When Is a Deletion a Real Deletion
June 29, 2019
Years ago we created the Point (Top 5% of the Internet). You oldsters may remember our badge which was for a short period of Internet time a thing.
When we started work on the service in either 1992 or 1993, one of the people working with the team put the demo in the Paradox database. Hey, who knew that traffic would explode, and advertisers would contact us to put their messages on the site.
The Paradox database was not designed to deal with the demands we put upon it. One of its charming characteristics was that when we deleted something, the space was not reclaimed. Paradox — like many, many other databases — just removed the index pointer. The “space” and hence some charming idiosyncrasies remained.
Flash forward decades. A deletion may not be a deletion. Different “databases” handle deletions in different ways. Plus anyone with experience working with long forgotten systems like the Information Dimensions’ system to the total weirdness of a CICS system knows that paranoid people back up and back up as often as possible. Why? Fool with an AS400 database at the wrong time doing something trivial and poof. Everything is gone. More modern databases? Consider this passage from the Last Pickle:
The process of deletion becomes more interesting when we consider that Cassandra stores its data in immutable files on disk. In such a system, to record the fact that a delete happened, a special value called a “tombstone” needs to be written as an indicator that previous values are to be considered deleted.
When one digs around in database files, it is possible to come across these deleted data. People are amazed when a Windows file can be recovered. Yep, deletions don’t explain exactly what has been “deleted” and the conditions under which the data can be undeleted. Deletion allows one to assume one thing when the data have been safely archived, converted to tokens, or munged into a dossier.
Put these two things together and what do you get? A minimum of two places to look for deleted data. Look in the database files themselves, and look in backups.
In short, deleted data may not be deleted.
How does one know if data are “there”? Easy. Grunt work.
Why is this journey to the world of Paradox relevant?
Navigate to “Google Now Lets Users Auto-Delete Their Location and Web History.” Note this passage:
Specifically, Google account holders will be able to choose a time limit of either 3 or 18 months, after which, their location, web, and app history will automatically be deleted.
Some questions?
- Who verifies that the content has been removed from indexes and data files?
- Who verifies that the data have been expunged from metadata linked to the user?
- What does deletion mean as the word is used by Google?
- From what has something been deleted?
Like hotel temperature controls, fiddling with the knobs may change nothing.
Stephen E Arnold, June 29, 2019
Comments
2 Responses to “When Is a Deletion a Real Deletion”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
Great information. Lucky me I discovered your website by chance (stumbleupon).
I have saved it for later!
Having read this I believed it was extremely enlightening.
I appreciate you taking the time and energy to put this
information together. I once again find myself
personally spending a significant amount of time both reading and commenting.
But so what, it was still worthwhile!