Hacker News new | past | comments | ask | show | jobs | submit login
Is social graph portability workable? (digitopoly.org)
58 points by avyfain on July 7, 2017 | hide | past | favorite | 39 comments



This highlights a recent episode I went through.

I decided to delete everything on my FB account. I didn't want to close, just delete years of photos, posts, likes, comments, etc. Everything. Then these two surprising things happened:

Having used Google Takeout many times, with great results, I decided to do the same with FB. For years I used FB as my primary way of storing and sharing photos from my phone (vs my desktop for a DSLR), so there were a lot of photos there that I didn't have anywhere else (pre-Google Photos times).

When I downloaded, I was surprised: Facebook downsized all my photos to 90's-era 800x600 pixels, and stripped ALL metadata from it (date, hour, location, camera, etc.). This pissed me off so much. FB wasn't just "not doing a good job", they were purposefully sending me a big "fuck you" letter as I removed my content. FB still had all my images in the original size, and with the metadata, but they chose to give me a shitty version of them.

This isn't laziness or incompetence, it's just plain evil.

My second surprise was how hard it is to delete content on FB. There's no "delete all", or even a "select and delete". You need to manually click 2-4 times for every single comment, post, picture, etc on a laggy user interface. It almost looks purposefully designed to prevent you from doing so.

There are some Chrome extensions that try to automate that, making deleting 1000's of items faster, but FB is constantly trying to break them.

So the message is clear: FB doesn't give a fuck about you, and will try its best to fuck you if you want to delete content or even get it back. Don't use FB as content storage, don't expect to have your FB content back.

PS: I went ahead and deleted everything anyway. Out of spite.


Why not use a Facebook App to extract all your photos and videos?

You can have a desktop app get an api key and grant it photo permissions etc. and go to town.

Full resolution everything.

Also you can remove all your posts and comments this way.

Perhaps such an app already exists. But if not, do you think FB would revoke keys of such an app?

FOUND THEM:

Odrive

Fotobounce


The existence of workarounds for this behavior on Facebook's part makes it zero less shitty of them. Those images are mine (the general me); Zuck, et al, do not get to leave me a down-rezzed, EXIF-stripped version as a parting gift, in the event I choose not to play in their sandbox any more.

Except, of course, that they do. Because we let them.


But they aren't exclusively your images, you gave them to Facebook when you uploaded them. Facebook is not a free image hosting or backup tool. You should never have had an expectation you could get back anything that you put in.


That argument would be far more (read: "at all") compelling if they weren't keeping the full-rez, EXIF'ed images. At that point, they're de facto acting as a backup service, and obscuring my access to my content [0] in the form in which it was originally submitted.

Your argument is the very thing I was referring to with the last sentence in my previous post. "They can do what they want with it, and you're wrong for thinking otherwise" is, besides being factually void, the kind of rationalization that lets FB jerk its users around like this.

[0] Yes, in fact, it is my content; I granted them a specifically constrained license, allowing them to use that content, in order to serve my and others' use of it. That's all. There is nothing whatsoever about not having to give back exactly what I uploaded, under any circumstance.

From the Fine ToS [1]:

> 2. Sharing Your Content and Information

You own all of the content and information you post on Facebook, and you can control how it is shared through your privacy and application settings. In addition:

For content that is covered by intellectual property rights, like photos and videos (IP content), you specifically give us the following permission, subject to your privacy and application settings: you grant us a non-exclusive, transferable, sub-licensable, royalty-free, worldwide license to use any IP content that you post on or in connection with Facebook (IP License). This IP License ends when you delete your IP content or your account unless your content has been shared with others, and they have not deleted it.

Nothing in that snippet, or anything else in the ToS grants them the right only to give back "lesser" versions of the content I uploaded. To suggest otherwise is tantamount to saying that, should I use the tool they provide to download all of my content, I have to suck it up that it will have all vowels elided.

[1] https://www.facebook.com/legal/terms


You own the content, not the files, there is a difference. Nothing in that snippet says they have to give you back anything.


Actually, Zuckerberg said:

“People own and have control over all info they put into Facebook and “Download Your Information” enables people to take stuff with them,”

“At a high level we’ve built two different things, Facebook Connect — which is our real effort to bring our sites to other sites, and “Download Your Information” where you can download your information and upload it to another site. Stuff that you put into the site, you should be able to take out.”

As usual, it looks like this was just PR.


> PS: I went ahead and deleted everything anyway. Out of spite.

Yeah, they still have the photos though. IIRC, when you upload a photo to Facebook you grant them a worldwide unlimited license to your photo, so there's really no reason for them to delete it. I suppose they can even use it for marketing purposes.


I've deleted all FB content last year (from activity history and timeline) and I regulary find few old posts in my already empty timeline. Looks like they will emerge from some dark corners of database for years.


I wonder if EU data protection law has something to say about this, especially with the GDPR coming into force next May:

http://www.bbc.co.uk/news/business-40441434

"We can also demand to know how our data is being used and withdraw consent whenever we like."

Combine that with the controversial "right to be forgotten", and I imagine that someone could successfully sue Facebook for lack of a "delete all", and for keeping images around on their servers after deletion.


It took me two days to delete all of my old Facebook content.


Zuckerberg for president!


[flagged]


[flagged]


I don't think this is useless.

A few weeks ago there was a similar case, and surely there are many others. Comments like this should make users think how such use case is radically wrong, for more than a reason.


I'm glad to see that most people here disagree with you.


You are complaining about free stuff.

Stop complaining about free stuff.


What is sad, is how similar the ideal portable social graph looks to email with your own domain name. You own the identifier, and you can select which provider hosts your email. In addition, you can send an email to somebody which will just work no matter what host they are on. You have the option of having a local copy of everything.

When we look back, I think we will see that in abandoning email as the medium of social interaction, we basically gave up on the open internet.


Email really isn't totally abandoned.

People in various compromised situations use it because it doesn't need a phone number (unlike a rapidly growing amount of social networks.) I've set up dates over email, most of my work is done over email. I tell people to email me rather than text me because I am /much/ more likely to get an email (and get it quickly, thanks to IMAP IDLE.) It's the only thing I get notifications for on my laptop and the only thing I reliably get notifications for on my phone.

It's not abandoned, it's just not as hip as snapchat.


I think it's more the abandonment of federated protocols in favor of centralized services, than the abandonment of a specific protocol.

An only tangentially related anecdote: SMTP and later XMPP offered beautifully portable social graphs. In fact, in 2012, I found it so portable that I decided to migrate off the self-hosted postfix/dovecot/ejabberd install I'd run on my own domain for years to GApps. Cue the replacement of GTalk with Hangouts, and now I really don't have much of a choice but to pay Google each month, or discard/re-establish the social connections I've built over the last decade and a half.


Email is the only truly successful decentralized protocol which is why the OP's statement mentions it.

A) Wide global adoption

B) Decades of use


> Email is the only truly successful decentralized protocol

HTTP?


HTTP is a transfer protocol, not a semantic protocol. Email at least lets you identify the sender/receiver and other metadata and separate them from the message, HTTP is just "this is the filename, this is the data.." It's not much more than a TCP stream really.


> Email at least lets you identify the sender/receiver

I'm not sure of your point. Http also allows you to identify the server. Forwarding mail servers are analogous to caching and forwarding proxies.

> other metadata and separate them from the message

Just like http.

> HTTP is just "this is the filename, this is the data.." It's not much more than a TCP stream really

No, it isn't. No more than SMTP, IMAP, and pop3 are.


By "sender" and "receiver" I mean person, not server. I truly don't understand why you're on about servers and transport protocols when this thread is about social networks.


> By "sender" and "receiver" I mean person, not server.

What's the difference? An email address isn't a person, nor does it represent one.

> I truly don't understand why you're on about servers and transport protocols when this thread is about social networks.

Because it was mentioned that SMTP is the only decentralized protocol, and I disagreed.


Email is still the de facto protocol for business communication, so you could probably build a professional network on top of it.


The social graph is only a small part of it. To make this work would require full federation so that, for example, someone on Facebook could friend someone on G+ and posts would flow in both directions so the timeline would work as normal. Like Ma Bell before them, Facebook will fight this to the death.


And it isn't even remotely clear how this would work. Like all "let's just semantic web all the things!" projects it all makes sense at a super-ultra-high level, but when you try to manifest this in the real world there's an immense explosion of tricky details, by the literal thousands.

Making social networks speak together is probably a good order of magnitude harder than unifying the IM networks, and even after a decade+ of effort and tons and tons of work, the multiprotocol IM clients are still generally quirky, involve losing useful features from any given specific network, and face constant issues where the semantics of the networks differ. (Especially around conferencing.)

Shared social graphs is a nice idea, but it would take enormous effort backed by legal force to happen, at a minimum.


The 80/20 solution would be to "just" free the authentication aspect from Facebook's control. If Facebook were forced to support something like OpenID Connect, then anyone with one of those IDs, from any provider, should at least be allowed to log in to Facebook and view posts from their friends who have accounts there.

Facebook would be forbidden from storing any information about you, except for the fact that your friends on Facebook had manually specified your ID as someone who could view their posts.

If your ID provider also ran a social network, and allowed people to log in with their Facebook accounts to view your posts, then that would cover a lot of use cases.

The next step would be to allow/require Facebook to fetch RSS-style feeds of events from your social network, so that Facebook users would be alerted to your new posts, without having to manually browse to all the different sites that their friends are on. This would require more engineering work, but could perhaps be implemented as a browser plugin.

At some point, if Facebook are making it unreasonably hard for competitors' users to interoperate with users on Facebook, there would be grounds for an anti-trust case against Facebook, which I'm sure the EU would be happy to consider.


I think that solving the technical problem is in the same order of complexity as the IM problem actually. It is solvable if you have some strong backers and not just the nerdy volunteers doing it in their leasure time, the ones who are telling you to use XMPP since 2004.

Actually the XMPP protocol already has a distributed social graph, and there are micro blogging extensions (eg. Movim), so we can get a functional distributed social network basically overnight.

The problem is that (a) you can't monetize an open, distributed network, so the incumbents will do everything they can to prevent federation, and (b) the only driving force that remains are the nerdy volunteers, who have a limited impact, even if they are able to tackle the UX challenges.


You don't need them all to work together. You just need one federated one that is functionally better.

It seems to me that a big problem is avoiding feature creep. I don't want or need a full browser experience for everything because it will break any customization I have set up.


Not necessarily. Linux arose despite Windows' monopoly. Wikipedia arose despite Britannica et al.


Linux didn't arise where Windows had a monopoly.

Wikipedia didn't arise where Britannica had a monopoly.

The industry changed, and the segment where Windows had a monopoly became less relevant, the same with Britannica.

Windows still retains 90%+ market share on desktops.


The same can happen with Facebook.

By the way - I am surprised Windows is used so much on laptops because I see mostly Macs since 2008!


Yes, it can! And that's why it bought WhatsApp, Instagram and tried to buy Snapchat. Defensive moves all the way.


It gets even more complex when we think about people’s private and public identities. Some of my Facebook posts are public and I read many public posts from the media, fan groups and companies. That is all part of my social graph but how would we work all of that? That said, there may be solutions there. The larger issue is how these links work is constantly evolving yet having a consumer controlled social graph may make it difficult to be responsive. After all, think about how you manage the social graph that is your pre-programmed fast dial numbers on a phone (if you even do those things). They quickly go out of date and you can’t be bothered updating them.

This is a hard problem to solve. We solved it at Qbix. We're still working on releasing our Platform to everyone, but for now it's already open source and we are dogfooding it in all our own apps.

https://github.com/Qbix/Platform

We are also working on an interoperable auth protocol where you do control your identity and keep your contacts private, but discover them across communities (venues and interests).

https://github.com/Qbix/auth



Interoperability is already settled law (railroad gauges, telephone networks.) We just don't care to apply the law 'cause democracy is remarkably corrupt in our time.


Different social networking applications imply different social contexts. I may follow you back on Twitter, but that doesn't mean that I want to be your friend on Facebook.


You could still have multiple different social networks per social context, which would compete on features, as opposed to lock-in and network effects.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: