deleting from kappa is actually quite easy. you just choose not to replicate a certain piece of data and only keep the hash (some lawyer would need to answer if a hash violates GDRP)
Event Sourcing / Kappa Architecture with GDRP is going to be a fun challenge. How do you delete something from a sequence of events that's intended to be immutable? Simply tombstoning entries apparently isn't enough; the law will actually require wiping data, hah.
-
-
-
Oh hey, yeah that actually makes perfect sense! Glad hyper{core,drive,db} support sparse replication
End of conversation
New conversation -
-
-
First blog post in my young lawyer life:https://writing.kemitchell.com/2014/10/23/deep-data-privacy-risk.html …
-
Thanks for sharing; enjoyed reading it!
End of conversation
New conversation -
-
-
And I'm curious about the blockchain. GDPR applies to "organizations" only, and blockchains are owned by no one in specific. What about organizations that need to store the blockchain? Are they forbidden to?
-
Yep, from the articles about GDPR I've read so far you can basically be fined into oblivion for negligance if you store your customer's data in a blockchain. It doesn't mean you can't store _anything_ in a blockchain; but gotta be careful with what you store.
- 2 more replies
New conversation -
-
-
One could choose to mutate the entries anyway. A more pure solution: don't store personal data in the log (or blockchain/ledger), store references to (personal) data defined elsewhere if you must. References can be kept, but underlying data can be changed, anonymized or deleted.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
part of the hope is that, for any given input set, the output produces the same. ideally this means one can run a simple filter on the original data set into it's new form, and re-run all the downstream compute on that new filtered copy, & presto chango it all works.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This isn’t so hard. Store the personal identification (IP, name, email, etc) separate from the behavioral labels and other data. The data without ‘PII’ is just a ghost / persona.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.