Slashdot Mirror
Facebook Competitor Diaspora Revealed
jamie writes "A post has just gone up on Diaspora's blog revealing what the project actually looks like for the first time. While it's not yet ready to be released to the public, the open-source social networking project is giving the world a glimpse of what it looks like today and also releasing the project code, as promised. At first glance, this preview version of Diaspora looks sparse, but clean. Oddly enough, with its big pictures and stream, it doesn't look unlike Apple's new Ping music social network mixed with yes, Facebook."
Facebook has things pretty much on lockdown, as far as "full feature" social networking is concerned (not to mention the fact that, if wanting to be visible on a social network, most people already have a Facebook account.) I realize that at one time, MySpace had things all sewn up as well, but still...you know what I'm getting at. Anyway, like so many other things, hopefully Diaspora will bring serious competition, and help dictate the way some things are done.
If nothing else, it could at least become a social network for FOSS folks, which would be pretty cool.
Living With a Nerd
Now, all they have to do is to convince 500 million people (or whatever it is FB claims today) to move over to their service that has no whistles or bells. Umm.. 1/ Build competitor 2/ Release to world 3/ ??? 4/ Complete and utter failure.
Yawn.
Wake we up when something interesting happens.
Oh, it's written in ruby? Never mind. /starts language war
Get my friends to join it. And then my friends' friends. And my friends' friends' friends. And their grandma.
And then Diaspora would look like a serious Facebook competition.
If this really wants to be a "competitor" to facebook they are going to need a lot more than just software. Of course they need users, but they also need a central organization and a LOT of servers. Facebook is more than just a software interface, they have a massive # of globally distributed data centers that cost a ton of money. I doubt any one organization is going to put the same amount of resources behind this project. More than likely, if this amounts to anything it won't be a facebook competitor but instead a platform for much smaller communities to use. TFA even mentions this(but its not in the summary. Of course being open source it is theoretically possible then to "transfer" your profile among communities, but that remains to be seen.
Monstar L
I don't understand how a piece of unreleased software can be considered a competitor to a service that (claims) to have 500 million active users.
Are there a load of open source social networks? I wasn't aware of any (not that I've looked past the articles on /.)
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
Now, all they have to do is to convince 500 million people (or whatever it is FB claims today) to move over to their service that has no whistles or bells. Umm.. 1/ Build competitor 2/ Release to world 3/ ??? 4/ Complete and utter failure.
That is utterly correct. It's too late. Facebook has hit critical mass and Dispora is too late to the party. Facebook has pretty much crushed MySpace and every other social networking site - LinkedIn is hanging on because it has the "professional" crowd - but even then, I see a lot of folks who are using Facebook for that purpose and businesses are finding it more and more important to have a FB profile.
Being Open Source is nowhere near a good reason for folks to flock to it. So what? What features will Dispora offer that will make it compelling for folks to cast aside their social "investment" in Facebook? None that I can see.
RIP America
July 4, 1776 - September 11, 2001
I gave the developer preview code a run today, and all my hopes as to what Diaspora could be died. It took too long to produce so little that everyone's outrage at facebook's privacy has been compartmentalized into a hollywood movie on the subject, and thus rendered irrelevant.
To be a seed you are going to need a hosting provider that supports ruby on rails with a freakishly huge list of gem dependencies, that is also running the thin webserver - that's right it doesn't work on apache (parts of it worked, but most of the ajax stuff didn't because it requires the eventmachine interface). In fact, installing all the dependencies on an ubuntu server running a LAMP stack still required an extra 350+Mb of extra packages as all the ruby and mongodb dependencies, for a so far tiny web application. Talk about bloatware!
So although it may look good, it's been put together by crApple fanboys, aka morons. WTF were they smoking at burning man to make them think this was worth it? Gimme some of that sh*t!
Now I can network with all 3 people that care about both FOSS principles and social networking!
Diaspora allegedly gives one more control over their data, and how it is used, because as we all know, Facebook discussing "privacy" is like McDonald's discussing "nutrition"
Reply to That ||
I don't care at all about the source code being released. Sure, they've released some Ruby code, which you can run, but that's not the important bit. We don't all use SMTP because Sendmail is open source (although that did help adoption), we use it because the protocols are well documented and different implementations can all interoperate. Release the protocol specs as RFCs, merge in feedback, and encourage independent implementations. Until there are two independent implementations, the protocol isn't worth anything.
I am TheRaven on Soylent News
We should all help participate and support any open source initiative. I can't wait to see it working. The name is inspiring as an image of a great number of people migrating over in rebellion - but doesn't mean much alone, defined apart from FB, stand on its own merit, which would be better. So relies on FB popularity for meaning. Plus it sounds vaguely religious or biblical. RebelBook, PiratesBook, CorsarBook, PeaceWar perhaps.
Build your own energy sources from scratch. http://otherpower.com/
Where the usefulness of a service increases with the number of people using it.
http://en.wikipedia.org/wiki/Network_effect
i.e. everyone but Facebook, are money down the drain. They would have to fuck up monumentally to break the effect.
Deleted
Last I checked, the hosting was either going to be you download and run it on your own server, or you pay them X dollars for them to host it for you. Is that still going to be the case? If so, this thing is dead in the water because Aunt Jane has no idea what a web server is, and she's not going to buy hosting from Diaspora when Facebook is free.
The problem is that most people don't really care about something being open source and, unfortunately, these people usually make up the majority of the friends of people who do care. In other words, I'll use whatever everyone else is using.
I just liked how it described facebook as an intranet. Further the article posits that facebook will follow the path of AOL. I am not agreeing or disagreeing, just sharing a link.
http://www.codinghorror.com/blog/2007/06/avoiding-walled-gardens-on-the-internet.html
slashdot: where everyone yells sarcastic metaphors to themselves to understand the issue
I find your lack of faith disturbing.
RIP America
July 4, 1776 - September 11, 2001
"Come to Dysphoria; it's great!"
Just do what pidgin did and let people connect to all their social networks from the social network they control.
While the source being open is pretty important, the really important thing is that anybody can host a Diaspora node and link to anybody hosted on any other Diaspora node. And Diaspora will also include ways to link to people on things like Facebook. The idea is that just because all your friends are using X, you can still be linked to them effectively without using X.
Need a Python, C++, Unix, Linux develop
Right, agreed, but Open Source will encourage developers to get involved. If a critical mass of them do (as in Linux Distro Critical Mass)
then they can hopefully start selling people on the privacy aspects of it. Of course just as with open source a lot of people don't seem
to care about privacy either. But, who knows, perhaps with the open aspects of the system you'll get some killer app that will draw in
the masses that don't care about the underlying principles. I'm not saying it's a sure bet but even Linus didn't think Linux was going
anywhere big when he started it and look at it now...
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
I have always wondered why we needed Diaspora when there are already so many projects. Why not work on one of the existing ones.
After looking at the source code i got the impression, that they took that long to release just the output of the generated rails code.
And do you honestly expect the typical FB user to do that - set up and manage their own server?
So it will be a social network for people who are at least semi-competent with technology (or at least are friends with such a person). Sounds good to me - I already run a web server at home, and the kids know how to put stuff onto it.
I tried MySpace for a few days; told them to wipe my account (it was before they had an option for leaving). Tried FaceBook for a month or so, then zombified my account.
Those who can make you believe absurdities can make you commit atrocities. - Voltaire
Things we are working on next for our Alpha in October:
from http://www.joindiaspora.com/2010/09/15/developer-release.html
i love the alt text on the image - "aaa"
weinersmith
Seriously. That's a huge mistake., Scalability should be built into the core of it. Its okay if its just the interface, but the core of the app should be written in Scala, or Go. One can only hope that this is just a prototype that they will rewrite with a more solid base.
Well.. maybe. Or Maybe not. But Definitely not sort of.
Of course there are open source social networks. I cannot believe you don't know GNU Social!
The distributed nature of diaspora means that you can write a server in another language and still interact with it.
Well, what are you waiting for?
The key to social network interoperability is the 'OStatus' suite of protocols and formats. Diaspora will be implementing this, but what really exited me the other day was the first open source implementations of OStatus communication between Status.net (wot powers identi.ca, etc) and itself (screencast is available via the link above) based on the Federated Social Web's SWAT0 test, and then shortly afterwards other systems (MiniMe). Work is under way to implement this in other systems such as ELGG, Drupal, WordPress, Google Buzz, Diaspora, etc. Some probably slightly out of date info on this can be found on the Status.net wiki.
MilkMiruku
Well those are current and trendy. But is it the best? I suppose php and mysql is too boring.
And here I was thinking that the likes of Diaspora could be nicely installed on my router. With a load of luck and a pitchfork I might be able to get it on there because this router has more memory than my previous laptop but you might as well forget about getting this incarnation of Diaspora running on a WRT54GL. If lightning had not struck last month I'd still be running one of those with no plans to replace it until, well, lightning would strike.
I will try to keep an eye on what they are doing but I'm really more interested in the protocols and APIs they use and develop. One it all settles down I'd create something which interacts with their implementation without all the buzz they deem necessary in some nice, compact and high performance language. It might even fit on a WRT54GL then which would give it an instant base of who knows how many nodes...
--frank[at]unternet.org
Overengineering your first proof-of-concept implementation is a good way to get overwhelmed by its complexity and never complete it.
Thanks, had no idea. Looks intriguing.
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
That is of course 100% true.
However, there is always the danger that management falls in love with the the proof of concept, sales thinks its ready and promises it in two weeks. Then your "proof of concept" ends up getting shipped as the final product.
Ruby w/Rails is a very dangerous language to start a project like this in. Its very easy ( and tempting) to add in so many things that would be very difficult to scale. Its so easy to build a beautiful edifice on a house made of straw.
Well.. maybe. Or Maybe not. But Definitely not sort of.
It's awesome if this first implementation is solid and fast, but not really a requirement: As someone else already said "it's the protocols, stupid". Assuming the design is good, competing implementations will take over if the original is somehow not up to par.
A number of other social networks let you do just that.
I really like pip.io, for example. very pretty but just starting, they lack a number of functionnalities. and a mass of users.
I'm not a professional developer, but I know my way around some programming languages (C/C++, python, Java) and I wanted to play with diasporas code base. Unfortunately I do know jack about ruby and don't have the time to learn a new programming language. Well, too bad I guess :(
Maybe this: http://groups.google.com/group/diaspora-dev/browse_thread/thread/3eceb21134faada1#
I just want to let you all know that I'm offering free Diaspora seeds at http://diasporahosting.eu/. And yes, the name is temporary. :P
From what's been revealed of Diaspora it looks exactly like Facebook. So what's the point of switching? I realize it's open-source, but that doesn't matter to the vast majority of people. Most who will come across Diaspora will see it as a Facebook-clone, with the huge shortcoming that no friends are on it. Most people simply want something that works sufficiently well and is used by a lot of other people. Companies are lured to Facebook because of the significant potential for marketing.
This is a problem with the majority of open-source projects I've come across. They don't try to improve on an idea or at least reinterpret it. They merely recreate it.
That said, I think the project is a good one. Perhaps the eventual release will be more compelling than what's been revealed. It's certainly got its strong points but it's got to have some hook that will lure people.
Google might pick it up. Android shows Google's willingness to adopt openness as a "scorched earth" policy against competitors who're doing end runs around Google's core business. All the other social networking sites like hi5 might adopt it for strength in numbers vs. facebook. You could even imagine IM programs like skype jumping on the social networking bandwagon through variation on Disapora's protocols.
Also, friends-to-friend file sharing is the untapped killer app for social networking, as easy invisible widespread friend-to-friend piracy could finally muzzle the MafIAA bullshit. For example, Skype could steal facebooks thunder tomorrow if they built social networking into their client, but supported filetypes beyond merely photos. Friend-to-friend file sharing might emerge in Disapora if people started using stand alone clients, just support more filetypes than merely photos.
The Christian religion has been and still is the principal enemy of moral progress in the world. -- Bertrand Russell
So.... you like it, but it doesn't have the people you want to connect to or the functionality you're looking for.
So.... why do you like it again? Honest question - I'm now quite curious. It doesn't seem to come close to meeting the needs you had...
You're special forces then? That's great! I just love your olympics!
The options for running this are:
A) Run your own server to host your "node" of data
- The majority of people do not want to take the time to learn how to do this.
B) Pay somebody to host your data.
- Most people will always choose a free service or a paid one if they can get the same or better.
C) Host your data with a free but ad-supported provider
- Here is the most likely option, but I can't see many advertisers choosing such a model when they can instead go to Facebook and have displayed ads customized according to the user, which of course is defeating the entire point of the privacy fight with FB.
I wish the Diaspora guys luck, but I doubt this will amount to much long term.
No, there won't be a competing implementation, if the first one isn't capable of supporting a non-trivial load. This is social networking, you don't have any value associated with the product, until there is a sizeable user base.
Well.. maybe. Or Maybe not. But Definitely not sort of.
WTF is an Aspect?
The only one I'm aware of is Appleseed. It's also distributed, it's in development for several years now, has working beta-servers, and is probably much closer to a final release than Diaspora.
Few people care about Open Source, but more (still a minority, but it's a lot more) care about not-getting-fucked-over. Free Software is the solution to getting fucked over by the software that you use.
"Believe me!" -- Donald Trump
So.... you like it, but it doesn't have the people you want to connect to or the functionality you're looking for.
So.... why do you like it again? Honest question - I'm now quite curious. It doesn't seem to come close to meeting the needs you had...
I use it to access my facebook account because the people I know are on facebook, but there's a number of things I can't do anywhere else than facebook.com (posting something to only a subset of my contacts, setting up parameters, editing my profile...).
Those are things I can do in pip.io with my pip.io account, but since almost nobody I know is on pip.io, it has almost no impact.
When I say I like it, it's because I like the way it looks and it works. I can't really use it, I only sign up because I wanted to know how they do things, and I like what I saw. As a web developer, I'm really curious about how different people solve different problem, and I think social web problems ar really interesting.
pip.io, or diaspora may never end up being realy use by a lot of people, but each contributes to try to solve the problem that is "how to do a social network the right way?". I like both of their answers, and hope that in 5 or 10 years, the social network where most people will be will have inherited something from them.
many Muslim hardliners frequently misuse the term "holocaust" to define obviously inequivalent events.
Yes they have, and just for the record, in Hebrew there is a severe difference between 'shoah' (holocaust) and 'ha shoah' (THE Holocaust), and this in itself has been misused as well, in subtle ways that non-Hebrew speakers would not be aware of.
Reply to That ||
And Facebook is written in... PHP. Youtube is almost entirely written in Python. Twitter and LinkedIn are both written in Ruby.
Scalability is one of the least important factors in Diaspora's future, in my opinion. If, and that's a big if, it gets major uptake we might start thinking about developing multiple compatible implementations in different technologies - but right now, a language like Ruby allows fast development and the possible involvement of a huge community of coders, which neither Scala, Go, Erlang, or any other of those languages have.
Dilbert RSS feed
Yeah, that's the really nifty bit. However, I've heard that facebook is basically putting up servers as fast as they can build them. Distributed or not, it'll take a lot of server power to match facebook. Certainly possible though, and I hope it catches on.
A social network that limits it's audience to a specific group of people isn't very 'social'. It would fail if it was only for those interested in FOSS,
While everyone can install and run their own copy of Diaspora, they won't have to. Just as most people do not run their own mail server, I'm sure there will be plenty of hosting companies through which you will be able to get service. And if you are not happy with that company's service, you change providers.
Facebook was "Sparse and Clean" once upon a time. I remember this is one of the things like liked over the yahoo.geocities like pages of MySpace. Of course look at Facebook now that it is big, and trying to find a way to make money. Apps, and a mess. This could be a social media cycle thing going on here.
Can you play Mafiawars?
it will have (open-source) CommuneVille?
From what I've seen, they're missing the point. By about one pacific ocean.
Open Source is great when you're talking about an app on your computer that does stuff. For a website that is primarily a service, it matters little. As others have written, the interfaces are more important. The data is more important. Critical mass is more important. Image is more important. Heck, pretty much everything else is more important than the stupid code that runs the stupid site.
Building a "Facebook competitor" is about pretty much everything else but the code.
Assorted stuff I do sometimes: Lemuria.org
What's so interesting about it anyway?
This project could be technically perfect, but it will still fail, due to one thing alone:
Really, Really dumb choice of name.
Did you know that, for instance, in Brazil, Google's Orkut is the social network used by everyone and Facebook is barely used?
Facebook can (and probably will) be replaced when a better alternative comes up.
It's facebook in black & white. That's me convinced this is a win.
First of all, Twitter started out in Ruby, but really ran into sever problems very early in its growth. Its a poster child in for not using ruby to scale. Right now, its ruby only in the front end, the back end is scala. After many years of trying to make ruby scale, they gave up. Its been much more stable since they put scala in. I was hoping that a new company would be able to learn from the mistakes of a company like twitter.
Facebook, did a really, really good job. They used LAMP to its utmost potential, and deftly transitioned away from it. If you ask me, Facebook succeeded primarily because they did the scaling correctly. The load killed friendster, and hobbled myspace. Note they used PHP, not Zend Framework, or any other MVC.
So yeah, they could do it all in PHP, Ruby, python or similar. Point some what granted. Its just alarming that they seem to have embraced rails. I really think the key is to not over engineer, but also not use an existing framework that will stop you from growing very large. Considering that this is supposed to be a distributed system as well, any architecture changes that they make to the reference system would have to be duplicated by others.
Well.. maybe. Or Maybe not. But Definitely not sort of.
Typical of much open source software the name of this app is the "Kiss of Death", it's a sucky name, it will never catch on due to that.
But what are the #s? Or rather what is the user/server ratio? Why it's X to 1 of course, so as long as at least 1 out of every X users of this new distributed site runs a node, they're equalling the "server power" of facebook in a sense.
What's is X? Well I have no idea personally. But I can say that, as a business, facebook would be likely trying to minimize X to save costs. OTOH, someone likely to run a distributed node is only looking at one box, and if they are going to run the node it's because they want to run the node, not because it clears some corporate budget.
The real question there is what value of X is the critical threshhold of where the tables turn...assuming "server power" is the magical metric in the first place.
There's no ads in the mock-up.
Thing is, I use Facebook to keep in touch with friends and relatives. Only a minority of my friends would be interested in the fact that some software was open source, and not all of those would find whether it was OS or not a significant reason to pick a social network. My relatives are even less technical.
There's a lot of intelligent and competent people in those groups, and some have serious geek credit or the equivalent in their fields, so many of them are the sorts even Slashdotters might like in their social networks, and most of those wouldn't understand why I'd move to a social network because it was Free Software or Open Source.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
I think their plan is to release a standard protocol as well as one implementation of the protocol, and I certainly hope that the protocol will be implemented by other people.
But a standardized protocol is not everything. We've had a standardized protocol for instant messaging for some time (Jabber aka XMPP), but there are still many incompatible proprietary protocols out there (MSN, AIM, Skype, ...). Jabber is gaining some momentum because Google is using it for Google Talk, and the same thing might happen with Diaspora, but I don't think Facebook will switch to an open social networking protocol anytime soon. (However, they did move to Jabber for IM -- but their server is not connected to the rest of the Jabebr world.)
No, it doesn't, as long as the protocols are independent from the implementation, which any good protocol is.
You know what's the largest decentralized system on top of the Internet (TCP/IP) today? The Web.
Yet, and although there are thousands of different servers in hundreds of different languages, from C to Lua, a browser can talk to any of them interchangeably. Why? Because the protocol (HTTP) is independent of the implementation.
In fact, while REST is now too much hyped, I think sticking to some of its constraints (like for example, having an uniform interface) would be beneficial to this project.
Dilbert RSS feed
Given how facebook have fought against third parties vacuuming out data (even if the person is only doing so with their own account) i worry that the first point will be one serious cat and mouse game.
comment first, facts later. http://chem.tufts.edu/AnswersInScience/RelativityofWrong.htm
Comment removed based on user account deletion
Just port FarmVille and a few other addictive flash games to it, and the masses will follow !
No seriously, for Diaspora to have success, it needs a few things, which might help
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
It looks like a Ning site without the ads.
-- $G
Or to give a more recent example : blogs
Blogs where pretty much standard. Using basic technologies like RSS feeds, ping backs, backtracking, etc. or even pretty basic URL links, it was possible to organise a network.
Wanted to change the engine ? Just change the code behind your home page on your own server. Your readers will still be able to read you from the exact same homepage, no matter what the engine behind.
Don't want to host your own stuff ? There are lots of large scale providers hosting blogging platforms.
Thanks to standards it wasn't even that difficult to port your data from one hosting service to another.
---
Even more recent example : Chats.
It used to be that every single new comer had it's own secret protocol. Needing third party clients to reverse engineer the protocols. And displaying strongly the "but all my friends are on XyZ" phenomenon.
Now enters Jabber/XMPP.
- A single standard protocols meaning that any one could connect to this network with any 3rd party client. The popularity is so strong that even FaceBook and StudiVZ are starting to provide XMPP compatible gateways.
- A standard supporting "federation" : for service which did enable server-to-server communication, that means one can talk to all friends, no matter who is on GTalk, Jabber.org, etc. (although not all currently have it. Facebook user only chat to other facebook user, although XMPP is a possible back-end).
---
Social networks are just going through the same stages :
people are going fed-up of cycles repeating it self (newer platform where everyone has to move and rebuild a network, just after the previous one lost momentum).
some solutions are popping up. some more efficient than others. (opensocial failed to live up it's hype).
Now the only question is: will Diaspora manage to attract enough supporters (even former networks moving to diaspora-seed-compatible interfaces to allow accross-network integration) and have enough killer features, to gain momentum and take over the scene (just like XMPP is slowing taking over, at a time where MSN and the like where declared winner over ICQ and other older versions).
Or is it to late and FaceBook will become the "next microsoft" : too big and too popular for anyone wanting to move on, become the "by default" standard, even if it's not that much open (impossible to run away from it to another provider without losing everything) and has lots of short-comings (in the privacy department).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
From their website: http://github.com/diaspora/diaspora "DISCLAIMER: THIS IS PRE-ALPHA SOFTWARE AND SHOULD BE TREATED ACCORDINGLY. " So if you're expecting a world-class program that clearly outshines Facebook, I think you're being a tad premature.
(Note to the Diaspora developers: if you like that idea please use it, I promise not to sue.) :)
Eric Baird
Bah, you didn't understand the point I was making.
This is the reference implementation. This will be the gold standard in the early days of the diaspora project, if it takes off. Now, if the main Diaspora service hits a bottleneck which requires them to seemlessly transition form say a ruby back end messaging system to a scala one or from Mysql to cassandra, then everyone else who's running the reference implementation will have to undergo those same procedures to make the upgrade to the next release.
That is all. I understand its a protocol as much as it is an implementation of a protocol. Just because that is the case, it doesn't mean you'll have hundreds of different implementations tomorrow to choose from. The initial one, will have to be a good one inorder to get user and developer buy in.
Well.. maybe. Or Maybe not. But Definitely not sort of.
Has anyone noticed you can sign up at http://pivots.joindiaspora.com/ ?
but I don't think Facebook will switch to an open social networking protocol anytime soon. (However, they did move to Jabber for IM -- but their server is not connected to the rest of the Jabebr world.)
They moved to Jabber, because lots of users were complaining of the stability of the normal one and running away to other chat systems.
when the next usual "Huge! Privacy! Scandal!" hits FaceBook, if there's some viable alternative, some small portion of the users might start going to that alternative,
specially if, as announced, there's some inter-operation with facebook (so they don't completely lose their social assets during the switch) and if there are enough addictive-as-crack farmville-like applications.
if a huge potential competitor like google is behind one of the Diaspora-seed-compatible service, FaceBook might feel a little bit threatened and could open up a little bit their platform.
If diaspora is a big enough menace to facebook, FB might consider playing along.
But it needs to be a viable alternative, with a strong backing.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
I suspect the main developers are the children of moguls or millionaires with media clout?
There are a dozen better suited, further along, open source distributed social networks in existence.
Why this one gets all the hype can't simply be random.
"doesn't look unlike"
Isn't this not unavoidant of the double negative rules, making it unnecessarily confusing to read?
...they can keep their house of cards until it collapses. To compete with facebook doesn't have to mean you are a facebook clone. Besides, I think in the long term (perhaps way out there many years, but eventually) technology will develop and enough people will clue in--"get it" about what the internet should be--that facebook will be obsolete anyways. Why aspire to be a dinosaur?
People HAVE "got it" to some degree for decades now I think--it's just that it isn't all one big picture yet. Think about the way the internet came to be--designed to be fault-tolerant, distributed, many relatively autonomous, small components working together. Client-server vs. mainframe timeshare systems and serial terminals--the big mainframes replaced with smaller servers and clients growing from dumb units to capable computers. Think the "UNIX way", where all the tools are small and work together piping data amongst each other. Object oriented programming, shared libraries and so on--the pattern is to make the blocks smaller and stronger and more numerous, instead of carving your creation out of a giant granite boulder.
Facebook doesn't fit the paradigm of the future. Facebook is a mainframe, or a giant boulder, or the Titanic, or a 1973 Ford Thunderbird. It is big, slow lumbering and dinosaur-like. "Full featured" can be a competitive disadvantage at that point. Remember when the Web was still kind of like the internet was intended? Lots of smaller websites, personal pages dedicated to people's interests that resided on Freenet servers or university systems or local mom-and-pop ISPs. Corporations HAD websites. Today it seems the hip thing to do is regress to the dark days of mainframes. Instead of people having their own web page, URL or email address they have a facebook page, twitter ccount and a GMail address, almost creating an internet run by three companies. Isn't that what the creators of the internet, www and personal computers were trying to avoid?
I think if the infamous three get too big, too complicated and leave people feeling like they've lost too much control over their online presence they'll implode--especially should more distributed alternatives arise. For example, should Diaspora take off like, say, the Apache HTTP server, and it stays open and has a consistent interface that encourages following and sharing and interacting between different Diaspora sites then what would the point of Facebook be? Same with StatusNet--if enough of those servers were out there why put up with Twitter fail whales? Small distributed parts are much better than the "barge with barnacles" approach taken by Facebook any day.
Think about it: If Facebook changes something and you don't like it you just have to get used to it, or else give up facebook and lose all the content you've created, all the links you've made with others and so on. If it was a million Diaspora's instead of one Facebook you could just move to another server--take your whole profile with you--because it is YOUR data and it is a platform NOT controlled by one large entity with a vested interest in doing as much as possible to keep you from moving. If facebook has technical problems huge swaths of users suffer. If it was a million Diaspora's and one goes down nobody would notice beyond the dozens or hundreds on that site.
Such a social network model means that Diaspora doesn't have to cater to FOSS folks only. You could have many Diaspora-based sites that cater to different interests or communities or corporations or whatever,, but since the platform is open it could be a consistent/standard platform allowing full interactivity amongst the different sites. Your profile could reside on a FOSS social network site and your sister could have her profile on, say, a gardening social net site, and you could be both on each others contacts and chat and interact and share and follow and EVERYTHING that you can do between users within facebook if you realy wanted to.
If Diaspora evolves to have a well-defined API for apps
However, I've heard that facebook is basically putting up servers as fast as they can build them.
Facebook could have thousands--it doesn't mater. The point is you must go to facebook.com to get to any of them. Facebook owns all of them, their load balancers tell you which of the boxes handles your requests, they control access to your information and decide for themselves how it is stored, processes, indexed, shared. Facebook can cange the site however they want, even if you don't like it. All the severs are on THEIR network, so if their network goes down we ALL fall down too. The same staff has full access to all the data too--so don't do ANYTHING on facebook you wouldn't do on prime time national TV and expect it to stay private.
"strength in numbers" isn't the whole point of this thing--personal control and prevention of lock-in are what appeal even more. Whether you are a personal user of facebook who wants more freedom or a business that wants their own "in house" social networking server/site to do social marketing, customer support/service, etc. that has to be pretty enticing to a lot of people.
Given that it's something that is designed to be installed on any domain at all, it may not be required to be called Diaspora. I don't know many people who refer to their gchat account as jabber.
And do you honestly expect the typical FB user to do that - set up and manage their own server?
And if specialized service providers sprout up to host this data, wouldn't that be creating the same situation that this software is supposed to be trying to get away from: other having control of your data?
No, only nerds would host their own server at this point (like me). However "normal people" more and more have "appliance" servers in their homes today (home routers, PVRs/digital tv receivers, etc that have the capability and perform some server-like functions) so the day may come where self-hosting your own services becomes fairly common.
This would be like email or generic web servers fro the most part as has been said. ISPs and hosting companies and corporations would host their own public social servers, and an open platform would kill lock-in and enable full interaction amongst users on every server as if they were all on facebook, except without entrusting one giant system with all the data.
Control over your own data doesn't have to mean owning and maintaining the server yourself. The open and distributed nature of the platform gives you the control you need. Just as the case is with any free market you could pull up stakes and move, or maintain different profiles on different sites that can be kept in sync or whatever. None of this nonsense like between Facebook and Apple or IM systems like MSN and Yahoo and AIM changing their protocols for the express purpose of breaking third-party clients that provide inter-operation and so on.
Yeah, I thought you were talking about forks, not installations.
I think you over estimate the performance requirements. Due to the distributed nature of the system, the load is balanced between many servers, which will probably have only a few hundreds or a couple of thousands of users.
The way I see it, the Ruby implementation will be good enough for most servers out there.
Only when some nodes start reaching the tens of thousands of users will it become a real issue, and in the mean time either:
1) The project will die, turning the point moot
2) Multiple implementations will appear
Just like there are some implementations that can be faster than CPython (Stackless, for example), doesn't mean the reference implementation changes. Just like CPython is the best for most apps, it'll be the most appropriate choice for most nodes.
In fact, I wonder if they shouldn't have done it in PHP. Not because it scales better (I doubt it), but because it could rely on the hundreds of cheap hosting packages 'round the world to form new nodes.
Either way, they need to package it properly, making an installer with all the dependencies and a very simple web interface to control it, allowing people to run their nodes easily.
Dilbert RSS feed
I'm of the mind, that PHP would have been a better choice, simply because the libraries for it suck so much that they wouldn't rely on them.
There seem to be two schools of thoughts on how it would end up. You have the hotmail/ yahoo mail/gmail folks who think it will end up being a handfull of large players backing it. And you have the web hosting folk who think each implementation will be small. If its fewer larger players, you still have to fear for privacy, but not worry about the implementations. If its more smaller players, then you do have to worry about the ease of installation and upgrades.
I stil think it will be still born ala google wave.
Well.. maybe. Or Maybe not. But Definitely not sort of.
I just don't see the business case for a major player investing in it.
A large company would want to run it in something closer to a cluster than a decentralized network and without encryption between the nodes, meaning the whole back-end would have to be redesigned, which would be probably more work than re-implement the whole thing from scratch.
End-to-end encryption also doesn't cope well with advertisement needs.
And then there's the question of it being Affero GPL licensed, which means they'd have to distribute their fork.
It'll either die or, in the best case scenario, it'll keep living for a small community, like Freenet does. The rest of the world will continue to obliviously share their personal data with FB and its advertisers. Privacy simply doesn't sell.
Dilbert RSS feed