Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[liberationtech] Fw: [progressiveexchange] Facebook interfering with activism Pages

Adam Fisk a at littleshoot.org
Wed Sep 22 18:01:53 PDT 2010


Hi Yosem- I think we're conflating a few different concepts here, so
let's dig a little deeper. As you mention, most of the existing social
networks offer APIs and support various protocols for federating or
sharing data with other social networks. That's great, and it's easy.
Diaspora is talking about a distributed social network in a much
different sense, however, where users themselves run servers. That's a
different ballgame. Some users will be behind home routers, some users
will be running old versions, and all of them have to somehow
interoperate. Depending on how far they take it, this means you have
to do fancy NAT/firewall traversal to get at those servers and beyond
that somehow coordinate all that data across potentially millions of
individual computers. The most advanced distributed databases around
barely due this for multiple data centers across the WAN, let alone
integrating peers behind heterogeneous network configurations. That
makes some of the harder things you can tackle:

1) NAT/firewall traversal that has to work all the time
2) Distributed databases that incorporate high numbers of peers and
have reasonable access times. You can't scale Cassandra like that
(could never have more than a couple thousand nodes because their DHT
maintains O(1) lookups through connecting to all nodes), and even
BigTable would be hard-pressed to do it.
3) Combining the two -- never been done to my knowledge

This is also unlike Zuckerburg's early forays into file sharing on
Facebook. He wasn't sharing the data of the social network itself but
rather specific files on users' machines, likely relayed through
servers. That's kid's stuff compared to a fully distributed social
network Diaspora is proposing.

I do agree this is largely an organizational issue, however -- in fact
I agree 100%. From a technical standpoint Facebook is the correct
solution or very close to it in my mind, but the organizational and
even legal issues are where things get sticky. I just want to be clear
that from a technical perspective it doesn't get much harder than what
Diaspora is proposing.

Now there might be some shortcuts and different designs (don't use a
fully connected database, for example), but you're still looking at a
monumental technical challenge.

-Adam


On Tue, Sep 21, 2010 at 12:49 PM, Yosem Companys <companys at stanford.edu> wrote:
> My thoughts:
> I don't think it's a difficult technical problem.  All the parts to put
> together a distributed social network exist.  The challenges, in my view,
> are more organizational (and financial).  Take a look at the folks from the
> Federated Social Web, which includes Google and Facebook.  They are all
> pushing data portability and distributedness.  They have already built the
> infrastructure to enable a distributed social network to develop.  In fact,
> Zuckerberg donated to Diaspora, no doubt because he was the first one to
> create a distributed social network, simultaneously while creating Facebook.
>  The distributed social network was killed by Facebook, however, over
> concerns that entertainment companies would sue due to the potential file
> sharing applications available in such a network.  But this is why, though
> Diaspora is painted as a Facebook killer, Facebook seems to be supporting
> Diaspora's efforts.
> I also think there is a key difference between Diaspora and Haystack (and
> please correct me if I am wrong Haystack folks).  My understanding is that
> Haystack sought out media attention.  Diaspora has not.  The only reason why
> they have gotten so much media exposure is because the media sought them
> out.  I don't think it's bad to work on something that may end up being sexy
> to the media.  I would be surprised if someone could pinpoint an example of
> a Silicon Valley startup like Google, FB, Apple, etc. that became successful
> without media awareness.  And in those cases, they had a public relations
> person running the show.  Diaspora doesn't have any PR people; it's only
> members are all programmers.  The more real danger is the hype building
> great -- or unrealistic -- expectations of what Diaspora can accomplish.  So
> far, from what I have seen, however, Diaspora has kept the problem small and
> focused.  Also note that Diaspora is open source, so anyone can read and
> improve the code.  This sounds like another distinction from Haystack.
> Sure, the Diaspora guys have limited experience.  Did Zuckerberg when he
> created the same thing 4 years ago?  (And I mean the distributed social net
> and not FB.)  Did the Yahoo guys?  I think organizational aspects -- putting
> together a structure to get a product done, marketing it effectively,
> raising money, etc. are much more important in this regard.  And anyone who
> sets out to compete with FB will need to deal with the critical mass
> effective:  How do you convince users to change from FB to an alternative?
> I agree with Adam that we're not close to a solution.  Lots of things need
> to happen before Diaspora or anyone else can be successful at tackling this
> problem.  But like Adam, I think identifying the problem is important.  It
> helps people focus on what needs to be done to improve existing systems.
> BTW, Jim, I don't have anything against corporate structures.  FB has
> created a private company to profit off of social interactions.  They watch
> everything you do and then sell that data or allow people to target
> advertising.  Before the Internet, many of these activities took place in
> public spaces.  And when the activities took place in a private setting such
> as a coffee shop, I think most people would object to having the coffee shop
> read their personal papers, or observe them via video cameras to sell the
> data to other coffee shops or firms, or to take their pictures and use them
> for advertising because the coffee shop's privacy policy, which you sign
> without reading when you first walk in gives them a right to do so.  It
> would be creepy.  Kudos for FB to get people to give up on these rights
> online and build a business model around it.  But there is no reason why a
> public space social network, like a park in the real world, as a non profit,
> for example, could not be as effective as FB without engaging in these
> privacy issues.  FB does it because they have the pressure to achieve
> profitability.  Otherwise, as Zuckerberg has said countless times, he
> wouldn't worry about advertising or making money.  Wikipedia seems to have
> done just fine as a non profit.  And I don't hear folks questioning why
> Wikipedia is not monetizing its knowledge base?
> Best,
> Yosem
>
> On Tue, Sep 21, 2010 at 12:21 PM, Adam Fisk <a at littleshoot.org> wrote:
>>
>> I've been resisting raining on the Diaspora parade, but there are many
>> areas of concern.
>>
>> The first issue I'll call the BullsEye/Haystack pattern. Attach a
>> project to a sexy headline that will get infinite press. Do it before
>> you have a line of code. Start the project with inexperienced
>> developers who are in way over their heads (any Diaspora readers on
>> the list, please use that as motivation -- I actually want you to
>> succeed). Ring any bells? It would be fun to graph the ratio of words
>> written about projects to their lines of code and maybe mix in the
>> lines of code their developers have ever written in their lifetimes.
>> Both Diaspora and Haystack would be firmly in the "mayday/abandon
>> ship" area of that graph.
>>
>> Digging a little deeper into the Diaspora premise, Martin Fowler's
>> First Law of Distributed Object Design comes to mind: Don't distribute
>> your objects! In other words, if you can avoid doing peer-to-peer
>> computing, you should avoid it. The reason is simple: P2P is hard. It
>> makes *almost* any project far more complicated and time consuming
>> than it otherwise would be. A social network might be the
>> quintessential example of when not to use P2P, as the concerns include
>> privacy, security, data, big files, etc. Those are hard problems to
>> begin with, and tackling them in a distributed architecture is just
>> way harder.
>>
>> Do I really want to be a part of a social network where I'm relying on
>> my semi-tech literate friend down the street running his server at
>> home with half his ports opened up with access to my private data? Any
>> semi-proficient hacker would be chomping at the bit for this to become
>> popular. Oh, and now there's a big security breach that leaks a
>> million e-mail addresses. How do I stop it? Send out an e-mail blast
>> to get everyone to update their home computers? This is not Zuckerburg
>> coding thefacebook.com in his dorm room. It's about an order of
>> magnitude harder. It's pretty clear now that Zuckerburg was a serious
>> bad ass technically for his age, and I think he would have been in way
>> over his head if he tried to do Diaspora instead of Facebook. There's
>> no doubt it my mind it would have crashed and burned.
>>
>> The only exceptions to Fowler's rule come when the benefits of a
>> distributed architecture make the complexity overwhelmingly worth it.
>> Over the years it's become clear that P2P works well for one thing in
>> particular: distributing large files. In that case, it can save a lot
>> of money and is extremely efficient. Otherwise, P2P architectures cost
>> much more money and are far less efficient. Things like SETI at home and
>> folding at home are interesting edge cases, and I'm a huge fan, but their
>> task is very specific and again well-suited to P2P.
>>
>> The Diaspora home page should ring the alarm bells: "The privacy
>> aware, personally controlled, do-it-all, open source social network."
>> The "do-it-all" kind of sounds like the kid that will show Iran how
>> the Internet works, no? It's great that it's open source, and it's
>> great the team is soliciting feedback, but note that Facebook also
>> runs largely on open source software, and its open source software has
>> actually benefitted companies outside of Facebook in major ways. It's
>> also telling there appear to be almost no P2P components in Diaspora
>> to speak of. It's basically an open source web site you can run at
>> home too.
>>
>> I agree with you on the problem, Yosem, but I don't think we're close
>> to the solution. All that said, I'd be more than happy to lend a hand
>> with any of this, so please don't hesitate to get in touch if anyone
>> from Diaspora is reading this.
>>
>> -Adam
>>
>>
>> On Tue, Sep 21, 2010 at 9:26 AM, Yosem Companys <companys at stanford.edu>
>> wrote:
>> > This is why we need peer to peer solutions like Diaspora to provide a
>> > platform for activists.  Otherwise the corporate structure of these tech
>> > firms can always interfere in one way or another to direct activity.
>> >
>> > On Mon, Sep 20, 2010 at 9:58 AM, Jillian C. York
>> > <jilliancyork at gmail.com>
>> > wrote:
>> >>
>> >> Hi - Jillian York here (have been lurking for a couple of weeks).  I
>> >> was
>> >> interviewed for that article and would like to add this just-released
>> >> paper
>> >> I wrote on the same subject: http://bit.ly/8ZN8PH
>> >> The paper ("Policing Content in the Quasi-Public Sphere") looks at the
>> >> content regulation policies of 5 social media platforms--Twitter,
>> >> Flickr,
>> >> Blogger, Facebook, and YouTube.  I won't give away the ending ;)
>> >> -Jillian
>> >>
>> >> On Mon, Sep 20, 2010 at 5:28 PM, Katrin Verclas
>> >> <katrinverclas at gmail.com>
>> >> wrote:
>> >>>
>> >>> Thoughts, colleagues!
>> >>>
>> >>> ------Original Message------
>> >>> From: Colin Delany
>> >>> Sender: Progressive Exchange
>> >>> To: Progressive Exchange
>> >>> ReplyTo: Colin Delany
>> >>> Subject: [progressiveexchange] Facebook interfering with activism
>> >>> Pages
>> >>> Sent: Sep 20, 2010 3:34 PM
>> >>>
>> >>> Activists upset with Facebook
>> >>> http://www.politico.com/news/stories/0910/42364.html
>> >>>
>> >>> Claims they've violated terms of service.  Discuss.
>> >>>
>> >>>
>> >>> --cpd
>> >>>
>> >>>
>> >>>
>> >>>
>> >>> ~~~----~~-~-~-~-~-~-~-~-~-~-~-~-~~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~
>> >>> Group Links:
>> >>>
>> >>>  * To reply to sender (Colin Delany):
>> >>>   cpd at epolitics.com
>> >>>
>> >>>  * To view sender's profile, click this link:
>> >>>   http://www.progressiveexchange.org/profile.htm?mode=vp&pid=2401
>> >>>
>> >>>  • To visit this group on the web, go to:
>> >>>   http://www.progressiveexchange.org/group.htm?igid=3
>> >>>
>> >>>  • To send email to entire group:
>> >>>   Discuss at progressiveexchange.org
>> >>>
>> >>>  • To email the group moderator(s) (Eve, Jonathan, Steve, Marc):
>> >>>   Discuss-moderator at progressiveexchange.org
>> >>>
>> >>>  • UNSUBSCRIBE or change delivery settings:
>> >>>
>> >>>
>> >>> http://www.progressiveexchange.org/p.htm?m=qs&em=katrinverclas@gmail.com&sk=vTtA2Z
>> >>>
>> >>> Powered by GoLightly, ProgressiveExchange's Collaboration Platform of
>> >>> Choice.
>> >>> http://www.golightly.com/
>> >>> ~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~~-~-~-~-~-~-~-~-~-~-~-~-~~~~-~--~
>> >>>
>> >>>
>> >>>
>> >>>
>> >>> Sent via BlackBerry
>> >>> _______________________________________________
>> >>> liberationtech mailing list
>> >>> liberationtech at lists.stanford.edu
>> >>>
>> >>> Should you need to change your subscription options, please go to:
>> >>>
>> >>> https://mailman.stanford.edu/mailman/listinfo/liberationtech
>> >>
>> >>
>> >>
>> >> --
>> >> Berkman Center for Internet and Society |
>> >> https://cyber.law.harvard.edu/people/jyork
>> >> jilliancyork.com | @jilliancyork | tel: +1-857-891-4244
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> liberationtech mailing list
>> >> liberationtech at lists.stanford.edu
>> >>
>> >> Should you need to change your subscription options, please go to:
>> >>
>> >> https://mailman.stanford.edu/mailman/listinfo/liberationtech
>> >>
>> >
>> >
>> > _______________________________________________
>> > liberationtech mailing list
>> > liberationtech at lists.stanford.edu
>> >
>> > Should you need to change your subscription options, please go to:
>> >
>> > https://mailman.stanford.edu/mailman/listinfo/liberationtech
>> >
>> >
>>
>>
>>
>> --
>> Adam Fisk
>> http://www.littleshoot.org | http://adamfisk.wordpress.com |
>> http://twitter.com/adamfisk
>
>



-- 
Adam Fisk
http://www.littleshoot.org | http://adamfisk.wordpress.com |
http://twitter.com/adamfisk



More information about the liberationtech mailing list