Hacker News Reader

Comment by jacquesm | Hacker News Reader

Key quote (for me) from the interview:

"A social network is disempowering because you put a lot of energy into it, all your personal data out there, and tell it who your friends are. You can only use that information inside the silo of that particular social network."

source: https://www.technologyreview.com/s/604052/webs-inventor-tim-...

View on HN

Tim's vision for a Giant Global Graph (2007)

http://dig.csail.mit.edu/breadcrumbs/node/215

and something we tried to do as well: http://yansh.github.io/articles/moana/

Yeah. I wish we had something more decentralised, like e-mail, but for social networks. Like, a standardized social network protocol that most people use (popularity is important here). There is nothing special about what social networks like Facebook do, the biggest challenge is probably serving so many requests and storing lots of data with high availability.

Even email is no longer really decentralized. Google and a few other email silos now make up the bulk.

This is why I use Fastmail, and encrypt my Google Drive data. It isn't any more secure as far as I can tell, but I'm tired of 2-3 companies holding all my easily accessible data in one place. Especially when they are openly using that data for advertising.

These days I'll always take a paid service over a "free" one where I'm the product.

But that requires scrutiny as well. Just because you're paying does not mean that they are not profiting off of your data.

I trust google reasonably well when they say they do not scan G Suite (paid) products for advertising. https://support.google.com/work/answer/6056650?hl=en

After Snowden leakage, I don't think you should take their word for it.

This is absolutely true. I guess I'm avoiding people who explicitly use my data for profit, in favour of those whose profit model doesn't require it. The trust is that Fastmail et al. are successful enough, and honest enough to not use it.

Ultimately I'd love to host everything myself. But, at present, I don't have the technical knowledge to trust myself to secure everything.

On top of that, on an even more emotional level, the principle of myself being the product doesn't sit right with me.

You still can have all your own data though in a useable form, all you need is IMAP. So it's better than social networks albeit not as decentralised as we'd like

At least it's still mainly interoperable with 3rd party solutions through open standards. I host my own email and as long as you adhere to the guidelines (use DKIM, make sure your IP range isn't part of some blacklist, triple check your postfix config etc...) you don't feel like a 2nd class netitizen.

Not quite. You still (in my experience) need to go through obscure web-forms to avoid having gmail/hotmail black list your domain. And even then gmail is way too aggressive with its spam labeling. I've seen replies to an email thread be marked as spam by gmail when the sender is in the recipient's address book.

Gmail considers anything beyond big corporate email providers as second class participants - and so does the other big providers.

Maybe I've been lucky but implementing DKIM, DMARC and SSL SMTP support seems to have been enough to make gmail happy (I even get the small "lock" icon in gmail inboxes as a bonus).

I've used one of the many blacklist check websites to test my domain and only ended up with two false positives which I've rectified through their web forms.

It was mostly smooth sailing from there.

Note that this is using a dedicated server as a host, a while ago I tried the same thing from my home ADSL (with a static IP) and that was a pain since those "home" addresses are generally blacklisted by default since people tend to assume that emails from those addresses are sent by compromised computers.

Nowadays I know many ISPs even block port 25 by default so that's not even an option anymore.

Sorry, but no. That is entirely false. I run a mail provider for companies, and as long as you adhere to standards and do not emit spam, you don't have to do anything like filling out obscure web forms. You may, on occasion, hit a spam trap due to a compromised account or similar, but de-listing on good RBLs is time-based and automatic. Be vigilant, take care of security faults in your network and, when solved, take the affected IPs out of service for a while.

This may not apply if you are a low volume sender.

I used to think so too. But sometimes, at seemingly random email ends up in outlook/Gmail spam boxes. My solution has been to not really care - I don't email that many outlook users from my private email. Let Ms/Google fix their end, after I've done my bit (dkim/ssl - I'm not a fan of dmarc).

But if I were running email for any new organisation I wouldn't have that luxury ofcourse - I'd have to pay for a Google/outlook or other major provider account - or spend a lot of time "reversing" the reputation of the ip/ip block my dedicated servers had - and as none of the major providers provide a simple check like rbl etc - it's a frustrating guessing game where your test-mails to your own Gmail(etc) account go through, but mail to new senders is silently dumped to the spam folder.

I don't mind greylisting, mail bouncing or blacklisting. It's the magic "intelligent" filtering which is annoying - because the only indication of error is that you don't get a reply - because your recipient haven't seen your mail.

In a way it's worse that it's rare - you never know if you've fixed the issue, or some arcane combination of your sending configuration and email content (an email not in English?! Probably spam!) will mean a new recipient will never see your mails.

I'll echo Simias. If you behave like a first-class citizen and follow all the current standards and protocols, then you'll be accepted as one. I've been running private mail services for decades, and by keeping pace with changing expectations I've never had a problem with deliverability.

That might be mostly true but it's not always completely true. I'll raise your anecdote with my own:

We've been running on the same domain+IP for over a decade. Hotmail block us when we're whitelisted by the original sender and also responding to an email and also have corresponded before. We've never sent even a multi-person email never mind spam, always as a response to an enquiry, very low volume.

We're blocked apparently because Hotmail has an intermediate level warning on a sender domain on another IP address managed by the same ISP; we were un-blocked before but they now say they won't allow our email through - we have SPF but can't work DKIM with our shared status. The only silver lining is they actually responded inside a fortnight to tell us that they're going to continue blocking our [at most] once a week emails to our customers who happen to have hotmail/live/outlook addresses ...

To work around I made a live.com address to send through for customer who use MS email. ~4 days later they started blocking it from remote access ... something like not enough direct website logins for the amount of IMAP lookups (a single client checking every hour or so during work hours). Few log-ins seemed to fix it.

We can pay to have the spam service MS use whitelist us though, apparently.

By sharing hosting with someone with a bad reputation, and then deliberately working around their controls, I have no surprise whatsoever that they drop your email. That is not behaving like a first-class citizen, that is behaving like an underhanded reprobate. Sorry to say that your anecdote is an illustration of "how not to do it".

You have to accept and work with changing expectations.

If you are unable to relocate to more reputable hosting, and your messages are legitimate, solicited, authorised, well-formed transactional email (no promotional mail shots), I suggest you switch to delivering through a squeaky-clean provider like Postmark.

Using there own service within their terms and conditions is being "an underhand reprobate".

They could just obey their own customers white-listing, or over-ride blacklisting for domains that act appropriately.

So you really think that having a domain of 10+ years standing that doesn't send unsolicited emails at all is "not behaving like a first class citizen"?

If someone in your ISPs /26 sends spam from a different IP and using a different domain, that totally means you're a massive spammer who's ruining the internet??

I could of course change ISP but which of the 15+ yo ISPs will have a spammer somewhere in their IP block, no way to tell.

So the only option left is pay to be able to send email to MS addresses. And you're fine with that? It's all our fault for replying to emails?

I think you have numerous misconceptions here. You can't pay directly for status. Longevity without change doesn't entitle you to anything. Ten years is not great longevity; try doubling that. Microsoft doesn't not believe you are a "massive spammer"; they are just playing the percentages based on your choice of associations and your choice to abuse their service. Finally, you seem to think you have a contractual relationship with Microsoft that obliges them to accept your email on the basis of "Terms and Conditions" which, if I'm not much mistaken from what you've described, don't apply to your relationship with them.

Your expectations don't match reality and you seem to prefer complaining to change.

I have basically no sympathy at all, I'm afraid. It is your fault, not for "replying to emails" but by failure to update infrastructure as the context changes. This is your problem to fix, and it appears you've made poor choices in that direction based on a misattributed sense of entitlement.

Perhaps the key here is "for decades" and "reputable ip". I do run my email from a dedicated box - but I host with hetzner. Their ip blocks might suffer from "bad neighbour" effect.

But how is such an effect not an issue? It makes it hard to "elevate" a box at home to a proper mail server, and makes it hard for new generations to host their own mail (until an ip6-only world (and probably also then) all new users will need to recycle old ips).

As for dmarc, I'm still not entirely convinced it's a great solution to smtp/spam issues. Dkim is already annoying enough if you occasionally want to send mail through/from a different smtp - like Gmail if you don't run your own webmail.

I've heard of diaspora being described like email.

That was essentially the promise of RDF, OWL, The Semantic Web, Web2.0 and so on. After swimming in that stuff for 3 years in the early 00's, I could tell back then that it was mostly a boondoggle, and even though those technologies have found their niche, the project failed for the same reason that the original WWW was successful.

It was like 'AI' in the 80's, but with XML.

The feeling I got was that most of those were just too general to be suited for any specific purpose. You could, in principle, build a piece of social networking software out of them, but there was never a product that you could just deploy to your web server and use.

diaspora probably was the most popular failed project which tried to achieve this. It did not really fail per se, but it's pretty clear by now that it will never reach impactful adoption.

I was rooting for those guys and I'm still following their releases, but I can't see it ever being accepted by the masses.

As a matter of fact, this is what Sir Berners Lee is working on right now at MIT.

https://solid.mit.edu/

The issue is not storing the data, just like "email" doesn't store data. It's just a protocol and a data format. This makes it interoperable across whichever industry player wants to spring up and compete for your service.

Ex:

Don't like gmail? Go to ProtonMail. And you don't lose the ability to interact with people who do use Gmail.

Don't like Facebook? Go to Ello. But now you've lost your entire network.

The decentralized web will be built on data formats and protocols that allow you to take your data with you and force companies to compete with the quality of their service, not the size of their network.

So does one export their existing network information to these open formats? I don't think the companies who currently hold that information will part with it willingly.

If it catches on, I assume a bunch of services will pop up offering to export you data. All they'll need is your password.

But how would you verify that YourFacebookFriend is YourFacebookFriend on some other service?

This is a solved problem, but not one that would catch on. See: PGP

If my friend can sign a message using the same key on both services - and I already trust that key is them - I can be almost certain it is them. Regardless of the service.

This is one of the first use cases for keybase.io - verifying identities across a few main social sites. (E: For clarification, Keybase makes it easier to find these proofs. It isn't necessary as part of the proofs.)

For example, you can find me on Reddit or Twitter and know it is me. You can also see my website URL and know that I own it - and a bitcoin address where you know I will receive the money. This is because I've verified that I own these accounts using PGP - and I've done the same for HN in my user profile. See: https://keybase.io/nadya

>This is a solved problem, but not one that would catch on. See: PGP

It's only 'solved' if you don't care about usability. PGP quickly becomes a nightmare when you consider the day-to-day things 'normal' computer users will go through (private key stolen, password to private key forgotten, phishing attacks to sign replacement keys for friends, etc).

They already do allow you to export your data, right?

Yes, some of it, but not your actual network information (friends, contacts), at least not in a way that is meaningful outside of those services.

That's Sir Tim(othy). The "Sir" attaches to the given name, not the surname.

I've never quite figured out how sirs work. I know Knights were addressed with sirs and the Queen had something to do with it, but how does it work in a multicultural UK?

Not sure I'm following you -- yes, the Queen knighted him. How does the multicultural part come in? (Women receive the same award but it's called a Damehood and their prefix is Dame.)

If Kazuo Ishiguro is knighted, should he be addressed as Sir Kazuo or Sir Ishiguro? The point being that people of Japanese descent might place their first names after their last names.

That's a language issue, not a culture issue. Normally, one uses the name ordering appropriate to the language one is speaking at the moment. Therefore, the question is equivalent to "What is the Japanese translation of 'Sir'"

This question doesn't seem very relevant to knighthoods. Whatever would be appropriate to call someone as their given name, rather than family name, would go after "Sir". It's not very complicated.

Sir Kazuo. With some exceptions - Japanese people "normalize" their names to English name order of Given Name, Family Name when writing or speaking in English. Some choose not to, usually out of national pride, but it is most common to normalize the name to the target language.

[0] https://en.wikipedia.org/wiki/Japanese_name#Japanese_names_i...

An example: Sir Ong-Siang.

https://en.wikipedia.org/wiki/Song_Ong_Siang

A sidebar issue: all these rules apply to knighthoods from the UK crown only. And under these rules non-Brits may be knighted but do not use the title 'sir'. Thus Bill Gates KBE is not known as Sir Bill.

(edit: Today I learned that Kazuo Ishiguro is British. The statement below is completely wrong. I've managed to get his nationality wrong while enjoying his books for 20 years!)

So in the specific case of Kazuo Ishiguro, the sir does not apply. For Brits with family-first names, see the other answers.

Kazuo Ishiguro is British. He's not called "Sir" because he has an OBE. He needs a GBE or DBE for the "Sir".

https://en.wikipedia.org/wiki/Order_of_the_British_Empire

And were he not British, he would not be entitled as such - Gates and Geldof iirc are such examples; sometimes (incorrectly) titled 'Sir' in press.

It's fortunate, then, that it's Sir GivenName, not Sir FirstName.

Knighthood is a life award. You are given the rank of Knight Commander of an appropriate order or made a Knight Bachelor (no particular order; lower precedence when it comes to deciding who walks in front of whom). The knighthood dies with the knight. Again, the "Sir" attaches to the given name ("first" name, even if that comes last or somewhere in the middle, or is a compound name in itself). In general, the "Sir" is only used for address or casual reference; you'd use the post-nominal letters in a more formal written setting.

Debretts is probably the place to find this out.

Here's the rules for putting the letters after names: https://www.debretts.com/expertise/forms-of-address/letters-...

The Joint Forms of address gives you some clues about the order of words before the name: https://www.debretts.com/expertise/forms-of-address/joint-fo...

Note that most people don't know about this stuff, and don't really care about it.

I've never understood why people (particularly outside UK) care about some "sirs" at all. "Mr. Tim Berners-Lee" sounds just fine to me.

Why use a title at all if you're going to use the wrong one?

He's still a mister. It's whether you buy in to "this is a male adult" vs. "I must respect this person's title because HM The Queen bestowed it on them".

Mr Berners-Lee surely won't feel offended because the person addressing him isn't a [UK] royalist?

Calling him Berners-Lee sounds like you would rather not use titles at all, which seems like a defensible position. Calling him Mr Berners-Lee sounds like you're going out of your way to deliberately use the wrong title, which just seems rude.

All adult males in the UK that you don't know the first name of but do know the surname of are called "Mr $Surname" as a social norm. Adjusting that designation because the Queen says so is very common too, but maintaining the status quo is just being apathetic towards the Queens edicts.

Most of her subjects do not complain. Some of them think we should have done away with such notions of reigning over other people already.

FWIW outside celebrity circles it is often considered quite gauche to insist on such pompous titles.

If you insist I'd be happy with calling him Professor Berners-Lee (even assuming he doesn't still have an active official professorship). If you want to talk about proper titles then is Sir Berners-Lee also rude when it should, by the right of HM EIIR be "Professor Sir Berners-Lee" [when spoken]?

If you're worried about recognition I think Mr Berners-Lee is not lacking in that department and was already party to the Queens inner circle as a holder of the Order of Merit which appears to be a far greater royal honour than a "mere" KBE. He has my utmost respect, if you doubted it.

That's not the case though, just as, more obviously, 'Mrs' isn't limited in meaning to 'this is a female adult'.

Increasingly and to the point of arguable totality, such titles are bestowed by popularity, committee, and HM Government. Damned shame, since it would mean more what it should were it not given to celebrity riff-raff for 'services to sport'.

I digress. You say it's fine as a non-'UK royalist' to use an improper title; I say I bet most of the world doesn't use even Mr, and I'd do my utmost to pay proper respect to local custom and any honours.

Does local custom work both ways? I think most Americans would consider "Mister" perfectly courteous.

I'm not sure what you mean by working both ways, I'm saying as the speaker one respects the subject's title, even if not 'recognised' natively.

For example, I'm not Catholic, but I would of course refer to Pope Francis; not 'Mr Francis', or anything involving his birth (as opposed to regnal) name.

I think the norm is the current surroundings, not wherever the subject is from. Dr. Smith is Smith-sensei in Japan. The New York Times is known for its use of courtesy titles, and it calls knights "Mr.".

Pope is a high office, and Francis is his chosen name. It would be rude of you to call him by his birth name; it would be rude of him to insist you call him "Your Holiness", although you probably should if you're visiting the Vatican.

Knighthood isn't a royalist issue. Australia has knighthoods bestowed by the prime minister. Unfortunately our PM decided to knight Prince Charles...

A knight of Her Britannic Majesty Queen Elizabeth the Second's Order of the British Empire. Yeah, nothing to do with royalty /s.

The could become a republic and everyone would be a knight of the president instead.

Do you similarly never refer to people as Doctor, Reverend, Minister, or Coach so-and-so? It's part of his name. If Sir Tim didn't want to be knighted, I feel confident he wouldn't have been. You'll note that his wikipedia page calls him Sir Tim, and I feel confident he could change it if he so chose.

You are clearly free to call him whatever you want, but I would advise you to be honest enough to realize that you're projecting your feelings about royalists onto him, and not calling him what he has chosen to be called.

In practice you're right, i call people what they indicate their designation to be. Where I work we've had a person call themselves something like "Grand Wizard-Admiral of the Terran System" - some people like a pompous name.

If a person is an MD, professor, and such then that's totally acceptable. If the designation is "the Queen says you have to call me sir now" I don't really see what business it is of the Queen's or that it's more respectful to recognise a person because "the Queen smiled at them" than recognising their actual work/effort as fellowship of a learned society does.

It's an anachronous system of nepotism; bleurgh. What's not to like./s

Because it's a nice way to recognize some distinguised people, that's why people care about sirs.

Sort of off topic, but this is exactly what I want in electronic medical records. Being able to take my data and give it to whichever healthcare provider I go to.

But half the point of a social network isn't the ability to communicate, but rather the history you've built there. Likewise, half the value of email isn't the ability to send a message to anyone, but the history of emails you have. Sure email isn't a data storage method, but in reality? Yeah, email is a data storage method. That's why Gmail gives me 25GB for free. Not because that's the size of my inbox, but because that's the size of every email I've ever gotten.

If I move from Gmail to Outlook, I can still email my friends but I can't go back and search our old conversations unless I export everything from Gmail to Outlook. If I switch from Facebook to Instagram, I can still message my friends but I can't see all of the pictures I posted on Facebook unless I export the pictures from Facebook to Instagram, but even then there are a lot of features of Facebook that I can't export data from because they just don't exist on Instagram.

History creates lock-in, and history is really hard to move around.

I don't see what would make basic facebook history hard to move around. Facebook of course already offers a "Download a copy of your Facebook data" function. Future-facebook could import it, or the most fundamental/basic parts of it, probably -- the problem is not the data itself, but the connections in it to other users, it's not that valuable without it. I suspect there isn't much interest from users in things-trying-to-be-facebook of importing the data.

I think it's the ability to communicate that's harder to move around, and why there's such facebook lockin. Your social network and ability to communicate with them in a public/group fashion.

>I suspect there isn't much interest from users in things-trying-to-be-facebook of importing the data.

And the legalities of that.

FB did a major legal smackdown on someone trying to do that.

Do you have a link to the history here? Would love to hear about it.

Curious for a link to that. Facebook lets you download your own data. But if you upload that data somewhere else, they're going to sue someone?

> history is really hard to move around

Is it? I think there's a way to get around that, if the protocol defines standards for verbs.

Of course, it's difficult and can grow wild, yet very much possible.

> History creates lock-in, and history is really hard to move around.

It doesn't need to be. For example, I moved my email between IMAP servers multiple times simply by dragging & dropping a swath of messages in Outlook. That Outlook (of all programs!) is the only tool that I know of that makes this easy is not a problem with the decentralized nature of email but simply with mediocre client implementations.

> That Outlook (of all programs!) is the only tool that I know of that makes this easy

Not sure what you mean.

It works as easy with Mutt. Open IMAP folder, select all, save to [other IMAP folder or local folder, as you wish].

Works reliable even with very large folders, you just have to wait a bit longer until it's finished.

Practically every email program can move email easily. It is somewhat trivial.

Recently we moved all our company mail with IMAPsync. Automated commandline tool. Awesome.

Thunderbird can also do this.

History is crazy easy to move; any difficulty we have moving history is intentionally caused by the companies with a vested interest in keeping you from moving that history.

History should be decentralized too, then.

Who remembers RSS feeds? I miss those days. We should have some superior spin on that but for social networks.

Imagine Facebook, but your history/data were your own property. A friend would find you on Facebook, or Ello, or AnythingAnyWhereBook ... add your feed, and done. Then it is up to the Social Networks to make a superior product for you to enjoy your friends data and interaction.

Right now it's about the monopolisation of our data.

Zuckerberg. If you're reading. Your current model is bullshit.

What do you miss about RSS feeds? They're still there... It's very rare that I come across a blog that doesn't have one, even now...

You're right, of course. It's possible I'm still pining for my perfectly set-up Google Reader account which I cannot quite replicate. That still hurts! RSS felt ubiquitous (to me) back then.

Anyway, to the point, try using RSS with facebook in the above context.

I hate being locked in - hence I gave up on FB. It's trash to me until it solves this issue.

> Anyway, to the point, try using RSS with facebook in the above context.

For me they're different things. RSS is for things that I want available until I get around to reading them or until I manually mark them as read, this is precisely what I don't want out of facebook.

I suspect parent was talking about the distributed nature of it. Subscribe to whoever you want, allow anyone to subscribe to you.

My RSSing calmed down, it's just the XKCD feed now. But it still works perfectly.

Intruigingly given how many RSS feeds are out there from WordPress blogs, about the only way I found to get news on WordPress releases is via their RSS feed.

>Who remembers RSS feeds? I miss those days. We should have some superior spin on that but for social networks.

I think a big missed opportunity for Google on social was the failure to cultivate Google Reader into an open social network (it already had social features), and build each aspect of it on an RSS/OPML-like structure.

Gmail supports the imap and pop3 protocols, which allow all history to be trivially imported somewhere else.

That's the importance of open protocols.

The goal is solid, but I wonder how it might prevail. The flexibility to change at any time doesn't seem a priority for most people. Most people search for a satisfactory solution and stay with it until it's no longer bearable, or a competitor many times better appears. The same for most things in one's lifestyle.

> There is nothing special about what social networks like Facebook do, the biggest challenge is probably serving so many requests and storing lots of data with high availability.

The difficulty with decentralising is that these are the biggest challenge to centralised systems like Facebook; they're problems which are very nice to have ("oh no, we have a billion users!").

The biggest challenge for decentralised systems is getting momentum behind an agreed-upon protocol.

Email was born when there were few people who needed to agree, and their major concern was being able to send/receive messages. As the Internet expanded, the momentum behind email ensured that it could ride that wave as well.

These days the Internet, and hence the number of people who need to agree on the protocols, is huge. Think about, for example, HTTP2; how many organisations and individuals were involved in its definition and publication? How many more have been involved in implementing it for various browsers, servers, crawlers, libraries, etc? How many more have been installing updates, editing config files, etc. to use it?

Another change is the concerns of the stakeholders; if all we cared about were sending/receiving messages, then we could use something like telnet, HTTP, email, etc. Instead, the concerns are more fuzzy (e.g. "presence", "likes", etc.) and more political (silos vs federation vs p2p, expectations of privacy, encryption, archiving, searchability, etc.). Reaching agreement on such things is very difficult, as everyone wants different things.

The results either:

- Have buy-in from too-few people (e.g. diaspora, pump.io, gnusocial)

- Are so general as to be inappropriate for most particular tasks (e.g. RDF; an interesting example considering that RSS is a stripped-down subset of RDF, and is/was widely adopted; also compare to ATOM, which took that subset, threw away the ties to RDF, and became more interoperable!)

- Only offer tiny featuresets (e.g. the various microformats)

"email" isn't even a single protocol. Even just talking about SMTP you'd run into lots of "common edge cases" where different servers respond to errors differently, a lack of TLS uniformity, variations in auth, variations in host detection, etc. And that's before you take collection into account (POP3, IMAP, IMAP+, ActiveSync, etc).

Email is a mess of kludges and frankly could benefit from a ground up re-implementation. In fact many have tried. But like with creating a decentralised social network, it doesn't matter if you have an arguably better implementation if you don't generate the snowball effect attracting users to your "better" platform.

> an interesting example considering that RSS is a stripped-down subset of RDF, and is/was widely adopted

RSS 0.9 was RDF, as was 1.0. 0.91–0.92 (and the 0.93/0.94 drafts) and 2.0 had nothing to do with RDF.

>These days the Internet, and hence the number of people who need to agree on the protocols, is huge.

These days however it's not the people who decide, it's organizations.

>Think about, for example, HTTP2; how many organisations and individuals were involved in its definition and publication?

Many may be involved, but very few can decide whether to use it or kill it. Namely, Google and Facebook would be enough. Descending down to individuals, it's about 3-5 from Google and one from Facebook.

I expect getting that many requests would be a bigger challenge than serving them.

Have you heard about Mastodon?

https://github.com/tootsuite/mastodon

It's like a decentralized version of Twitter, with federalization. Might need to find a instance that's not busy with signups though.

Mastodon is (mainly?) a new implementation of GNU Social. It federates with existing GNU Social servers (which, as StatusNet go back to like 2007 or so). I hear it's a good quality implementation, though.

That's the nice thing they did in the 90s and before. They created standards like FTP, HTTP, Gopher and whatever and people could build upon these standards. Now we are back to everything being proprietary,

Help build Urbit.

What about GNU Social? https://en.wikipedia.org/wiki/GNU_social

I run a GNU Social site. It works well. It's a bit slow and the themes are out-of-date, but if it had more of a following I'm sure we could turn it around.

This this this.

I really like the idea that Eben Moglen seems to pitch in every of his recent talks when talking about FreedomBox:

A decentralized network of cheap "plug-computers" that run Facebook-like functionality on the hardware of the users. And yet, while FreedomBox is a great project and does many things right, it omits the one crucial aspect of this: the decentralized social network.

Let's hope Solid or something similar takes over the social networking world eventually and we can get rid of all the megacorporations hoarding our data.

This is why I'm a fan of the indie web (https://indieweb.org/).

I like indieweb but I'm not on board with their 'design first' slogan. I think to be most effective such a thing should be 'content first'.

Content, in terms of quality; and subsequently curation/censorship?

Otherwise, content falls into a specific set of mediums (text, images, video, sound), all of which are assumed on a modern platform and aren't worth reconsidering. Design and presentation, however, has infinite possibility.

> Design and presentation, however, has infinite possibility.

Most of the possibilities get in the way of the content.

How do you get curation and censorship out of a bunch of people publishing things separately on their own websites?

That's my point. Why should content be the first priority if it's qualifications are not open to discussion?

I'm sorry if this has to do with my sleepiness, but from that page I don't understand at all what this is. A kind of social CMS? A bloggers community? The page is rich in word count but poor in being informative. I'd be very grateful if somebody could give a little explanation of what this is, because, it seems to be an interesting thing.

IndieWeb is a social movement rather than any one particular technology or project. It's a group of people that believe in publishing as the web was originally intended: a diverse network of participating individuals, not corporations, bound by open standards. There are a number of projects under the general Indieweb banner or just sharing the IndieWeb principles.

The other reply gives a good answer. I'll suggest reading over the about page (https://indieweb.org/IndieWeb:About), I think it's more informative.

Indie web is about people owning their data and having control over it. It's about being able to build what you want for yourself. It's a direction, a set of ideas and principles that some people enjoy.

then you'll like https://yunohost.org

Tim Berners-Lee seems to be much more idealistic about how the web should be in that regard, because he knows it's the right thing to do.

And yet he decided to be "pragmatic" on web DRM, instead of taking the same idealistic approach because it was the right thing to do.

http://gruik.io/view/1238

I had a similar realisation.

What's the product in a social network?

It's you.

:) That goes for television as well.

Ad-supported TV. Other models exist: BBC, PBS, HBO, etc.

This is true, unless you pay for the social network, or it's p2p.

It's not purposefully evil or inherently wrong, just an inevitable consequence of the business models involved in free-to-use social networks.

I'm trying to explore whether a paid social network could be a workable model with https://postbelt.com

Please check it out :-)

Appreciate the realisation but this really is repeated ad nauseum and is not an original thought in the slightest

a thought/concept doesn't need to be original to a wider group and can still be profound for the individual. see "just about any subject".

thats true, but they also provide lots of useful features to users.

It may be good, after all we are talking about private information. Some degree of control and exposition are welcome.

The problem is that such control is not imposed by you, but by the network owner.

Reply on news.ycombinator.com