Hacker News Reader

Comment by bonestamp2 | Hacker News Reader

I would actually consider using Facebook if I could pay for it and not be tracked or datamined. There is a let a legit use case for sharing photos, personal stories, and discussing news and information with a curated list of your friends and family.

View on HN

"Social" should be an Internet protocol. The only reason it's not is that we basically stopped making protocols (well, ones that gain any meaningful traction, anyway—I'm aware there are some lightly-used efforts at social protocols) because all the companies in a position to push them to a meaningful number of users are better served by making interoperability difficult. The "free" services spyvertising economy, where captive non-paying user count & eyeball time is what matters, is why things are this way.

At it’s simplest, the protocol for social media is RSS with feeds for each of your contacts.

And I guess this is sort of how fediverse is working.

Protocols don’t make much money compared to walled gardens selling max data possible.

I see this argument made often on HN, but it's not clear to me how an internet protocol would make social networks more accountable towards their users. Do you mind explaining your reasoning here? Specifically, how would a protocol prevent motivated companies from tracking your personal information?

> Specifically, how would a protocol prevent motivated companies from tracking your personal information?

They could still try! But you'd have options.

Take email, for example. I cannot imagine something like that coming into existence today.

I can use my own client to avoid ads and tracking from my service provider—did I download this message? Sure, the server knows that. How long have I looked at the message? Which message did I look at next? Did I follow any links (yes, someone might track that part, but my email provider's going to have a hard time doing that)? What mouse movements did I make while looking at it? No such luck there, and yes websites and closed-platform services do track that stuff.

I can switch providers. Say my email provider starts injecting trackers into all links. I can just dump their ass if I don't like it. I keep using email, and now they receive zero info about me (I mean, they might get a little if I send emails to their users, but you get my point). If I have my own domain name I don't even need to tell anyone I switched.

I can email someone using a different provider. Yes blocklists or whatever might cause a problem but, fundamentally, this does work.

Protocols force providers to act like a telco, at least, except that the situation's even better for software because the barriers to entry in the market are so low... unless all your competitors are giving away access to their strictly closed ecosystem for free, and not supporting open protocols. Then you're screwed, and that's exactly what's happening now and why the Internet protocols are largely frozen in time.

I see, thanks for the detailed reply. Yeah it's sad that companies have no incentive nowadays to support open protocols (besides the ones that already exist). I wonder if regulation could solve this problem or at least encourage healthy competition into the marketplace.

at the very least it'd mean you could take your data and connections out of one service and go to another which would mean there is genuine competition that isn't hampered by network effects of platforms.

When you say a protocol, do you mean in the same way http is a protocol? Like a new way to communicate that web browsers would start adopting?

Because in my opinion the browsers would need to make it as easy to use as any other website or phone app if it were ever to gain traction.

Usenet, email, http, XMPP, IRC, et c. Yes, just like those. In combination, that bunch is already not too far off. The trouble is that anything trying to do that is competing with "free" spyvertising services, which have no incentive to integrate with them (i.e. implement the protocol), unless it's to do it temporarily to eat their meager market share before cutting them off. IMO that dynamic is why protocols have stagnated for decades. Working on a client or server for some new protocol is thankless when you know it'll be niche at best, and more likely DOA. Making a go at a business with one is insane in this market. So they stagnate. No new ones catch hold, and old ones make slow progress at best, or gradually die.

Have you looked into ActivityPub?

ActivityPub is more than lightly used, it just isn't used by people you want to associate with. You pretty much have to either accept radical leftism and hang out with tankies or you get to hang out with the actual Nazis that are blocked by the "mainstream" leftist servers

Yes yes, there's radicalization for everyone! God forbid you want to use social media for something besides political fighting though

There's such political fighting everywhere on the 'net. I use ActivityPub for musicians, programmers and Cory Doctorow, and don't have to deal with much of that, because I just avoid it.

The Federated Timeline is slightly cursed (I haven't curated it for a while), but less so than Twitter. (My instance, Fosstodon, doesn't really federate with the Nazi instances; those it has ended up federating with have been swiftly blocked, because Fosstodon's admins don't like admin-condoned harassment of their users.)

> My instance, Fosstodon, doesn't really federate with the Nazi instances; those it has ended up federating with have been swiftly blocked, because Fosstodon's admins don't like admin-condoned harassment of their users.

And I'm glad that with email we don't have server operators deciding on politics for their users. ActivityPub has a long way to go if labeling instances as "Nazi" is enough break federation.

I see this often, but the trust problem still exists. There's no universe where I can trust Facebook (or Google or...) To not track me. Others have their own standards, but I don't feel like it's unreasonable to consider them APTs no different than the spyware of old. They have datacenters dedicated to exactly that. How can my data be reliably segregated so that it's not 'accidentally' included later?

Even if the data stays all first-party, they make so many copies and it gets put in so many datasets across the world that co-mingling and abuse of it just feels inevitable at that scale.

Uploading your daily business, personal photos, and down to what you eat to a profile on someone else's servers was never a good idea from the very start. I find it pretty amusing how it becomes outrage and parroting after the fact that bad news comes out about Facebook. We knew what Facebook was doing in 2008-2012 era as well but nobody parroted these old and tired narratives then.

We knew what Facebook was doing in 2008-2012 era as well but nobody parroted these old and tired narratives then.

We didn't know that Facebook was compiling dossiers on every human being on the planet, even the ones without Facebook accounts.

There were plenty of greybeards making claims about FB tracking etc, they just got written off as crackpots. The below link describes the history of some of FBs tracking and includes publicised examples from 2010 and 2011. To me it’s just that we believed FB back then when they claimed ‘it’s a bug’, ‘a limited number of users were affected’ or ‘we don’t don’t do that’.

https://www.propublica.org/article/its-complicated-facebooks...

I created a fake account for testing a fix to a bug for my company. I used a fake email, fake name, fake location, etc. I used a photo of something I was given for graduation and a date of birth that was within 6 months of my real one.

They started recommending my friends from my high school. I went to a magnet school for my junior and senior years, and Facebook even recognized that. I graduated 10 years before I signed up. And the one piece I made sure I faked was the school I went to. I'm shy. If I wanted to communicate with someone I figured I'd add them myself.

Nothing magical to it. They had some other piece of data from someone else (or me from some other source) that matched up (the data was fake, but I had used it previously; I don't make that mistake anymore). But it showed me just how dedicated to getting all of this data they were.

Wow, thanks.

Everyone paying attention when Path et al were uploading whole phone address books (which lead to Apple adding it as a specific permission in the next iOS) knew this data was getting stored indefinitely.

I've always taken Zuck's infamous "dumb fucks" comment to come from a place of astonishment, more than malice. Posting tons of private info under your own name online was strongly against Web norms at the time, but all those newbies didn't know that. FB and others made it normal, but the reasons it was a bad idea to begin with didn't go away.

>against Web norms at the time

I would even say it was against social norms. Don't talk to strangers. Don't give away your phone number. Don't give away your address. Don't tell state your mother's maiden name.

It was such a simple thing to get all of those new people on social media to give up all of that very personal information without hesitation. Yet, if some random person at the supermarket asked you any of those questions in person you'd definitely raise an eyebrow and walk away. SocialEngineering++

Exactly. If you used the internet from the beginning you clearly saw how Facebook is a data harvesting scam. By late 2010 I'd chatted with several engineers at Facebook about how they were clearly storing all chats/messages indefinitely, even if you deleted your PMs. So to the dude up there downvoting me, yes, many of us in tech knew this very early on. It was literally the goal of Facebook from the time you ever heard about Facebook.

Sure, but “and not be tracked” is the part @Maarten88 (and I) don’t believe will happen even if you do pay.

I think for me that a pay model without advertising would've been enough to prevent me from quitting, but now that I no longer have to put that energy into the website and my emotional associations with it, I am gone.

Reply on news.ycombinator.com