Hacker News new | past | comments | ask | show | jobs | submit login
Diaspora version 0.7.0.0 released (diasporafoundation.org)
128 points by chtfn on Aug 20, 2017 | hide | past | favorite | 62 comments



Someone should run a Diaspora pod with the following model:

1. They market it like crazy in the hopes of getting something like a few million users

2. They publicly state that they will mine all metadata for the purpose of generating an "inference report".

3. Every other week they release an "inference report" that reveals a new, dangerous way the seemingly innocuous metadata can be used. Some examples would include a) accurately gleaning more private data from the metadata, and-- if enough people join-- b) using that data to subtly influence the behavior of the participants.

Outside researchers are given access to the process in order to audit it. Anything revealed in the inference report would be assumed to already be happening on larger commercial networks.

Users would remain as long as they believe the value of the inference reports outweigh the risk to them of using the network.

Edit: formatting


1. They market it like crazy in the hopes of getting something like a few million users.

That's the problem. Nobody uses Disapora*.

I got an account a few years ago, but abandoned it after discovering there was nobody on there worth talking to.


The idea sounds like Matadors day job.

Everybody want to see the results, rarely who will want to actually be there.

How do you incentivise it?


I like these guys' perseverance. But part of me can't stop the feeling of sadness that they've spent a better part of a decade working on it, perhaps the best years of their lives. I know perseverance and failure is seen as good things in certain circles, but I don't think I've ever read a biography where the author spent ten years on something that ultimately went nowhere and then went on to do things of enough noteworthiness to merit said biography.


I spent 8 years working on Appleseed (also a federated social network, 2003 to approx. 2011) and it was worth every minute, even though it never went anywhere. Working on a problem that nobody has been able to solve gives you really good insight into software engineering that you can't get when you're implementing other people's solutions, no matter how complex those solutions are.

Even though I failed, I benefited quite a bit by taking on something so ambitious, and whether Diaspora succeeds or fails, the people working on it will most like be able to say the same.


Hey Michael, good to see you around :)


Oh, hey Matt! Good to see you, too!


> I don't think I've ever read a biography where the author spent ten years on something that ultimately went nowhere and then went on to do things of enough noteworthiness to merit said biography.

Didn't it occur to you that even if diaspora goes nowhere, the work these guys have been doing may be influential or even the foundation for something that will actually go somewhere in the future?


Nelson Mandela spent 27 years in jail before becoming president and father of a nation.

While "only" 4 years, Muhammad Ali sacrificed what would have been the prime of his boxing career as a conscientious objector. He was famous as a great athlete before, but transcended it after.

Alexander Calder was an engineer for about ten years before becoming one of the great artists of the 20th century.

Many other examples.


That you don't hear about the ten years of failure before the success says more about what sells biographies than it says about the founders.


As far as I can tell from the commit log it's a quite different group of people who are maintaining this now as an open source project than the four people who launched it from the Kickstarter 7 years ago.


Yep, we (the community) are now working on it as volunteers. That means no promises and no pressure. If you look at the github contributors graph, you will see that many of us often take a break of a month without contributing and then come back, fresh and more motivated than ever :)

Any way, the goal is not to kill facebook. It is to experiment how a decentralized social network could look like, and provide a solid alternative to those who don't want to see their data analyzed. It's nothing more than that at the moment.

For the first point, our protocol is improving (see https://blog.diasporafoundation.org/43-our-federation-protoc...) and is the only one really production ready imo, and for the second point, the user experience is getting better at each release, with new features.


"Ten Year Overnight Success" is almost a cliche in the music world, and of course most musicians (even very talented ones) never hit it big at all.


so, my life goal should be to spend my time such that you (or anybody else) would be willing to read my biography?


Isaac Newton?


You should read some Newton bios.


Abraham Lincoln, albeit not a scientist.


We have fun. ;)


While I completely support the goal of diaspora, I don't think that I can use it. one of the main benefits of Facebook is that I can look up someone's information / posts without them being informed. Imo that is what made Facebook so big in the first place.


Zuck et al probably realised very early on that casual stalking was a major use case. Not in a dangerous way, but maybe you wanted to find out a little more about that guy or girl before you asked them out, that sort of thing. They were all in college at the time after all. So he was careful not to include any feature that would inhibit this.


This is my biggest pet peeve about Facebook. I'd pay money for better analytics about my posts.


I would love to see the same stats on my personal posts as I do on my page's post. How many people have seen it, the breakdown of their demographics, etc. The problem is, seeing those stats on a page (and a page that I pay FB to advertise for me) is really disheartening: the numbers are incredibly low even though I pay FB to show it to people. I'm sure the numbers are similarly low on my personal page as to how many of my 100+ friends actually saw my post come across their feed.


For me it's the sense of ownership that's missing. If I hosted it myself and just looked at data in the server logs - not even fancy javascript analytics - I'd get more information. Not having access reminds me that Facebook is using me to collect data on all my friends.


I think that's going a little far... HN/Reddit doesn't give me server logs either. Very few web properties do. Very very few.


You can create a "Facebook Page" or convert your account to a Page instead.


Is there a bigger community actually using Diaspora? Maybe I'm outside of that bubble that's why I'm wondering.


I'm only peripherally attached (I've got a ... long-dormant Diaspora account), but there are a number of semi-inter-related open-social projects, which talk to various extents. "The Fediverse" includes a few of those, and I think it has bridges/gateways to Diaspora.

Frendica is another, I think, related protocol.

The problem generally is that the communities are small, diversified, hosting is a hurdle for virtually anyone, and individual instances can be finicky.

(On Mastodon -- a different technology entirely, but similar in concept -- I have two accounts on different instances, and since April have found that one or the other has been down, unavailable, or technically unusable for up to weeks at a time.)

Should some nucleating group decide that they were going all-in on Diasapora (or a compatible tech), that might make a difference. Meantime, everything seems stuck in slow-start mode.


Cool. I just signed up on Mastadon. Have you tried patchwork yet?

The slow-start mode seems to be default mode for all new social media platforms while FB is dominating the space. I could see a service business rolling Diaspora out on-site to organizations. I think poaching users is fair game and should be perfected across all the platforms. Basically, sharing on FB links that are hosted on a Diaspora instance. So the act of sharing on FB increases exposure for the instance.


I haven't.

https://mammouth.cafe/@dredmorbius

https://mastodon.cloud/@dredmorbius

First is primary, though I hop back and forth a fair bit.


Mastodon instances are actually part of the Fediverse (OStatus)


As I'm aware.

Largely through the whinging of previous Fediversalists.


Wikipedia gives 677,000 as the size of the Diaspora community

https://en.m.wikipedia.org/wiki/Diaspora_(social_network) citing The Federation: https://the-federation.info

That's ... on the order of Usenet ~1990 or so, per personal conversations with Gene Spafford a few years back. Where "OTO" could be 500k - 5m users. Much of Usenet was far smaller.

Microsoft did some studies on Usenet nodes and behavioural patterns in the early 2000s, and got some usage numbers out of that, though I'd have to dig to find them again.


I may have those too; I was at a company that bid to host microsoft.public around that time and their views on Usenet were very enlightening. I'll see if I can dig them up.

(Funniest part of the requirements spec: it had to "appear to be" hosted on a Windows server...)


That is ... amusing.

As I recall, there was some degree of contention over whether or not Unix (a/k/a not infrequently Linux), or servers from a certain Redmond, WA, based company were superior and/or at all capable of general Internet hosting, at the time.

Hotmail comes to mind as well. FreeBSD.


Hotmail and Windows Update used to be hosted on FreeBSD, yes.


It's not an overwhelming lot, so I wouldn't expect you to find someone on there that you already know, but it has this sort of early internet forum charm, where you can easily get to know a handful of strangers and hop on there every now and then to talk to those.

Also, its community does very much consist out of the more technical crowd, with a focus on privacy and it also helps if you don't mind the occasional free software activist shouting about. That's just the group that's most likely to sign up to it, which can be a blessing as even with so few people, you always find someone to talk about tech stuff, but it can at times also get somewhat old...


I think it would be really interesting to build something like dispora as a suite of dockerised micro-services which would allow you to add and extend functionality based on a new social-networkong protocol in a language-agnostic way. It would also be very easy to incorporate email and xmpp which are already federated protocols.


Docker should help in general with making services like this more accessible. I can imagine someone like Fastmail using docker to allow a one click process to set up a Diaspora or Mastodon node. That should take the hassle out of it, and if you could just watch a 3 min video to explain the concept/metaphors, and set it up in 30 seconds, it expands the possible audience from system admins and software engineers, up to anyone technically inclined.


I don't know if these guys have funding or not, but they deserve it. There is too much centralization nowadays. Too many eggs in the same basket.


Yes, too much centralization going on.

But ... isn't Diaspora trying to solve a too big problem? Ideally, shouldn't the decentralization-space be filled by smaller services? For example, in my perfect world, instant messaging and newsfeeds would be two separate projects. That way, not only is the information decentralized, but also the development of the decentralized web becomes more decentralized (in a way).


What people like nowadays is integration. They don't want to have multiple accounts on multiple services they each have to trust. They want one sign-in and everything available. That's why google is so good to trap you.


I think Github could create a good nucleating community for federated networking. They could host a giant pod of Github users (could be the next venue following slashdot>kuro5hin>reddit>hn), and they could also create 1 click tooling for Github users to create pod for their own project communities.

Would be a good kickstarter for generating sufficient network effect value to use in the first place, evolving the platform software through real world tire kicking, and getting it out to the wider world. And being github, as well as a sufficiently technical userbase, there's already a business model around freemium hosting which could be applied.


diaspora is flawed by design. it will only ever be used by techies b/c either you take care of your own server or you have to trust somebody with your data.

what is needed is a facebook e.V. with a strict data safety policy.

https://en.m.wikipedia.org/wiki/Registered_association_(Germ...

it would cost transparently what is needed to sustain the service.


That's not helpful. "data safety policy" is only as safe as you trust the people running it, and as long as you trust the government not to change the playing field.

Historically, neither of these tend to be sustainable.


With your attitude you cannot trust anybody.

Also government is nothing bad. A part of a good data safety policy is absolutely also including cooperation with law enforcement. If you think that your government is abusing its power then choosing super-crypto is not the right battle field - you have to become politically active.


> you have to become politically active

Okay, fine, if you have the energy for it, but even if you dedicated yourself to it fully, the most that you can hope for is a small incremental change back from the brink, around the time you are old enough to retire. What are you going to do with your life in the meantime? If your government is abusing its power, which most generally are, then for practical purposes you have to treat that as a given and work around it.


The problem with Facebook is that one entity has access to all the info.

I don't trust any single entity; they are too easy to pervert. Just one more kind of bad monoculture.

Strong crypto and political activism are not contradictory; they are complementary


Your comment contradicts itself. A facebook e.V. with a strict data safety policy would be nothing more than another organisation running something similar to diaspora.


I don't think it's necessarily a contradiction. You either have to run your own server, or you have to trust someone else at which point you might as well just use Facebook. The entire point of Disaspora is that you don't have to trust anyone. Once you have some random, unknown person hosting it for you, you're just using Facebook but with fewer people.


You need some server that these private pods can push to or pull from.

Or you have software that manages your local disaspora server or in the cloud or darknet.


here are the options:

data is controlled by:

a) yourself b) your hacker-friend c) a profit-oriented company d) a non-profit-oriented organisation

a and b are the restrictions diaspora faces.

Facebook is c - not good for well-known reasons.

d is an organisation responsibly handling your data.

no contradiction.

the organization though could in fact be based on Diaspora and maintain pods.


That's what https://framasoft.org is doing. They have a https://degooglisons-internet.org/ project (degooglify the internet) and are providing alternatives to every big corporate services out there. https://framasphere.org is their installation of diaspora* built against facebook. And they are non-profit. (You can switch the language to english if you do not speak french)


MediaGoblin and OpenBazaar also fit their Fediverse.


I'm so happy to see this release, especially with the codebase cleaning. The diaspora* protocol is now rock-solid (see https://blog.diasporafoundation.org/43-our-federation-protoc... ) that means the federated social web is now free and can extend!


Does event management exist yet? I.e, can my friends and I create events and invite people to them like you can on Facebook?

Also, ISTR from last time I played with it, there was no ability to create photo albums/collections. Does that exist now?

Without those two things I couldn't even begin to consider getting people to start using it.


Groundwork has been done, so it is definitely planned: https://github.com/diaspora/diaspora_federation/pull/44 Hopefully, this gets a bit more attention now that the big cleanup work of the codebase has been done.


> Does event management exist yet? I.e, can my friends and I create events and invite people to them like you can on Facebook?

No, not yet. See https://github.com/diaspora/diaspora/issues/1359


Lack of integration with their calendar right from the off is why Google Plus flopped, I reckon.


How about for messaging and updates


If I and my friends are on the same pod, and I get banned from it, can I still talk to them?


What do you mean by "get banned"? If you mean you violate the term of use and the podmin closed your account, you can create one on another pod and re-add your friends for sure.


I'm just not clear about the federation works I guess. If a power hungry mod (or maybe a good guy coming under very intense pressure from someone else) were to hate me, could he block me from sending messages to my friends by refusing federation somehow?

I dont want to be on a platform that can hold my friend hostage, is why I'm asking.

EDIT: If hub federation is needed for asynch messages, and unblockable (by hub owner) p2p used for chat when both users are online, then I guess that would be acceptable.




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

Search: