[This is every section of the Searching Archives Treatise regarding Chansociety that was removed due to its uselessness.] Firstly, Chanarchive was the definitive selective archive for 4chan between 2006 and 2013, covering every board and saving around 18000 threads at its peak. Its death and complete loss of data in late 2013 was one of the greatest losses in 4chan archivalism. However, the majority of threads from the site were saved on the Wayback machine, and although most thumbnails are lost, let alone full images, the KB, filesize, and time of an image post should be enough to determine if an image is the one we're looking for. For Chanarchive, the Wayback machine will have to be used to find threads. In terms of searching through boards that had fuuka archives on July 15 2012, I would consider it unnecessary and unlikely that a thread from chanarchive, which only saved important, big threads, would be missing from a fuuka archiver. However, there may be gaps in one of these archives, and if one were to exist around the day July 15 2012, using the URL prefix feature in approach 1. below would be warranted for that board, just to check. There are several approaches to scanning Chanarchive for threads: 1. Most intensive. This would entail using the URL prefix feature on the Wayback machine to find every board (done by taking this link here and changing the letter to a different board) and searching through every single thread to see if it had a post on July 15 2012. Of course, sorting the threads by the date of their first save in the Wayback machine greatly reduces the number of threads needed to search through, following the logic that if the thread was saved before July 14 2012 no posts afterward would be in it. However, checking all threads afterward would still take a significant amount of time. The only other possible shortcut that may be taken would be to exclude searching through boards that had a fuuka archiver on July 15 2012, although this still leaves the herculean task of searching all of /b/. A working theory is that the number in the URL is indicative of its placement in the archive and thus can be used to determine when a thread was posted, but this is still uncertain and I don't recommend using them. This is a particularly intensive and lengthy means of scanning the site, and would likely be futile considering the 85% chance that the thread was not archived on chanarchive. 2. Least Intensive: The URL prefix feature can be used to sort by date (as was done in approach 1.) and find a thread saved on a particular day (say, July 15 2012, or later for slower boards that have threads lasting over several days) and find the "latest threads of thesame board" box on the bottom right, where all recent archived threads from that board are shown. Example: /b/, /x/. An additional recommendation that should be done for either approach is to view the frontpage of each board (which can be found for any board via editing this link for whatever board: https://web.archive.org/web/20120715000000*/http://chanarchive.org/4chan/[board letter(s)] - and then click for the date closest after July 15 2012) in order to check for any threads that may have been removed before being granted full archival via the voting system, and thus not visible through the latest threads box. This is most useful for approach 2., but would preferably be done for both considering how time unintensive it is. For a more reasonable means of completion for functioning members of society, the separate parts of the search may be divided amongst many members of a large group as opposed to one person. For example, for Chanarchive, one person may be designated to search through the URL prefix for one specific board, such as /x/, scanning posts and reporting if they found anything. For a group search, I would recommend the most intensive approach (without the fuuka archived boards and sorting by date to get to July 15 2012) for Chanarchive and the advanced lazy approach (2b) for Yotsuba Society to maximize what can reasonably be found without overexerting everyone.