Email or username:

Password:

Forgot your password?
Top-level
22

@Kye fam, the app uses totally public data your instance provides anyone. It doesn’t ask for a token, it doesn’t authenticate with your instance, it won’t work if you/your instance/your follows/their instances hide social graph. It has nothing to do with Facebook or data harvesting.

I strongly value privacy and avoiding targeted advertising, but I fear that if you don’t accurately discern what’s a real threat vs harmless, your risk assessment is going to weaken.
@carnage4life

7 comments
bluestarultor

@22 @Kye @carnage4life So does every aggregation site. Sorry, not sorry, but enough public data in one spot all sorted out for ease of consumption becomes creepy. I'm with Kye on this one; the whole point of Fedi is organic growth, not algorithmic curation.

You've been around long enough to have heard the arguments against this stuff before. Fedi has long resisted scrapers and social graphing because of the abuse potential. The issue is privacy, not security.

22

@bluestarultor so my comment was mainly about the somewhat ad hominem claim that one shouldn’t use a tool made by a Meta employee and that therefore it had to be data harvesting.

You’re raising the broader question of whether Followgraph is in general a bad thing. Hmm, so if I’m reading you correctly, there are a few distinct pieces here:

1. “Public aggregation of public data is creepy.” I’m sympathetic to this. I wrestled with this when I added the feature to my Mastodon client to let you jump to someone’s oldest toots and scroll “newer” from there, there was a real worry that making it easy to do this would widen the abuse surface area. But I feel that that’s why Fedi apps including Mastodon give you tools like (1) follow requests, (2) auto-delete toots, (3) hide follower graph, (4) opt out of search engines, etc., to control how much is public, I made the decision (perhaps wrong) that tooling to more easily slice and dice public data was acceptable, and why I’m fine with Followgraph.
2. “We’ve always done it this way, part 1: scraping is bad.” I’m not going to go here since it has nothing to do with Followgraph and because the issue with Archive.org being asked to crawl that instance shutting down was… horrific in many compounding ways.
3. “We’ve always done it this way, part 2: organic discovery is good, anything else is bad.” I have little sympathy for this. Which is weird because, I personally like organic discovery, and likely the reason I’ve stayed on Octodon is because others there like organic discovery too, but I don’t think I ever thought that was normative. I don’t think it was some spiritual thing holding back the creation of Followgraph until now (“back in my day…!”), it’s just with a lot more users there’s a lot more divergent workflows and use cases, and talent to make new tooling.

But that’s just my personal view. I don’t think you or Kye are silly for having your views and avoiding Followgraph—diversity is strength. I personally found Followgraph’s analysis of my graph predictable, like, “yup I don’t follow that account for a reason”. But I’m not a normal social media user either (cf., above-mentioned Mastodon client/my pinned toot).
@Kye @carnage4life

@bluestarultor so my comment was mainly about the somewhat ad hominem claim that one shouldn’t use a tool made by a Meta employee and that therefore it had to be data harvesting.

You’re raising the broader question of whether Followgraph is in general a bad thing. Hmm, so if I’m reading you correctly, there are a few distinct pieces here:

Ressu

@22 @Kye @carnage4life @bluestarultor Am I understading your argument correctly? You are OK that the information is out there, but you have a problem if someone looks?

It's completely OK to resist the creation of these tools from the point of view of organic growth instead of recommendations made by a tool. But making this a privacy issue is a bit misleading. Mastodon has controls to prevent sharing of your social graph, I think what you are trying to argue for is that the default should be not to share your graph. If that indeed is the case, I recommend you file an issue with mastodon and collect your arguments there.

@22 @Kye @carnage4life @bluestarultor Am I understading your argument correctly? You are OK that the information is out there, but you have a problem if someone looks?

It's completely OK to resist the creation of these tools from the point of view of organic growth instead of recommendations made by a tool. But making this a privacy issue is a bit misleading. Mastodon has controls to prevent sharing of your social graph, I think what you are trying to argue for is that the default should be not to...

ericjmorey

@bluestarultor

How has Fediverse tools like Mastodon resisted data aggregation? It seems like it was designed to leverage aggregation as each federated instance is designed to aggregate data of it's accounts.

@22 @Kye @carnage4life

bluestarultor

@ericjmorey To un-tag a few people, scrapers are generally frowned upon because they're tools of _3rd party_ aggregation. It's like how you know the people in your own neighborhood, but you wouldn't expect someone from across the country to take an interest in everyone.

There's simple aggregation and then there's analysis and curation, which I think is more the implication of an online aggregator these days and maybe a limitation of my language. Data vs. information.

ericjmorey

@bluestarultor
That doesn't address the question at all. Mastodon doesn't seem to do much at all to prevent data aggregation. It seems to facilitate it.

bluestarultor

@ericjmorey It aggregates like a phone book, but not like a census. You can find someone in a phone book easy. Connecting them to their family and neighbors is possible, but much, much harder. A census does that connecting.

I can follow anyone on Fedi, but tracking that down for dozens to form a social web would make the average person question why I put forth the effort. I don't like that being given at the push of a button.

Go Up