Skip to content

Wikitext standard

From the Quicksilver Metaweb.

This page is open for anyone to edit and improve.

A wikitext standard would let different wiki software suites - including offline readers or full offline clients and text corpus software - read the open content of this and other webs.

There is an effort towards this starting at Meta-Wikipedia. This may eventually relieve the dependence on mediawiki and let other suites such as tikiwiki or MoinMoinWiki be used, or entirely new programs to develop to manipulate the Metaweb.

See 'linguistic democracy', wikitext DTD, Wikipedia DTD, 'simple ideology of Wikitax', (wiki) talk media, and other comments and proposals linked from there, if you wish to contribute to the Wikipedia-centric discussion.

All-language support is being discussed at Unilang - see EU.Unilang.org's wiki. This may be the most fruitful place to discuss Unicode and the representation of pronunciation and such.

Meatball:WikiMarkupStandard lists the common sorts of formatting people want to use, the diverse ways of representing that formatting at various wiki, and discussion about which way is "best".

There is an even more strategic discussion on mediawiki features going on at Consumerium, where Vibber and Juho have discussed the need for authentication information to be part of the standard. Since Consumerium, a moral purchasing project, has more significant needs for authentication than any other mediawiki-based project, it may lead in this regard. Authorship tracking and authentication of the veracity of comments, and group affiliations between authors (in a faction will probably have to be state of the art at Consumerium, given the latter's mission). Metaweb authentication can simply track that - in the meantime, assisting Juho and Brion Vibber in using Consumerium as guinea pig.

Also the Disinfopedia, another mediawiki-based service, is devoted to exposes of propaganda techniques and propagandists, will probably set standards for compiling "profiles" of people. It seems likely that at some point that project will come under legal attack, if only for telling the truth. There has been discussion of 'profiler' technology to draft biography at Meta-Wikipedia, but it's not a priority. Metaweb would be best off developing standards for source authentication (as also addressed in simple ideology of Wikitax) and linking all mentions of one particular human body (as per meta-wikipedia's person DTD proposal).

Wikipedia and Disinfopedia have started to converge on how they represent dates - Disinfopedia uses a cut-down version of Wikipedia's standard, and doesn't have comprehensive timelines and "year articles", etc.

The first useful map generated automatically from Wikipedia's geographic tags is also evidence that it sets standards in the GIS area that will probably be adopted by other mediawiki projects including Metaweb.

It is also possible that some standards will be set by some other, non-mediawiki project. But at this point, Wikipedia has over 300,000 articles in many languages, so conversion would seem less onerous if it was the base standard, at least for other mediawiki users.

Tracking the projects which have the strictest requirements, and noting discussion at those projects about those requirements (only), seems like the wisest way to track progress towards a wikitext standard and associated capabilities like a wikitext DTD - on which a metaweb DTD would be based. Here's a list of leading projects that Metaweb can follow, and the capabilities it should track: