Re:Debunking the Myth of "Anonymous" Data
by __MatrixMan__ on Sunday November 12, 2023 @11:34PM
(#38246962)
So this conversation has me thinking that I could go a step further and create a master list of all of the hashes, sort it, and then replace the hashes with their string containing their numerical index in the list. It just needs to be a 1-to-1 map. So maybe:
As for the utility... it doesn't break foreign key references. Null values still stay null... You can run the app, it'll just present you with gobbledygook.
So suppose the complaint is "two weeks ago workflow xyz started failing to complete, but never explicitly reached the failed state so we weren't alerted to the problem".
Before I could hash "xyz" and get the sanitized workflow name and then load the app with the sanitized name to see what stands out about it compared to its neighbors. Perhaps the dependencies on it's inputs are null or there's some should-never-happen scenario like a cyclic dependency. I don't need to know the actual names of the depended upon things to identify a cycle in them.
After this upgrade (?) I would not be able to hash the name that the user has given me, but I could still go looking for suspicious behavior during the time internal they indicated.
Privacy wise, the game is to be as effective as possible without having to ask them to share their screen. And when it's a bug, to have something to test with so we can say "it's fixed now" instead of "is it fix now? How about now?"
> Baltimore=>0x01a1...1ef7=>001 > Aaron=>0x17bf...86f1=>0002 >{"Foo":”Bar"}=>0x19f4...9af2"=>0003
This would be nearly impossible to reverse.
As for the utility... it doesn't break foreign key references. Null values still stay null... You can run the app, it'll just present you with gobbledygook.
So suppose the complaint is "two weeks ago workflow xyz started failing to complete, but never explicitly reached the failed state so we weren't alerted to the problem".
Before I could hash "xyz" and get the sanitized workflow name and then load the app with the sanitized name to see what stands out about it compared to its neighbors. Perhaps the dependencies on it's inputs are null or there's some should-never-happen scenario like a cyclic dependency. I don't need to know the actual names of the depended upon things to identify a cycle in them.
After this upgrade (?) I would not be able to hash the name that the user has given me, but I could still go looking for suspicious behavior during the time internal they indicated.
Privacy wise, the game is to be as effective as possible without having to ask them to share their screen. And when it's a bug, to have something to test with so we can say "it's fixed now" instead of "is it fix now? How about now?"