Magnus Manske wrote:
Alwin Meschede wrote:
You know, deleting more than 200 articles by hand can be a painful job... Maybe some developer could invent a SQL query that gets rid of all those duplicates at once? And maybe the software can be somehow improved to make sure it updates articles correctly and does *not* create a duplicate in some weird corner-cases.
There's a "INSERT ... ON DUPLICATE KEY UPDATE" function in MySQL which should avoid double creation.
Ah, but that would over-write the older 'cur' row, no?
Alternatively, we could make the key pair cur_title,cur_namespace unique.
That sounds much better. :-)