On 7/3/06, Bryan Derksen bryan.derksen@shaw.ca wrote:
I'm the opposite, I think the second principle is very important. Anything that's not in the article namespace, template namespace or image namespace is stuff that's not really a part of the encyclopedia itself - it's just temporary scaffolding and support equipment we're using while writing it. When Wikipedia gets mirrored or otherwise distributed all that non-encyclopedia stuff should be pared away. If there are links to it then those links will automatically become broken.
Ok, so we have: Article namespace - encyclopaedic articles about encyclopaedic topics. "List of X" is about as far as we stretch. Template namespace - text that gets included into encyclopaedic articles. Plus, random cruft like userboxes. Image namespace - images that get included into encyclaedic articles (plus other cruft) Category namespace (you forgot this one) - meta information about groups of related articles
Now, where the hell does meta information about individual or groups of effectively unrelated articles go? Bizarrely, the category namespace seems like the closest one! What else is there?
Far better to switch over to the <ref></ref><references/> system, IMO. The main problem here is that there are tens of thousands of articles to convert. I don't suppose there are any bothandlers who'd be willing to look into doing automated conversion? Rambot articles are probably still fairly standardized in format. I've noticed that the old Rambot reference links use section numbers that no longer match anything in the reference "article", that could perhaps be repaired in the process.
I don't see how this helps. Putting <ref>s into an infobox is dangerous, as there's no guarantee the including page will actually have a <references> section...
Steve