4 ms·
> The concern is that articles that no one cares about are potentially places where damaging content could fester and dealing with them That could describe the
by xref 7y ago
> The concern is that articles that no one cares about are potentially places where damaging content could fester and dealing with them
That could describe the bulk of the millions of articles, why single out stubs for the axe specifically
- nullc 7y agoI'm sure someone could do a whole graduate thesis on that question, or any of a million other questions about how Wikipedia works. Wikipedia is evolved, not intelligently designed. It has a million behaviors that few can explain clearly-- some are historical baggage, others have critical subtle interactions. There are historical reasons for sure-- e.g. relatively early in WP's history there were a number of editors that mass created all sorts of quasi-random articles using automation. But are there good reasons to keep doing it? Dunno. Maybe someone more active than I has a better idea. If the meta of wikipedia is interesting to you, you might find these notes I made for a talk I wrote in circa 2008 or so which addresses why WP's behavior is often kind of inexplicable to humans and why that isn't a problem or at least why it's not avoidable: https://people.xiph.org/~greg/zen.wp.txt https://people.xiph.org/~greg/zen.wp.txt