At 09:10 PM 11/29/2003 +0100, Erik Moeller wrote:
This would place an unacceptable burden on third parties as they would have to carry along the complete history of every page thtey use (since there is no automated way to determine who is a principal author),
Not currently, but I can't see why an automatic algorithm can't be created that does an adequate job. Have it compare diffs and determine what percentage of the current article's text was originally typed in by each author, for example, and then pick the users with the five biggest scores. This would need some tricks to account for special cases, though, especially things like rearrangements of existing text that look like massive additions/deletions but which aren't really. Could take a lot of computer resources, so maybe it could be done only when generating "downloadable" archive bundles.
Does anyone know if there's an accepted legal definition of "principal author" anyway? If not, then would counting the number of characters in the current version of the article that an author contributed be a reasonable way of defining it for Wikipedia's purposes?
a history which on the English Wikipedia is now so large that we can't even store it in a single file anymore (over 2 gigabytes). Not to mention that having such a list in articles is cumbersome and annoying.
The licence only requires five names attached to the work; that's a single line of small text, which can be tucked down at the bottom of the page somewhere discreet. It could be a configurable option, even, if people don't like the clutter. The whole entire history would not be required then; if the page already lists the five main authors when it's displayed then you can just copy the entire page wholesale into your derivative work.