>>36128479Yes, I downloaded the pastes. Most of the links were collected earlier or in previous threads.
>>36129293Well, the initial panic has subsided. Have an update from me: my scraper finished its initial run, picking up 84k pastes and 4.7k users. I'm not done yet, though. There are another 4k users/pastes that I collected from lists posted previously. I ignored these before because they seemed mostly non-pony, but I decided to scrape them anyway. Assuming that nothing goes wrong, I should be actually done by tomorrow.
Also, here are the promised stats. Take these with a big grain of salt: there are tons of non-pony pastes, many "not found"s may be due to typos, and there may be a few duplicates counted (about 10-20).
SMART filter: 43711 pastes, 51.85%
Okay: 32257 pastes, 38.26%
Not found: 7907 pastes, 9.38%
Private: 317 pastes, 0.38%
Potentially harmful: 114 pastes, 0.14%
Still, this shows that the "deleted because potentially harmful" problem is actually really, really small. Again, it's good to back things up anyway, but I'm starting to think that the deletions may be due to manual reports. The timestamps that Pastebin reports for these deletions are confusing though. See pic related. (35 pastes were excluded because they had epoch times: "December 31, 1969 6pm CST". Probably a bug.) I'm pretty sure the "potentially harmful" notice is recent, yet the earliest deletions are from September 2016.
Still, keep in mind that some (or even most) of these deletions may be of non-pony pastes. I'll update once I've sorted through everything.
