Diaspora: One Month In | Hacker News Reader

Diaspora: One Month In

joindiaspora.com

212 pointsiwani86 comments

Glad to see an update from them—sounds like they're doing work. They've got a lot more to do in the months ahead, and I'm looking forward to following along.

As an aside, it's interesting to see how similar their message passing protocol is to one I'm beginning to plan out, though for a completely different application. Kinda like a "push to origin; fetch from origin" (from git).

EDIT: And hmm, every time Diaspora comes up, people start to say something about a "hype cycle" or something. I don't really know what that is, but it's rather obnoxious.

How about we don't take every mention of Diaspora as an invitation to count the ways they're sure to fail; because of the hypecycle or motorhyper or the gnarly-toothed funderrazer…etc.

To me, this is an interesting update about a curious project. Keep it going!

How can you tell it's obnoxious if you don't know what it means? Easy to find out, though - http://en.wikipedia.org/wiki/Hype_cycle You might find it inaccurate or otherwise useless but it hardly seems obnoxious.

As to assessing Diaspora's chances of success - it's a perfectly reasonable thing to do, especially in a forum of technology-, product- and business-focused people. It's Hacker News, not Cheerleader News, after all. Just because a reasoned analysis might be negative doesn't mean it's cynical.

So, if I'm reading this right, the servers will be totally distributed?

I don't care about a Facebook replacement so much as a totally distributed file-sharing system where trust and recommendations are based on my real-life friends. It seems like that could happen with this. Also, I think file-sharing is a great hook for the everyday Joe. And it's somewhere Facebook can't easily go.

Couldn't agree more. The system they cook up will be one of many websocket wielding filesharing services. What will they do for browsers without websockets, flash supports sockets (cofounder Tyler hacked out a quick fallback).

Maybe it's better to just not suppor older browsers?

My guess is that people who are interested in Diaspora, at least at first, are current on their web browsers. I think there's a trend among web services (esp. Google) to prod even everyday people away from older browsers, especially IE6.

Then again, if they want widespread adoption, they can't just support HTML5-compliant browsers. Hmm...

Yes, a private dropbox/pirate bay among your trusted network of friends is 100% win. Easy to use and virtually invincible to RIAA claims. Heck, include a VLC browser plugin to make it all seemless.

This was already done by Facebook and I assumed removed for liability reasons. See: http://en.wikipedia.org/wiki/Wirehog

Though I do agree that the distributed nature provided you have encryption will make it very hard to catch people sharing files. Though if you stream the videos/images/books from your site to read/watch them you can get caught at that point(If ISPs complied with CopyRight enforcement strategies.)

I actually like this refreshing new approach to writing open source software: keep it closed for several months & code like madmen. If they can get UX people involved in the closed stage we might even see an open source product with an elegant ui for once!

Totally agree.

I think the "release often" philosophy is a good one, but too often and you just open yourself up to "this doesn't work, that's wrong, this is broken etc"

I think knuckle down, code a bunch of good stuff then show it off. Of course you'll still get the complaints, but it'll be a lot clearer to a lot more people how it works and the potential it has.

All that said, all they've shown us here is a very simple demo, we have no understanding of the underlying message passing protocol and its robustness. It does look promising, but looks of course...

Sure. But the mantra is "Release early, release often".

And I guess early means that you don't need to wait until your code-base is feature complete or at 1.0 to release the code.

By the same token I think you're allowed to do what the Diaspora guys are doing and lay the foundations first, set the tone of the project and the general direction and style.

I hope these guys do well but if it ain't them it'll be someone else cuz peer to peer federated social networking is in the near future of social networking and I'd much much much prefer if it was open source than not.

Yea, this is pretty much the comment I wish I'd posted.

Sorry to say but "this doesn't work, that's wrong, this is broken etc" is going to be said regardless of when you release and regardless of how many users you have.

The number one rule of software engineering. Your code sucks.

Yes! It shouldn't be too hard for them to find some good UX people in the middle of SF and still flush with Kickstarter funds; they'll just need to know what to look for.

Of course, there's still the issue of actually countering Facebook with a more organic social structure. If you will excuse my carting out of this fine dead horse once again: http://interuserface.net/2010/05/how-diaspora-can-succeed-mo...

Indeed, they've already found good UI people, in the very building they work in: they're being hosted for the summer by Pivotal Labs.

It's disappointing to me to see them choose regular HTTP for communication. I think XMPP PubSub or just regular XMPP messages would work better, and you could use the existing federated jabber networks. With this, they are probably going to run into NAT traversal issues, etc.

I actually spoke to these guys directly about this at a party. Only for a few minutes cause I am sure everyone and their mother wants to bike shed with them. But yeah, xmpp might have been the better choice. Google certainly thinks so.

Protocol is all they have to get really right out of the gate.

If it can't be hosted on the crappiest major shared host provider you could name, it isn't a fit for their goals.

Which means it's a total mystery why they're using WebSocket, but yeah.

Despite these guys' good intentions, I'd actually place my bet on an unfunded startup in this space over these guys who have been put through the hype cycle before they were ready.

Yes but remember how important the "seed users" are, which is why Facebook started where it did. If they truly deliver on the non-corporate, privacy aspects and keep a slim interface then many "alpha geeks" will swap over. Do you really care about the 100 or 200 friends you have on Facebook or the core 10-40? Or if all your tech buddies and most of your clued-in friends are on Diaspora and posting content there they don't post on Facebook, which account will you log into first? What will your Facebook become then other than a viewer, rather than a place that you put content? Would you have much trouble shutting down your Facebook at that point?

I don't see any unfunded startup being able to do anything new in this space, Facebook offered interface over Myspace. Diaspora offers control over Facebook - and from this demo they seem to be dispelling the vapour accusations, although we will have to see how it progresses. Google and Twitter currently are offering things and will be offering more in this space also. What real other avenues or general areas do you see a newcomer playing to? Games? Dating? Work collaboration? An even better interface? I don't see many areas for traction. Yes there is probably something I haven't thought of, but I don't think you can place your "bet" on a hypothetical vs these guys with an interesting demo and plenty of money to keep developing, who have hype and who are attacking the most vulnerable privacy/information vector in this market without some reasoning to back it up.

Edit/addendum: I also think these guys rubbed a lot of HN people the wrong way - although I am not accusing you of this necessarily - because they got decent crowd-sourced funding in a seemingly haphazard way. Yes they are young, but if you go back to their pitch it was very slick, and as I said above, this tech demo goes a fair way to dispel the concerns about their youth/inexperience.

Personally, I use facebook to communicate with friends and family - neither of which are universally geeks or techy. So I'd still use facebook.

Just because a few privacy freakouts and geeks flock to this I don't see any reason anyone else will yet.

You sound like the slow-adopters of facebook. In the world we live in with security becoming more of a general concern, they could be on the front-edge of the revolution they think they are on. Plus, if they deliver on the UX/UI, people will use it.

"In the world we live in with security becoming more of a general concern"

Which world is that? The one where people post minute details of their life on twitter? The one where they send naked pics of themselves around, expose themselves on chatroulette?

It's going the other way - people are exposing more and more of themselves making more of their life public than ever before.

Some geeks are always concerned about privacy/security. How many "normal people" use PGP to sign their emails?

Today's concerns for security are not concerns for keeping our information private. Rather, they are concerns that we will maintain control over our private information, to be able to share it at our own discretion. Facebook removes much of this control from the user.

no evidence to show that facebook users see it that way, or care.

Also FWIW, I think the name "diaspora" is terrible I cannot see it working at all.

The name can be changed, what I want is the technology.

Show me the code and let me develop on it. Let me change the interface, add features, and do analysis on my social graph. Depending on the license they release the code under, there could potentially be other services that use the technology and have a nicer name than Diaspora.

Facebook users might even take notice if Diaspora has a "dislike" button and access to their Facebook data via Diaspora.

The issue is, you'll get complete fragmentation and confusion. It won't get widespread adoption. It's a fun thing for geeks to play with, but I don't think there is any way in hell this can succeed with the current setup.

Would you see a way to succeed for Google in 1999? Only geeks used it in the beginning

That's slightly different. Other search engines were poor. They focused on the wrong things, and didn't actually give good search results in the way users wanted.

Facebook actually solves most peoples needs very well for a social network / casual gaming platform.

Any competitor will have to work for a few years just to get the features that people will assume it has. Then comes the massive job of moving people from facebook.

Consider ebay. Ebay actually really sucks. It's terrible. Yet everyone still uses it for auctions. Why aren't there any auction startups? Because it's impossible to get people to move.

It doesn't have to get widespread adoption. It just has to be suitable for me to use. I don't use facebook hardly at all for posting content. I almost never log into it.

If this can pull from facebook eventually and perhaps even push to it then I'll be happy.

It's open source and open protocols. Because of that it's definition of success is not the same as a startups. Even if the project fails miserably but in the process causes Facebook to change a little then that will be a success IMO.

What is the point of a social network if only you use it? ;)

It'll be fragmentation, yes, but hopefully it will be smart fragmentation, not chaos. The purpose of a distributed social network is for there not to be a single centralized server. They're not building a Facebook clone, they're building an open social network.

The way people use social networks right now is fragmented. People have accounts on Facebook, Twitter, Linkedin, etc. and they have to continually check those. What Diaspora potentially offers is a way to unify that communication, if you so choose.

Diaspora doesn't necessarily have to move people over from Facebook to their service completely. People may not need to leave any service at all. All Diaspora needs to do, is provide the same access to their social network, but with a little bit more control and extensibility.

There are currently 3,208,579 people on Facebook who want a dislike button. If Diaspora can provide that in the form of a dislike plugin, there's a lot of potential users. If Diaspora does not force any changes to their interface, but instead allows users to theme and customize them, then there are a few million more potential users.

As far as how long it will take before Diaspora has all of the features that Facebook has, I personally believe that it will take less time. The technology is more established than it was when Facebook first started, and they were still trying to figure out what would work, what wouldn't, and how they can make money.

In Diaspora's instance however, there are quite a few case studies (Facebook, Twitter, FourSquare) that they can look towards for inspiration. People are in general a bit more open, and engaged on social networks than they were when Facebook was founded (Facebook after all was first limited to college students only). Plus, there are the untold number of developers who would be willing to contribute towards the project when it is finally open sourced (just look at how many backers they got, and that was money, not code).

Phew, this comment turned into more of an essay, but I think it sums up my views, and hopes on Diaspora. Agree or disagree, thanks for reading at least :)

The so-called 'hype cycle' is the biggest contributor to social network's success, at least at the very early stage. People join social networks because other people are there, not because it guarantees more privacy, faster communications, or special features. An unfunded/unknown startup is more likely to lack traction and fade out quickly, no matter how good it is.

What I'd really like to know is: what are these guys doing with the 200 grand people pledged? Is it just paying their rent in SF or what?

If you look at the rewards the "backers" get[1], I bet they'll spend around half their money on buying and shipping all of that stuff. Possibly more if they pay someone to do fulfillment for them.

Also, the best way to waste their money is to say "well, people expect us to spend this money, let's find ways to spend it that sound useful." They shouldn't be afraid to spend it on worthwhile endeavors, but they also should be in no rush to spend it.

What do you really want them to spend it on? A few more developers might be good if they know good people, but wasting much time on hiring when they've just got a summer isn't worth it. (Btw, anyone know if they're returning to school in the fall?)

So they'll pay for their plane tickets, rent, and food. Hopefully they'll contract someone to help make the UI gorgeous unless one of them is really good at it (though it looks pretty good right now). And if they end the summer with $75k in the bank that they can use going forward, that's awesome.

[1]: right sidebar of http://www.kickstarter.com/projects/196017994/diaspora-the-p...

It'll be a fair bit less than half, probably around $50k.

Shameless but very relevant plug: http://nerdlife.net/diasporas-upfront-costs/

What do you think they should spend it on? What did you think they would use it for?

Creating something like this - or most web startups - isn't expensive, it's just time consuming.

"Creating something like this - or most web startups - isn't expensive, it's just time consuming."

Which makes it expensive. Or is your time free?

Right. but the OP said "Is it just paying their rent in SF or what?"

Which seemed to imply that would be a bad use of money - salaries.

Aside from paying their own salary - that money could be a very good investment into their infrastructure. I know their primary goal is development of the protocol, but it'll catch on a whole lot faster if it starts with an official host for people to keep their seeds on.

What I want to see: DRM for the people! Incorporate White Box Cryptography and the whole suite of tricks so that users can share certain info then unshare it. It wouldn't have to be perfect to be of tremendous benefit to many. (Girlfriend revenge sites won't be thrilled.)

What a fantastic idea. I'd never thought about the positive applications of DRM.

Can you describe how you envision this working? Why would you choose to use a client that correctly implemented the drm features? How could a free/open source client even accomplish the drm?

I'd only implement this for completely closed platforms. All bets are off for hacked and jailbroken devices. The point isn't to provide absolute security. It's to provide an audit trail and some measure of security -- enough to foil an average consumer, but not enough to stand up to concerted attack.

Why would you choose to use a client that correctly implemented the drm features?

To avoid prosecution under the DMCA. To avoid having to jailbreak your device. Because the effort of hacking isn't worth it. Current streaming DRM technologies have a good-enough track record, in part because what they protect isn't that valuable.

Big companies don't have access to perfect DRM, but it does useful things for them. Why shouldn't individuals have access to the same tools to protect their own data against such companies? In particular, big corporations are just as vulnerable to legal remedies as individuals.

What is the nature of their relationship with Pivotal? Anyone know?

It is most likely a client relationship. Pivotal is big on getting their clients into their office so they can participate in the XP process.

Probably not. They don't have enough money yet to afford Pivotal. This is probably just good PR for both sides, as well as relationship-building, so that when Diaspora gets an A round, they can start spending it at Pivotal.

Happy to see them working with Pivotal, that gives me a lot more confidence in them. They're working with top notch mentors.

I don't think this will catch on. Diaspora is too hard for people to say... ;)

I was surprised by the way the bloke in the video pronounced it actually, was thinking more dey-ah-spore-ah.

I think there is an argument for a simpler name, but at this stage it's hardly the most important factor.

If you step back from the "hype", not to directly oppose the many views present on this board, and analyze their statements, it appears they have more reasons to fail than succeed. The most worrying thing I see is their reliance on pivotal labs. They mention it in passing, but it would appear that the 4 founders of diaspora have very little technical knowledge and have outsourced the majority of their code. They are relying on someone else to do the most fundamental part of their startup and that, in and of itself, is extremely worrisome.

Every startup has more reasons to fail than to succeed. That doesn't stop a few from succeeding though.

Why is it worrisome that they outsource a part or even all of their code ? They came in to a boatload of money, maybe they're just using that to accelerate their development track.

"Every startup has more reasons to fail than to succeed."

That is true. But I'm curious -- is Diaspora a startup? I understand they're a bunch of students with a right attitude and a wrong idea, trying to give something to the community for the greater good -- have I got it wrong?

They're a startup in the sense that they're starting something new from scratch; they're a few people taking a risky voyage into the unknown in the hope of it paying off big (in this case, via social benefits).

> "Every startup has more reasons to fail than to succeed. That doesn't stop a few from succeeding though."

That's true for an average startup. But a facebook competitor??? Really? cmon. The odds of success are billions to one. It's probably similar to a startup gaining market share from eBay for auctions. The network effect is just too great.

I'm not quite sure what makes you say "the 4 founders...have very little technical knowledge and have outsourced the majority of their code."

Here's what I read,

After getting settled in ole San Fran, our first day at Pivotal Labs was June 7th. Not only has Pivotal lent us desks and monitors for the summer, they push us daily to drive development from the interface and focus on the experience, rather than providing just a tool for developers to hack on. Getting periodic help from Pivots has already been transformative on the outcome of Diaspora.

So Pivotal Labs, thanks for letting us use your stuff, eat your food, and for teaching us your agile ninja ways. We owe you one!

That doesn't sound like Pivotal is writing code for them. It sounds like Pivotal is (1) giving them (free?) space to work and (2) mentoring them regularly. The bits below what I quoted - where they talk about specific technology and features, say "we did this..." and "we are adding that..." Again, no implication of outsourcing.

My recollection was that they were NYU undergrads in CS. Why do you think they lack technical knowledge?

I find it very encouraging that Pivotal is encouraging the Diaspora folks to focus on the user experience.

Pivotal Labs doesn't work for free. If you go to Pivotal's front page, what they do is pretty much do all the coding for a startup. They take a business person's ideas and then code them up and implement them. From there they then teach the business people either how to maintain it or how to hire a dev team that can (or work with the dev team they've hired to learn how to work with the codebase).

Its just what pivotal does for a living, so my logical conclusion was that Diaspora was using Pivotal for code, and their comments on their blog were filtered to with PR speak.

You may be right, but that's an awful lot of speculation. As for "Pivotal Labs doesn't work for free," the publicity of "helping the famous kids dream big" may be more than enough payment. (That's assuming we want to be ultra-cynical, which I have mixed feelings about frankly.)

Nope, Pivotal is just providing space and mentoring.

The mention of Pivotal's support actually improved my outlook on this whole thing. It doesn't sound like they are outsourcing the code, it sounds like they are being mentored, and that Pivotal thinks this crazy thing could work.

Anyone else turn HD on and look at the URLs? http://washington.joindiaspora.com/ et al. I've been trying to create an account without luck, but they're running the app in the development environment so I've seen a lot of in depth back traces and things like that. They're using Rails 3 (beta 4), Thin and EventMachine's async features, Mongo/MongoMapper, Devise, and Warden.

Is that asterisk part of their logo? It's really distracting. I spent a good few moments looking for a footnote, until it dawned on me that it was just a visual gimmick.

So instead of hearing about their progress, I'm scanning their footer. I say change the logo, because it's first thing on every page. Thus an awesome distractor for whatever they want to tell.

What did you expect to find in a footnote for a logo/company name?

Trademarks, disclaimers, parent company references. (X is a registered trademark of Y. Y and X are not affiliated with nor endorsed by similar-sounding Υ or Х. Y is a fully owned subsidiary of Z. All rights reserved. Blah Blah.)

I think it will be really interesting to see if they do the security correctly. If my entire network wants to remain anonymous to each other then that needs to be supported. If I want to create a separate public persona - I should be able to do that too.

All communication between nodes should be really hard to intercept/fake and should be strongly encrypted, minimal logs should be stored and all content should be strongly encrypted on disk. I'm sure there are those here that could tell us what that might look like from an implementation standpoint and I'd love to hear everyone's thoughts.

People are already comfortable with this mode of communication and we can make the world just a little better place if this new platform for communication was really secure by default and worked not only for the average internet user posting pictures of their cat but also the occasional government dissident organizing a protest.

An interesting note is that Pivotal Lab's VP of Technology, Ian McFarland (@imf), was Friendster's Chief Architect. Ian is one of the nicest people in the SF startup scene, and can share the experience of Friendster's growth, pain, and subsequent downfall.

What just happened in the message passing demo?

An offscreen computer on another seed posted some messages. They then propagated from one seed to the others and then websockets automagically sent the posts through to the users sitting at their browsers, which jammed them into the page as it received them (each of the six windows being a different user viewing their own seed and its feed.)

Looks interesting. I really hope they try to change it up completely from 'the facebook model', while still including essential features - personally I'd like to see more emphasis on the virality (yes, virality!) of groups.

These guys seem pretty smart, energetic, and optimistic. Anything that breaks the status quo is good.

However, they could have used some screen capture for the demo ... It wasn't entirely clear to me what was going on there.

They're focussing on features, not benefits.

One more thing I don't really understand -- what will Diaspora offer that other similar projects, already implemented don't already have? I'm talking the likes of Appleseed, PeerBook, PeerSoN, OneSocialWeb and Peerouette...

honestly? I think it's hype and polish (then again, i've seen the other sites, and they aren't far away from good polish, so mostly hype).

edit: i'm not trying to make a comment about hype cycles -_- or how they are going to fail. i'm just Diaspora is pretty much those projects, but with the attention and money they all deserved at a time when they were (are?) most active.

congrats on the success - good to hear you're on track, and I'm excited to try logging in!

how can average people like me help out?

patience. realistic expectations. encouragement.

... so when you post a message, it's like sending 300-400 messages to up to 300-400 servers?

Uh, yeah, that'll scale.

IRC servers can have lots of clients that receive messages, and they have no problems sending thousands of messages to connected clients. Diaspora's nodes seems to be both server and client, but the mechanics are the same. When it receives a message, it's just like a IRC client, and when it sends a message it is like an IRC server.

Unless each physical server is overloaded with busy nodes, I think it will scale well. There is a limit to how many friends that subscribe to an average user.

A possible improvement could be to have "repeater" nodes for popular users with tens of thousands of readers. Instead of sending 20.000 messages, you only send messages to 2-3 repeaters for example. The subscribers would be connected to the repeater node, and not the original node. I imagine that this could be controlled by setting a privacy level on your messages. Public messages would go to the repeaters, while private messages would be sent directly to your closest friends and family

OTOH, if you have 20,000 "friends" maybe Diaspora should force you to buy a more powerful server. Social networking is one thing and social marketing is another.

What language are they using?

From the FAQ:

What language are you programming Diaspora in?

Ruby, with a bit of Rails and other Ruby frameworks thrown in.

It looks more like a Twitter replacement rather than a Facebook.

Im excited to try it out.

Made me think about Web 3.0 possible definition - websites connected

web 2.0 - ajax

web 3.0 - comet

Sorry, but they're the same.

No, they are related, but not the same, particularly on the server side.

3.0 was supposed to be semantic, atomized data.

Reply on news.ycombinator.com