That sounds like a good solution to me.

Just knowing that any such data is collected or shared, and when, could help to improve one's privacy.

I see that, but I didn't vote on other platforms. I knew I'd be giving data directly to the owners and their five eyes operatives. I know this platform is public, so I'm careful with my words. Now I know I should be careful with voting, too.

The people who want our details are incredibly creative at how to interpret data that seems innocuous.

Edit: that link is very informative, thanks. I should confirm that I assume everything on the internet is public in one way or another and confirm that I don't have any major concerns with the general security architecture of the Lemmy software or the way Lemmygrad is run. I just thought it's something that we should talk about to make sure that we're not increasing the chances of being doxxed by giving away useful metadata.

Edit 2: when I say, 'I don't have any major concerns with the general security architecture', people should know that I'm not qualified to judge this from the coding side of things!

I see the irony. I suppose two differences are that, one, the survey was optional, and two, although participants might not necessarily know who was reading the answers, the answers weren't (I don't think) generally accessible.

Voting data is not generally accessible, either, but it appears to be accessible to admins of other instances. If this is the case, it wouldn't be difficult to set up an instance, make a post through any account about a specific topic, and observing the data to see who up/downvotes it. This could narrow down the list of people that a bad actor might want to target for further data harvesting.

Considering how many billions are put into surveillance, data collection, and controlling (social) media, I don't think we should discount this fear as farfetched. Even if, on its own, it's fairly benign data.

I've also noticed that some of the new instances/users are a lot happier with 'local' communities. If this grows into, say, a list of 'things to do in my area' communities, it wouldn't take much for a fascist to identify local targets by a series of what I'm going to call 'voting traps'.

Maybe I'm paranoid.

Exactly why I didn't; I still don't know how much data flows through lemmy-actual's veins to be comfortable like that.

It logs the timing as well, which could be sensitive data. For example if an employer were to gain access to this information and tie it to an account that someone thinks is anonymous, they'll know when you weren't working but getting paid to be at work. Or it could be used to determine when some is at home or out. Or be used as evidence for holding certain views.

It's unlikely for a single employer to get that data. But I wouldn't put it past the five eyes to set up an instance, mine data, and use analysts/AI to cross-reference it with other user metadata.

It's like J Sakai says in his security pamphlet. It's bad practice to give any information to feds, even 'benign' data because it helps them to build profiles on you and others. To offer a rather extreme example, if they know you were upvoting a comment at 16.07 EST, they can guess it wasn't you at a protest at the same time. This means they can narrow down the suspect list to the other handful of people with your build, etc, who go to protests. Being lax with data means the feds have an easier time undermining the efforts of people who are tight with their data.

It's a privacy issue that I hadn't considered. I knew our admins could see this data. I didn't realise it was visible to other instances.