Wiki Theory

WikiTheory involves thoughts about wikis and their common characteristics, especially featurewise.

WikiAxioms

A Wiki can automatically interpret a special language of gestures or codes

Different wikis interpret different "languages" of gestures and have a different range of actions they can take on the user's behalf. An example of a gesture would be to put double brackets around a phrase in this wiki. The wiki would interpret that as meaning that the phrase should be turned into a link.

A Wiki generates a browsable hypertext web by interpreting its contributor's gestures.

One primary gesture/action is for the user to create a document in an editor and have the wiki store that document on the server in some form. Within that document the user can enter certain codes and the wiki will respond by creating links to other stored documents. Different wikis have different vocabularies of codes that they can understand. Most wikis thus have possibilities to layout text, most can even insert images, and a few support customizable plugins.

A Wiki is an un-ordered and un-typed collection of pages

For the same reason that modern scripting languages are un-typed, the wiki does not require that the information submitted to it already be structured or organized into certain fixed patters. To be powerful and expressive the wiki system cannot commit itself in advance to certain kinds of uses or certain structures of content. There is no requirement that pages be related to each other or that they fit into any organizational scheme (like a table of contents). This is why tentative links (broken-links) are acceptable in a wiki. The wiki has no expectation of unity or completeness.

The content of individual pages is also un-ordered and un-typed

This no-structure-up-front logic applies to individual pages within a wiki just as it does to the wiki as a whole. There is no expectation of how a page will be structured or that it will be limited to a given topic or set of topics. There is no expectation that a page will be about something or that its content will in some way correspond to its name. The whole concept of "page" in a wiki is a guess at the level of granuality it will be useful to clump content in rather than a meaningful division of the content in the wiki into significant units.

Tentative links in a wiki are not "broken" and pages with no links to them are not "orphaned".

On the wiki the complete structure of "link/page" is not required before creating one or the other. The wiki's low requirement of structure means that a contributor can express the intention of creating a page, or an invitation to others to create a page, by entering an link which points to a page that does not exist. They can also create a page that has no links to it on any other page in the wiki without that page being an inaccessible (because the wiki auto-generates a page directory). Therefore, what would be a "broken link" or an "orphan page" on a different kind of site is no error, but an important mechanism that the wiki uses to generate itself.

A Wiki operates in terms of multiples

In a wiki it is always possible to multiply pages, to multiply contributors and to multiply topics. This same dynamic happens at the level of the individual page. The page can split and become multiple pages. The page can hold multiple texts that are "about" multiple topics. The texts on a page have multiple authors. The multiplicity applies at the level above the wiki as well. Either within the same site or by using a different site, it is always possible to multiply wikis. This multiplication is always inherent in the wiki, even if it has not actually happened to a particular piece of text, a particular page or a particular wiki. This axiom can be apply in reverse. A given page in a wiki has the potential of multiple contributors, even if only one person has written the text on it. This means that wiki text has not real "author", although signatures can be included in wiki text, they are a non-wiki element.

A Wiki is fast to operate.

The wiki's ability to interpret and execute commands and its ability to let its contributors add content and give it structure at the same time rather than having to structure the content before being able to add it make it fast. The wiki's technical simplicity (because it write the HTML of the pages for you) allow the contributors who have the content to contribute it themselves rather than passing it along to specialized people in the organization who have to add it to the system. This direct involvement can also dramatically reduce the time it takes to get results.

No existing information is inaccessible: A Wiki automatically builds several kinds of systems that allow navigation among its pages

Traditionally a wiki has a mechanism that builds hyperlinks, one that tracks reverse-links (pages with links to the current page), an alphabetic list of pages, a list of recently changed pages and some sort of search facility.A wiki has measures of accessing any page, even if it be the orphan pages. Therefore a wiki can provide intelligent tools to detect information that wastes storage space, if so desired. Most wikis, in the opposite, will preserve all versions of a page, despite Space consumptions. This behaviour is part of the Fire Brigade principle and might be considered an axiom for its own. However, a wiki is still a wiki without historical tools.

A Wiki is a "flat" information space, rather than the hierarchical spaces represented by file systems or databases.

A wiki is a "bag" (an unorderd collection) of pages. A wiki-farm is a "bag" of wikis. All the things put into a wiki are in the same space. The wiki resembles a "pile" of papers on a desk, rather the same papers filed into the folders of an ordered filing cabinent. The information universe a Wiki represents does not depend on the filesystem or database used, and can be abstracted to an interface that takes a page name or resource name (image, audio etc.) and returns either a wikipage or the requested data. The resource name is not a URLs, it's a Wiki Resource Name, WRN, containing namespaces. An absolute WRN must have the wikiname as the root namespace classifier.

A Wiki always includes a mixture of non-wiki elements

A wiki's low level of structure means that other elements can be included into the wiki by mixing them in rather than by building them as integrated parts of the wiki-like elements themselves. Given the radical nature of the wiki-like elements almost all wikis include other elements that make the wiki more palatable to the taste of the group using it. Examples of non-wiki elements are: saving past versions of pages, identifying contributors through logins or passwords, access control to the viewing and editing of pages, systems for tagging or indexing pages, systems for organizing pages into meaningful sets and systems for viewing the changes between different versions of pages. The addition of some or all of these non-wiki elements is very important to the functioning of any wiki. The logic, design and implementation of such non-wiki elements crosscuts the logic, design and implementation of the wiki-like elements. Which non-wiki elements are added to a wiki and the success of how the interface between them and the wiki-like elements is handled do much to establish the character of an individual wiki implementation.

WikiAnatomy

Namespaces

Every wiki has at least one namespace. Some fellow wikis cross reference each other using URLs and a colon separated notation for the linktext, others openly use namespaces to separate content.

The Spinal Tree

Every wiki is used as an information retrieval system, and the Spinal Tree, rooted at the MainPage, are the hypertext routes traveled the most. If a wiki has categories, this is, in fact, a strong part of the Spinal Tree. Both Spinal Tree and Categories need no special implementation. They are there, just because of the nature of the wiki, and sometimes the tree is institutionalized, for example as a page-tree, which is, essentially, a rather rigid structure for wiki.

Fire Brigade

Most wikis are exposed to a big number of people. Some behave badly towards the wiki, others provide dangerous information unintendedly. The Fire Brigade is, if existent, the sum of all methods to avoid and repair damage to the wiki. In Wikipedia this is mainly a large stem of users (Fire Fighters) watching the recently changed pages. This works extraordinarily well. For bots this would not work, but there are other countermeasures in the toolchain of the fire brigade.

This theory might form the basis of the Personal Information Wiki, simply PIWi.