Some notes on a survey of what all I find in IndexedDB storage in Firefox. That’s something like cookies, a way for a website to store data on your computer, only in this case it’s a fancy NoSQL database. I’m using it myself and have found Firefox’s Storage Inspector unreliable and limited, so I built my own tool to look through my profile directory and tell me what it finds.
Example report from the tool:
https://twitter.com dm_typeahead 20160920 98304 conversations title metadata - users name_lowercase, screen_name https://www.smithsonianmag.com ONE_SIGNAL_SDK_DB 1 49152 Ids - NotificationOpened - Options - https://www.theguardian.com test 1 49152
That tells me
- Twitter has a database called “dm_typeahead” with three tables; conversations, metadata, and users. Users has two columns, name_lowercase and screen_name. It has a version string of “20160920” and was 98304 bytes big when last vacuumed.
- Smithsonian is using some SDK to create some tables about notifications, but they contain no columns at all.
- The Guardian created a database named “test” with no tables at all.
So what’d I find?
- A bunch of empty “test” databases, presumably testing the browser can do IndexedDB at all. This may be for detecting if the browser is in private mode.
- A bunch of sites use One Signal, which I guess manages those horrible HTML5 website notifications that spam pop ups. I’m religious about never allowing that, which is probably why I have no data.
- Several sites using Augur, a web tracking system.
- Things called LPSecureStorage and fibet; trackers?
- archive.org seems to be storing PC emulator state
- Amazon is caching a lot of data about books, maybe for Kindle Cloud?
- Twitter has extensive app-specific usage
- WordPress’ Calypso
- broadwayworld.com, a pretty spammy site, has a database named J7bhwj9e with some user tracking stats.
- ft.com has a bunch of databases named next:ads-v1 and next:image-v1 and the like
- wired.com has something called workbox.
The tool I built is in this gist.