How to recover fic deleted from AO3 that’s NOT on the Wayback machine
Sharing this because I just found out about this and it blew my mind.
The short version of it is: The Wayback Machine is not the only backup/archive of AO3 content out there. It’s just the most user-friendly and immediately browsable.
What you’ll need: A browser for .sqlite3 files such as DB Browser for SQLite, an archive manager (e.g. WinRar or 7zip), good internet download speeds, and potentially a LOT of free GBs in storage space.
Not needed but heavily recommended: A download manager such as HTTP Downloader (so you don’t lose the entire download the second your internet stutters).
1. Click here to get to the archive’s files. It’s going to look something like this:
ao3_current.sqlite3 and ao3_old_files.sqlite3 are metadata files. The .zip files contain fic, most of them in simple .txt format. The metadata files tell you which fic is in which zip.
The “current” metadata file is recent backups. The “old” metadata file seems to be fics archived until 2020ish.
2. First, download either ao3_current.sqlite3 or ao3_old_files.sqlite3. Now launch DB Browser for SQlite, then File > Open Database Read-Only > open the sqlite3 file. Now click on the Browse Data tab.
3. It’s going to look like this.
4. The “Filter in any column” field can be used for keyword searches in, well, any column of this table. Be warned, it takes a while to update, give it time, it’s indexing.
5. Here I searched for all fic which gets a hit for the “Avengers” keyword (usually fandom). You can also search for a specific title, author, description, etc.
Let’s try to locate the first fic on the list. Click on the field on the left - row 1, column 1.
On the right you’ll see the full content of that cell. The most important thing here is the start - ao3_01. This means that the fic is located in ao3_old_files_part01.zip.
6. Download ao3_old_files_part01.zip and open it with your archive manager. It’s 5.5 GB. This will take some time.
7. There are multiple ways to find the fic within the zip file. Probably the easiest way is to use your archive manager’s search/find function to locate the fic by keyword - author is a good bet here, or title if it’s unique enough - and extract that. This way you don’t have to extract the entire archive. Be sure to add a wildcard operator (*) on either side of the keyword.
8. Extract the file and you’re done. Note: It will probably be in .txt format, and might be in one giant block of text. Just select-all and paste it to a proper word processor to restore the paragraph formatting.
+ I suppose if you’ve got like a free TB of space you could just skip the metadata step and download all the zip files and unzip them and use a command line search tool for keywords, too. This will work with keywords like title, author and fandom that are part of the file title. The metadata file just contains additional info, like character fields, description, etc.
This isn’t a perfect remedy, there are still fics that got deleted before they could get archived here. But it seems more complete than the stuff on the Wayback Machine on average.
