BannedContentBot

MeatballWiki | RecentChanges | Random Page | Indices | Categories

Would people like me to run my spamclean.py script on MB? The spamclean.py script checks recent changes for new text which matches any of the spam regexes from CommunityWiki:BannedContent. If it finds spam, it looks for an older version of the page which is spam-free. If it finds one, it reverts, writing "spamclean.py spam revert" in the summary field.

The script is able to be run interactively, with a human checking to make sure each revert is really spam, but because I don't have time to do that, I run it as a cron-job. So, one disadvantage is that if normal conversation is auto-detected as spam, it will be auto-reverted. However, if this happens, just drop me a line and I'll fine-tune the regexes on CommunityWiki:BannedContent to avoid catching that conversation (or, you can just mangle the spamy word, i.e. s-p-a-m to avoid it being caught by the regex).

The bot misses any spam which hasn't yet been added to CommunityWiki:BannedContent, of course. But it's a simple matter to suggest that a new keyword be added to that list.

So, shall I add MeatballWiki to the list of wikis that are being auto-reverted?

-- BayleShanks

One concern is with the fact that we sometimes lose versions of a page. For a brief window the lost version is retrievable from the diff, albeit crudely. Once an automated revert is done to the prior available version (note this is not the lost revision!), we lose all possibility of recovering the lost revision. I think it's best to wait until we move to the new server, where I expect the version loss will be more contained.

: That's a good point; I agree. -- BayleShanks

BannedContentBot

Discussion