Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Ask HN: Debriefing Diaspora; where did flexibility go?
7 points by dirkgadsden on Oct 12, 2010 | hide | past | favorite | 5 comments
In mere moments, it seemed, the Diaspora team managed to capture the hopes of thousands of people, who jumped on the Diaspora bandwagon to the tune of hundreds of thousands of dollars. The developer preview that the team then published was a catastrophic failure. It's time for a debriefing.

All the protocols for open social networks have two massive flaws, ambition and scope. Diaspora tries to replicate Facebook on a distributed level, but forgets about security. DFRN obsesses about the nitty gritty details of networking/connections, but forgets about the actual socializing. Diaspora lunged beyond the fundamentals of flexibility and extensibility and instead focused on a feature list of photos, messaging, and other social interactions. They missed the big picture. You can do virtually anything with the foundational protocols of the Internet, but Diaspora, for all it's hyped-up freedom, only lets you make photo albums? What happened to the good ol' days of hacking (in the classical sense of the word)?

My question to you, the brilliant Hacker community, is why are we not going back to the basics? Instead of trying to fight the fire (Facebook) with fire (duplicating it), let's fight it with something better. Why are we trying to prematurely define how people will interact? Why not build a namespaced, compartmentalized system that allows for any type of data, be that a definition of relationships with other profiles, high scores from a game, or photo albums? Flexibility used to be the name of the game, but it seems that it's now all about rigidity and locking everything, the user included, down.



FaceBook isn't popular because it's software, it's popular because it lets you talk to your friends. It lets you talk to your friends because that's what the software is made for. Software that lets you share anything won't be used to share anything.

Users often take their cues for what they're supposed to do with the product by what the product allows them to do. If the answer is "everything", like with Google Wave, then the normal user response is "???" Twitter succeeded because of the 140 character limit: it gave the user a cue that this software was supposed to be used for light, fluffy thoughts that you don't mind sharing with the world. Within months, FaceBook had come out with status updates, which are superior in every technical respect (they leverage your existing social connections, the length limit is much less restrictive, you can share other stuff with them too, and you can restrict them to only certain groups), and yet this hasn't made a dent in Twitter's growth. People continue to use it for light, fluffy comments because that's what it's made for.


Personally I hate extra features; but that doesn't mean that the possibility shouldn't exist. Diaspora's plans for opening up their system appears to be "we'll add plugins later." This seems fundamentally flawed to me. I'd rather use a protocol that had a minimalistic base that could then be extended with the desired functionality. I really don't care about relationship statuses, so why should I have my profile cluttered with them? The best extensions of a flexible protocol would grow in popularity as users demanded them; instead of having to wait for the Diaspora team or some other ruling body to approve your patch to add in a new feature, you could write your own extension to the base protocol and convince other people to implement that extension on their profiles. If other people liked them then they'd tell their friends, and the extension would spread organically according to demand, not by decree of some Overload of the Social Networks.

I don't really care about the users and features at this point; the foundation is what matters, and it seems to me that the current foundations are all low-grade concrete and not reinforced at all with rebar.


If they didn't care about the users and features, chances are there won't be any, and without users and features, the project is unlikely to continue existing for any length of time.


Usually such generalized approaches fail because there's nothing keeping the code connected to reality. http://www.joelonsoftware.com/articles/fog0000000018.html


The answer is "because the people that were charged with building the system are not capable of thinking on that level."




Consider applying for YC's Summer 2026 batch! Applications are open till May 4

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

Search: