Skip to main content

On the risks of a federated option

On the risks of a federated option

I expect most people want from their social media platform the same things they want from their car. They're busy and they want it to work every day with no necessary effort beyond routine maintenance. They want it to start up for them in the morning and to take their kids safely and reliably to school.

(Not everyone has that, but I expect that's what we all want.)

I drive a nice trusty Honda. It's almost 12 years old and has few bells and whistles. No Bluetooth. Manual seat adjustment. But it gets me where I need to go and has done so safely for many years.

Recently, there was a problem with the airbags from one of the company's suppliers. They sent me a note. I scheduled an appointment online. The dealership replaced the airbags in about 30 minutes at no cost to me.

Most of us here -- or at least lots and lots of the people who now have to move -- aren't programmers or sysadmins and don't want to be. As it happens, I'm about to marry a Japanese woman, and if I'm going to learn a language, it's going to be Japanese, not Ruby or Python or whatever. [Insert your tool of choice.]

Most of us have mortgages, children, difficult spouses, and careers in non-technical fields that aren't progressing as rapidly as we hoped and which have their own unique demands. We have non-technical hobbies. We make all kinds of art, or cosplay, or in the case of one of my contacts, woodworking.

And that's not counting the very old, the very young, people right now suffering life-consuming events (or who soon will be), anyone dealing with severe anxiety and depression, and so on.

All those people want a finished product: a user-focused solution, not a developer-focused one. They want their social media platform to work like their car. They want it to start every morning and to be safe for their children.

I love that so many of you are so excited about building something new. It's fascinating, and I hope it starts a revolution. I have no philosophical objections to a federated system. I do have lots and lots of practical ones, such as what happens when people running pods get divorced, or go on vacation, or move, or get sick?

Think. Once a community reaches any kind of critical mass, it will become a target for spammers, hackers, bots, creepers, and trolls. How do you stop that when there isn't even an option to block people?

What happens if a pod owner's machine gets stolen or their cloud account hacked?

What prevents pod owners from acting maliciously? What protections keep people from stalking my fiancee? Or my kids? What recourse do I have if that happens?

How do I get my data? Is everything I contribute lodged permanently on the platform? What about data loss? Are there backups, and if so, how do I know they are secure from all the folks mentioned above?

That's a sample. The list goes on.

I don't see where Diaspora, or indeed any of the federated options, have adequate protections for the mentally handicapped, for frequently targeted social groups, and so forth. Mostly they serve experienced, technically savvy users.

If that's you, great. I'll not stop you. Full steam ahead.

Everyone else, just please keep all of that in mind as you make your choices.

It would be nice to keep the community together, but right now it seems to be fracturing along these lines.

Comments

  1. Yeah, it's a give-n-take. The bad thing is that a neonazi can make a neonazi pod where neonazis can congregate and be terrible people. The good thing is that every other pod can refuse to communicate with it, leaving it a little walled-garden where they can be awful to each other without bothering the normals.

    Thinking maybe a "nice trusty honda" solution would be for me to pay an existing pod-owner to create and maintain my game-programmer-pod. And maybe get a few friends to kick in a few bucks. So we're no longer moderators or operators, but benefactors.

    ReplyDelete
  2. Donna Buckles of course. I hope people do.

    ReplyDelete
  3. Google+ Please control the amount of porn invading your site. Is there one decent technician left who can solve the problem?

    ReplyDelete
  4. Mari Christian Most likely not. I would imagine it's on autopilot now, and by next August G+ is gonna be such a mess that everyone will beg for it to be put out of its misery.

    ReplyDelete
  5. I think it has become clear that we won't find an alt soc. media site that has all the features of G+. They have different strengths and weaknesses so maybe we need to be realistic and use the best features on offer in several sites. G+ is not perfect, and what a headache it has been putting up with the faults and bugs caused by the upgrades.
    It's not healthy to become emotionally attached to a social networking site - they, like people , are liable to let you down or die at some point. We need to be realistic and resilient.

    ReplyDelete
  6. Very well said. Having been a 'victim' myself to a Friendica server suddenly gone, I can't stress it enough that not even a decentralized/federated system is going to solve the problem we have with centralized/proprietary social platform. Sometimes we have to tolerate the defects of an online home we wanted to move in as long as it provides us what we need.

    ReplyDelete
  7. Rick, hi! I'm a professional privacy advocate.

    Nearly every objection you have here to pods is an objection privacy advocates have to Big Data.

    We got here because Google abused our data privacy rights -- and didn't inform us.

    Google has regularly been compelled to hand our data to the feds without informing us. They post aggregated data on this quarterly.

    We had the #nymwars over their handling of privacy for people who wanted to use pseudonymity to shield their online identities from their identities on the ground for many of the reasons you cite -- sexual minorities, people subject to stalking, and so on.

    None of us have the recourse as individuals to take on Alphabet for reparations if we come to harm from this social network. If we did, we'd be discussing suing them to keep g+ open, or for damage to our businesses for shutting it down or something. After all, so many of us are Americans. :)

    So your argument is MOOT.

    It comes down to this: do you trust a giant corporation who serves shareholder value as a first priority legally as a first priority, where your data is a commodity and you are not a client?

    Or do you trust your fellow users, as in an open source community, where relationships can be fractious, and people can flake?

    Both have problems. It's paternalism vs. actual (not fake) community management.

    Here, we are the product. And what just happened was inevitable. It was just a matter of time.

    If we go to a private user-run platform? It running into problems is only a matter of time.

    Your car will not run forever. No one can make it last for the rest of your life. Even if you are an engineer with a good salary for the rest of your life, paying for it for decades won't keep it running for decades, it will wear out and the company will stop maintaining it.

    The car manufacturer may go out of business or start getting a bad rep for good safe cars. You may not want to put your kids in one of their cars anymore.

    Your analogy doesn't hold up.

    Besides, makers don't make cars.

    But our peers do make social net alternatives. Please evaluate apples to apples. You are using the politics of fear.

    The problem is, we have just as much reason to be afraid of Alphabet (and their alternatives) if not more. And ffs, we were just burned, I'd think that would be fresh in everyones' minds.

    There is no life without risk. As an engineer, you should understand that.

    ReplyDelete
  8. I've been advocating since the 1980s that we need not only federated systems but also redundant systems. You should be able to have your content stored by multiple federated servers, not just the one where you initially established your account, either as a mutual agreement (you store mine and I'll store yours) or for payment. This has always been the Xanadu publishing model. I continue to hope that we get incrementally closer as people see the merits.

    ReplyDelete
  9. You've pretty much listed my objections to "pluspora". There are things I like about it, and I'll certainly keep giving it a try, and maybe they'll be able to address some of this stuff.

    ReplyDelete
  10. Mostly they serve experienced, technically savvy users.

    If that's you, great. I'll not stop you. Full steam ahead.

    The rest of us who have other priorities are going to get left behind.

    ReplyDelete
  11. Shava Nerad if someone breaks into my house, I call the police, not my neighbor. If I get into an accident, I call my insurance company, not my neighbor. If I need a new car, I go to a dealer, not my neighbor the mechanic. YMMV. The analogy works fine.

    ReplyDelete
  12. Kathryn Huxtable people will miss it, so I'll keep repeating. I'm excited about the project. It's just far from ready. By all means, folks should set up an account if they want, and be active. Just recognize the risks and don't make it your only home -- for now. In the future, it may be different. I hope it is.

    ReplyDelete
  13. I would hope that people will be able to build services for all types of users in time, including those who want an easy way to self-publish their thoughts and those who just want to pay someone else to publish for them. That has certainly always been our goal, and I'm sure others are thinking similarly.

    ReplyDelete
  14. Andrew Pam Sure. The key phrase being "in time."

    ReplyDelete
  15. Rick Wayne

    So, you are putting Alphabet in the position of a world government now?

    See, if I have a problem that needs fixing, I evaluate it.

    If it needs the police, I call the police, but amazingly few problems do.

    But if someone is, say, picnicking where I think they shouldn't be, I might talk to them first as a neighbor rather than getting on my cell phone. Or if they are selling lemonade and I think they might not have a business license.

    Again, you are picking bad analogies. I can too.

    If you get into a car accident, and your car is old, you can settle with the other person if you don't have full coverage without consulting your insurance company. Obviously you've never owned a beater.

    I personally have never in my life owned a new car. I think it's an incredible waste of money -- the value goes down 50% when you sign papers and turn the key.

    Your arguments are arguments of privilege. Not all of us come from that position, even if we have that option. Many of us would rather rely on community and cooperation among our peers than on coercive authorities, than on Big Brother, if the choice is binary.

    Generally I find the tradeoffs good in some cases, bad in others. But to prevent them as all good or all bad is false reasoning, and as the group moderator, you are going to hold more weight here, so I'm going to engage that strongly.

    Indeed, YMMV, but that's not how you presented it.

    I am just trying to present the alternative case.

    I have made my living here on G+ for years, even though I am a critic of Alphabet for their privacy practices and various, so I made my own choices (privacy is a slider where you trade your information for value received -- but ideally you'd understand transparently what you are giving and receiving and no one does -- I am not an absolutist), and I will miss this platform incredibly.

    But I am here to validate the people looking at pods and other alternatives that give them independence from Big Data.

    They are taking no predictably greater risks than you will be going to another Big Data platform. Just different risks.

    I challenge you to come up with anything at all that contradicts that, in this environment of data breaches, API screw ups such as this recent one at Alphabet and Facebook/Cambridge Analytica, and so on.

    There is also the middle ground -- small, community responsive groups such as Dreamwidth where social blogging moved from LiveJournal when the Russian government mandated anti-LGBT sanctions on content to LJ. There are still "small data" corporations that are years long viable platforms for social networking, with strong community.

    Those might satisfy a good many of your and my concerns, although they are still subject to government intrusion and centralized datastore breaching, which bothers the true tin-hat privacy crowd.

    ReplyDelete
  16. Shava Nerad you don't seem to have actually read the post, especially the last few lines, so this will be my last reply. Feel free to take the last word.

    I'm not advocating a platform. I raised what I think are very important questions. You think the risks are worth. More power to you. I explicitly

    You seem to take exception to anyone seeing things differently than you. As someone who was doxxed, who was trolled mercilessly by the alt-right for months on a different platform, my answer is NO, I don't trust my fellow users more than Google. I've met several people in targeted groups who feel the same.

    I don't trust Google either, but I expect they'll at least act in their economic self interests.

    And you lecture me about privilege.

    I raised some very real, very open questions. Your answer is "but everything else is worse!"

    If that's true, it still doesn't abrogate those questions, and I think people should be aware. I like arming folks with information and letting them make the best decision for themselves, versus telling them the best decision for them.

    Clearly your mileage does vary there.

    ReplyDelete
  17. I'm sorry that happened to you.

    I also think it's impeding you from hearing what I said. But I hope that the people who go through our exchange can draw a balanced reading from it. Good luck.

    ReplyDelete
  18. I found what you said useful and helpful Rick Wayne

    ReplyDelete
  19. I also think what Rick Wayne said are all valid points to consider and all make sense to me. The flaw (if any) in the car analogy does not matter.

    ReplyDelete
  20. Far as developmental disabilities go:

    Our solution for our adult daughter has been a smartphone running an app-limiting and app-locking app. She can not be trusted with unlimited web access, she will certainly start pilfering aides' card numbers once she figures out The Internet Is Retail Heaven, she for sure gets no social media access, but I wanted her to have a phone and an ability to text or call family, and contact her house manager in case of emergency, and I (need to double check if) turn on GPS in case of all the nightmares that beset the family of any adult functioning at the developmental level of "Oh Look A Puppy!".

    Fuhu/Nabi used to have functional sandboxed Android apps with their hardware but I suspect Hasbro who bought them is going to "sunset" (thanks, Google+ , for that shitty verb) all the useful parts of that.

    ReplyDelete
  21. I think you're right on the money with these questions Rick Wayne, and I'm somebody who's quite happy with the notion of running Internet facing web services myself. (I might have had the first web server in a Texas university, back in the 1992 or 93...)

    A pod running in someone's house over their ADSL might disrupt something I use for relaxation -- and there's not any clear way for me to know how reliable the hosting of a dispora-like pod is (ala pluspora) or how dedicated the owner will be to keeping it running. There's also the fact that I'd be trusting the folks at pluspora with my data every bit as much as I'd be trusting a walled garden. It might be that I am taking a larger risk of a smaller group not having the time to maintain their server software from a security standpoint, as they're dependent firstly on the base software maintenance and then updating their particular instance. There's also nothing to prevent the owner from installing a fork of the code that starts putting advertising in the mix after a good sized community is in place -- except the pledges similar to what some of the for-profit walled gardens like MeWe are doing.

    The virtual of the federated model is that it's fail-soft. Lose one pods, the other pods are there. But that matters not all if all the groups I am part of happen to be hosted on one instance, unless there's replication across instances with a failover strategy. If that functionality has been part of the discussion of diaspora, I've missed it.

    ReplyDelete
  22. Martha Magenta I agree, G+ quite successfully integrated several different aspects of social interaction and became unique because of it. It's unlikely an alternative with the exact same set of features will appear within 10 months time. So we'll end up having to pick a few different choices -- some are better for communities (e.g. with group support) and some others better for "world-public" communication, or collection style sharing. In any case, having to use multiple platforms is going to be an inconvenience.

    As to Rick Wayne original point, I mostly agree. If you feel hosting your own pod on a federated network is beyond your technical skills, it's well worth considering if the federated network is the right place for you to be. Or to extend the analogies: you should know your pod host at least as well as you know your next-door neighbour; and if you find you don't trust your neighbour that much then a central regulator may be the better choice. Both have their downsides.

    Also remember if you choose something that's free, it's probably not the best option for your privacy.

    ReplyDelete
  23. Alan Peery exactly. You replace a relatively well-guarded single point of failure (low probability but catastrophic), with open multiple points of failure, where there is a high probability that at least one node will fall, but that's low-impact to the overall network. Of course, if you're one of the people who lose everything, as some of the commenters here and elsewhere have experienced, then it's catastrophic for you.

    I am POSITIVE there are technical solutions to each of these problems. No doubt.

    One could develops a scheme with Disapora pods similar to what Freenet does with its nodes, where there is encryption and redundancies. Pod hosters can come and go. All you need is some minimum number to sustain the system.

    That's obviously not the only way to do it. Just giving an example to prove I'm not against these kinds of solutions.

    But we're not even close to this, and it's not going to be developed, tested, and implemented by a group of unpaid volunteers between now and when this place closes.

    ReplyDelete
  24. Rick Wayne I see that Di posted last night about their future plans. They know they can't take everyone. There are changes planned for Diaspora which may address some of the user experience issues people are having. I remain open minded about all these platforms. (And I am alright with FB up to a point. I'm certainly there, and my experience there is alright.)

    ReplyDelete
  25. Just a note that there's a Risks and Concerns section of the #PlexodusWiki to which much of this discussion and thoughts, both ways, would be extremely appropriate.

    Capturing and distilling discussion rather than watching it wash away down the Stream, and eventually to Sunset Oblivion, is valuable.

    social.antefriguserat.de - PlexodusWiki

    ReplyDelete
  26. Edward Morbius Yes, and you're doing a good thing collecting this stuff where it won't go away with G+.

    ReplyDelete
  27. It seems the only way to provide this trusty Honda is building a distributed, federated, redundant network almost from scratch. However, accessing it might require a dedicated client (be it browser extension, or an electron-based app like telegram client, or a plugin that will connect to a browser and sit in the system tray). Although there will have to be some central portal site, that'll allow registration and perhaps even using the network without installing apps, the data won't be stored there - otherwise it'll be that single point of failure we're trying to avoid. The question is - how many people will agree to install an app or a browser extension, and share some of their PC's resources to power the community?

    Redundancy means that data is still available when your PC is off, as long as certain minimal percentage of them are online.

    Distributed means your PC will store other people's data (in an encrypted form, perhaps inaccessible to you).

    Federated means there can be multiple points of access to the network, like gateways, and also infrastructure nodes that will try to provide seamless experience and enable features usually missing in distributed projects, like search or discovery of new people.

    ReplyDelete
  28. Kathryn Huxtable I really need abuse/harrassement, and any other issues you can think of, addressed in detail there by someone who can give a direct, informed, and where relevant (and it usually is) directly personal perspective on this.

    (Qualia -- directness of experience -- is a concept I've come to appreciate of late. "Autopsy" similarly means "to see for yourself". Cut the body open and figure out WTF went wrong. )

    You've been active in that before and know many others as well. Shava Nerad is excellent on this (if she's got the energy, and I know she needs to conserve spoons), but may be able to recruit others.

    Donna Buckles and I rub each other ( I more her than the other), but I really do respect her views. Most of the friction, I think, is me digging into what are very painful areas for her, trying to understand. If she's game to participate or talk with you and others offline / elsewhere, I think she's also got very good wisdom. We've been giving each other space here, but that's out of consideration on my part.

    ReplyDelete
  29. Edward Morbius I'm not really the person. I've been very fortunate on social media thus far. But I can direct some others your way. (They may, of course, not feel comfortable putting their experiences out there.)

    ReplyDelete
  30. Kathryn Huxtable Understood. Organising, chatting, and passing on what's presented -- concerns, capabilities wanted, types of problems / issues encountered -- would be useful.

    ReplyDelete
  31. I was told to come here as I've had some issues on MeWe while trying it out this past week. Will have to come back and read later.

    ReplyDelete
  32. Edward Morbius I'll get my thoughts together over the next week.

    ReplyDelete
  33. Kathryn Huxtable Much appreciated.

    Phase 1 is likely to last at least until January -- that's exploring options and stuff.

    ReplyDelete
  34. "I don't see where Diaspora, or indeed any of the federated options, have adequate protections for the mentally handicapped, for frequently targeted social groups, and so forth."

    "Frequently targeted social groups" includes women. It includes people of color. And that doesn't even touch the kind of organized harassment we are seeing all the time. I am old enough to remember the failure of Usenet. Oh, it's still around. But it's a spammy sewer. And now I see Diaspora replicating the same mistakes, and the advocates of Disapora making the same arguments.

    A famous editor and moderator once commented, 'I’ve done my time and then some on Usenet. If learning to moderate online forums is like studying trolls and demons, then hanging out on Usenet is like living in Sunnydale: if you survive long enough, you’ll eventually come up against one of every kind of monster—and after a while, your reaction will change to “Bored now.” ' Been there once. Don't want to go back.

    ReplyDelete
  35. And, I did not specifically endorse diaspora* for example. I find the lack of moderation tools disturbing.

    I'll note that diaspora* development doesn't use diaspora* for their own development discussions:

    https://discourse.diasporafoundation.org/guidelines

    Why we don’t use diaspora* for discussions
    The fact that we do not use diaspora* for discussions, but we have many reasons for it. diaspora* is a social network, not a discussion tool. It’s not designed as a tool to have long discussions that can span over months or even years. Many technical issues and, in fact, some design decisions make diaspora* less then ideal for having relevant discussions.

    In diaspora*, there is no “single source of truth”, no single place where all discussions are stored. That, obviously, is made by design since we are a federated network. However, that turns out to be a problem if you try to have discussions where everyone needs to have access to all information. Within diaspora*, it is not possible to “get a list of all discussions with all comments”, since frankly, we do not know where all those posts are!

    In addition, the core team has no way of moderating discussions if they are spread over many many servers all across the planet. By having a single, central system where everyone can come together and participate, it is much easier to have conversations that have some weight to them, making this the only way of having “official” discussions.
    ===

    diaspora* has a lot more in common with Discord than G+, really in terms of how one might be expected to use it.

    I, at least, use G+ more like USENET of old -- in terms of culture of discussions and followups and community ties, and the expectation of the persistence of content.

    But diaspora* is not the only non-big-data alternative under discussion.

    If I thought my readership would go there, I might well go to Dreamwidth.

    ReplyDelete
  36. Shava Nerad Fair enough. But then, what do you advocate? I am not aware of any open source solution that is even accessible to most users, nor any that does remotely adequate usability testing. And this is without even considering the troll armies.

    I like Dreamwidth and it addresses the troll army problem. BUT.
    - No support for small-screen devices
    - Weak support touch-screen devices
    - No app support on iOS and Android devices.
    - The most effective use of DW requires knowledge of HTML

    It is not, in other words, a service that will ever be of use to the vast majority of social media users. In principle all these flaws could be remedied. In practice, both funding and organization are problem. Most open-source solutions that are adequately funded are funded by industry consortia. Mozilla is funded by Google search revenue. Organizationally, the tools to make such a thing properly democratic do not exist yet; the trolls and the sociopaths are an enormous problem. Mark Bernstein's critique of Wikipedia provides a bitter example.

    :-(

    I want to see better answers than these. This failure of the cooperative internet is a small part of the vast failures of democratic governance of our time, but it is my corner of the world, and I would like to see it thrive.

    ReplyDelete
  37. "diaspora* is a social network, not a discussion tool" -- I suppose that explains why so often social networks feel so dissatisfying. And I guess it shouldn't be news but it's still eye-opening when stated so directly.

    Back to cat pictures.

    ReplyDelete
  38. I am a programmer, but I don't use social media for programming. I am also searching for a new platform where it's easy to use and user-friendly.

    However, I will mention the impact that some of these "developer-focused' platforms are making. One of the big flaws to a platform owned by a corporation is that the direction becomes political. Look at Twitter and Facebook and how controversial their decisions are becoming. On a platform where there is a collection of private instances, there are many owners instead of one powerful overarching entity. If I don't agree with the moderation of a community, I can switch to a different one while remaining on the same platform.

    Ideally every user will hold their own content locally and it would be their responsibility to use storage they trust, whether that's Apple Cloud or Google Drive or their own external drive. This way everyone owns their content while being able to share it on privately owned nodes connected by a couple master-servers for the only purpose of discovery.

    ReplyDelete
  39. Kyle I'm leaning strongly to Git, GitLab, an SSG, and then syndication for visibility through other platforms. My home will be wherever Git is, but I can get reach (and engagement if I choose it) elsewhere.

    That's the plan, at any rate.

    ReplyDelete
  40. +Edward Morbius That's not a bad idea. I've had musings in which I considered one central platform where my content would originate and then use the APIs of other social networks to "broadcast" my content to all of them. That way I can take advantage of the communities it many networks while keeping my original content in one place I trust.

    ReplyDelete
  41. Kyle the problem with this approach is to collect back the comments.

    ReplyDelete
  42. פליקס כץ There are a few approaches. Curating & selecting comments has advantages.

    ReplyDelete

Post a Comment

New comments on this blog are moderated. If you do not have a Google identity, you are welcome to post anonymously. Your comments will appear here after they have been reviewed. Comments with vulgarity will be rejected.

”go"