the minimum required to achieve at least near-convergence. Other structural stuff like [[best cases]], [[worst cases]], [[user stories]], etc., [[trolls]] leave to you to manage. It's probably best to avoid that structure for now - it only starts to matter once there is a very rigorous mission, that fails easily, or, far too many people involved. | ||
=== serious === | === serious === | ||
Latest revision as of 18:52, 21 September 2004
On starting the Research Wiki pilot
It is clear that a policy that deals with deletions and such has to be formed.
Research Wiki will have namespaces for Product, Company and Group. Product Groups will be implemented as categories with nesting. Brands may also be implemented as categories, this is yet unclear.
A good plan for the order of adding things is needed. For example documenting (retail)channels should be quite quickly brought to as close to 100% coverage as soon as possible since when adding products we will have to begin with a goal of 0,01% coverage. Another reason for this is that the Consumerium:intermediate product page links to channels where it is sold starting with one or more retail chains where a product is found. This is necessary for providing a view to which products in a certain store have articles on them by utilizing the special:whatlinkshere function.
A simple plan for implementing what was in the old days referred to as The Consumerium Exchange is presented at Simple model for expressing opinions which can be implemented with only minimal additions/changes to the MediaWiki software. This plan must be meticulously checked before it is made into an official policy. It contains a few vulnerabilities to attacks, some of which can be detected and prevented (after or before they happen) and some attacks mostly diluting the reputation of the voting system by utilising sock puppets. A more elaborate voting system may be later on implemented, but this is what we start with.
On the programming and data fetching side auto stub generation is one essential feature to pursue.
The Research Wiki is the easiest element of a pilot project to start.
Accordingly, many poorly qualified people with no clue about wiki management seem to be starting some based on the principles of GodKing control, sysop power structure, sysop vigilantiism and sysop vandalism: Consumerpedia in particular has accepted all these norms.
CorpKnowPedia seems to be run by an experienced activist and may do better. However, its objectives are more limited and interaction with it more arm's length. Thus, it is not really an adequate Research Wiki pilot either.
We all seem to agree that a close relationship with some external sources of Research Wiki input is desirable and necessary. The Consumerium:intermediate pages are too difficult to maintain alone with a development-focused effort like Consumerium. It needs some convergence:
convergence
Convergence is far simpler if some basic conventions are followed very early, and some essential concepts have the same names everywhere: buying criteria, comprehensive outcome, economic choice, moral purchasing, political consumerism, fair trade, safe trade, and externalities like deforestation, extinction, slavery, etc.
From Consumerpedia point of view, different GFDL corpus might well be appropriate, and they might not always converge with those of Consumerium itself. But, it's easier to track that by making sure that essential projects have at least the most basic terminology in common, e.g. thaten: Consumerpedia: "we",en: Consumerpedia: "Policy" and en: Consumerpedia: "itself" have the usual meanings. An interwiki link standard would be handy about now!
Such wikis all need to get off on the right foot: Gus Kouwenhoven (perhaps CIV should be high on the investigation priorities list. If after all that work you're still buying from Gus via CIV, then, Consumerpedia will just be an enemy projects that distracts and distorts the reality. It does seem to be becoming exactly that sort of a project now!
near-convergence
Consumerium Service access that doesn't need worn devices could all happen through research sources directly if they converged perfectly on definitions of some concepts:
We must at least share comprehensive outcome and economic choice and very exact descriptions of externalities is a real mistake. It's going to take work to get everyone agreeing on all that! But without it, you have no consensus about what is actually worth documenting about a company, no standard terms to use, "what links here" becomes useless, and etc.
managing
These pilot wikis are small and don't have a lot of wiki management problems yet. The above are the minimum required to achieve at least near-convergence. Other structural stuff like best cases, worst cases, user stories, etc., trolls leave to you to manage. It's probably best to avoid that structure for now - it only starts to matter once there is a very rigorous mission, that fails easily, or, far too many people involved.
serious
A serious Research Wiki pilot probably must be run by people who share our values, such as Greenpeace, Adbusters, Earth First or some other principles-driven, not technology-and-status-driven, real-world group.
Ideally people who have never heard of Wikipedia would be running the thing.