>Please read the FAQ before posting.
Q: Do you have any statistical data for this archive?
A: Yes: http://thebarchive.com/stats/
Q: Do you have the threads from the other archives?
A: Yes, I have the entire database dump of 4ch.be
and the thumbs they provided, as well as about a month of archived posts and thumbs from archived.moe. I would gladly add any foolfuuka&asagi db&thumb dump if someone provides it to me.
Q: When did you began archiving?
A: On 05. June 2016
There are no threads older than 2015-10-24 on this archive.fgts.jp
had that data for sure but all of the collected data of /b/ is gone forever prior 24.10.2015.
Prior ~march 2016 there exists only thumbnails and a month might be missing for this archive. See http://thebarchive.com/b/statistics/activity/4archive.org
may contain what you are looking for, googling 2012-06-14 site:4archive.org
reveals some results from that date.
If it was on /b/, this creepshot thread exists on google from that date: http://4archive.org/board/b/thread/603783238
As for the search, the system follows in general a particular way of storing images->
4chan stores the image in a unix timestamp or smth like that,renaming the file for instance 1447554314617.png . The 4chan posts fetcher, called asagi, pulls them in that format and slices the name in to quad and double parts for the data directory, thus 1447554314617.png transforms in to 1447/55/1447554314617.png . This is done for easier file management.
The fetcher will first check if the image (based on MD5) is already in the DB and if it was it will use the DB entry instead of reloading the image again, which saves 20-40% more space.