User:Fmph/techsol

An outline for a technical solution to naming difficulties
There have been a number of naming disputes (Ireland/Republic of Ireland is the one that got my attention) all of whom are based around a technical limitation of the current implementation of MediaWiki, namely that we can only have 1 article ateach article name location. In other words, we can only have 1 Ireland article. This is such a fundamental principle of Wikipedia, that suggestions that we might change that principle are often treated with derision. "Why would you want 2 Ireland articles?" is the style of rhetorical question usually asked. I prefer a different question. Why would you want to be limited to only one?

Traditional paper publishers have no such limitation. But Wikipedia is not paper, so that's that then. Or is it? WP:PAPER is about content, not titling. The principle of unique article names is there for a very good reason. It's there to make editing and linking very easy for novices. So if I want to know what a novice is about, I can just add double square brackets before and after the word novice and hey presto a link magically appears. Trivially simple. Totally taken for granted. A master stroke by the original software designers, because it allows a huge level of user interface/user experience enrichment for apparently minimal effort.

I say 'apparently' because it is important to consider to what happens after the editor adds the brackets. Lets keep it simple by ignoring piped links and templates. A simple 'novice' link like the above is an ideal example. When the edited article is submitted with nthe newly linked word novice, the system needs to parse the markup to know how to display (or render) the page when asked. It needs to know if the word 'novice' should be linked to the existing novice article or if the article doesn't exist, the link needs to create a new one. In the Wikipedia environment this would be rendered as a redlink, like this link to the non-existent article novicf.

The important point to take out of this discussion is that every link is parsed and checked against the unique article index. If it finds a match one action is performed. If it fails, another takes place. This point in the process is key to my proposal. It gives us a point in the process where we can intervene, add a small amount of additional processing for non-unique articles. I'll return to this point later.

Natural Naming, and why we need it
Let's give it a name. I'm going to call it Natural Naming. But what is it? It's a proposal to develop a process to allow wikipedia articles to have non-unique articles titles.