News for hackers. Stuff that matters.
Re:GDPR, China and data sovereignty are ultimately wins for ... by bad_user on Thursday May 31, 2018 @01:34AM (#17194193)
Primarily data that can be used to identify the user and these have a pretty good definition.

If you have a database with users in it, that's basically the row in the Users table, along with everything in the database that has a user_id, which depending on the data you're talking about, might be enough to just set the user_id to `null`, but if unsure and if possible, then cascade delete is the safe choice.

Basically that's nothing that a normal RDBMS doesn't do efficiently already. If you are using a RDBMS, you should have had a sane deletion policy on your foreign keys already.

It gets more complicated with NoSQL databases of course, but you knew already that by using a NoSQL that's essentially technical debt you're postponing. But having worked on many such systems in the past, it's not a tragedy and certainly not something that couldn't be solved in the last 2 years since the GDPR was adopted.

Honestly, GDPR was adopted 2 years ago and people now act all surprised that it came into effect.

Re:GDPR, China and data sovereignty are ultimately wins for ... by dominotw on Thursday May 31, 2018 @10:01AM (#17196346)
> these have a pretty good definition

thank you for your response. I've been trying to find this in GDPR docs but i can't seem to find a clear definition. Can you link me to the definition if you have it handy by any chance.

I can tell you couple of examples where simple delete from database is not enough.

One example is a analytics pdf report that was sent to management with geo diagram with number of purchases dotsizes/zipcode in a city.