BuddyPress

I found this article http://www.nytimes.com/2010/05/12/nyregion/12about.html about a bunch of guys creating an alternative to FaceBook. Here’s the thing, BuddyPress is all packaged and ready to go. The only thing it’s missing is the “node” concept they talk about. So that each individual can host their own site and connect with the sites of others via a meta-network…

Any thoughts?

Wow, 4 guys with $10,000 and an idea good enough to impress a NYT reporter on deadline. The future is changed forever!

But seriously, I do recall an audio interview with Andy Peatling more than a year ago where I think he floated an idea of BP networks all interconnected and talking to each other, and for all I know, this is a plugin or core feature in some state of alpha or beta. Maybe somebody with more awareness of the overall project can clarify.

Seriously, they’re up to $30k+ now. SO what do you think about adapting BP…

i’d hop on the bandwagon if someone could point me in the right direction…

Always good to have options. Four coders on the project sound good.

Here’s one thing that caught my eye when the Diaspora coders were talking about similar projects that failed (DiSo):

http://joindiaspora.com/2010/04/30/a-response-to-mr-villa.html

To answer directly, I am pretty sure DISO failed because:

2) (a guess) they quickly tried to support all sort of features without building some sort of common infrastructure before trying to add functionality.

I would have replied and mentioned BP of course
But I also get their point.

“1) They tried to add on to Wordpress, a project which was not designed from the ground up to be a distributed network.”

This really feels like the “elephant in the room” every time I visit my wp-admin panel. My BuddyPress admin menu keeps growing and growing, disproportionate to the WP-specific admin menus.

I keep waiting for the discussions to start on how BP is “outgrowing” WP. I guess that happens when the limitations of WP start to outweigh the advantages, such as all the great WP functions we get “for free”. When you think about it, it is a little weird – for example, how blogs and forums are integrated with BP in such different ways and don’t really “talk” to each other in a meaningful way. The whole idea of “like” and “favorite” and “report” buttons that work on one content type but not another. Etc.

“1) They tried to add on to Wordpress, a project which was not designed from the ground up to be a distributed network.”

okay, here’s what I don’t get – isn’t MU a distributed network?

@andrea_r I’d say that MU isn’t distributed in the way that’s important here, because the data is all stored in the same db, on the same server (at least that’s how it works by default). In a truly distributed network, each individual/cluster of individuals would have full, sysadmin level control over his or her data. WP/BP is still largely under the control of the person who owns the particular machine where the stuff lives – if the owner decides to pull the plug, there goes the network.

Wow, amazing analysis by Boone… I really “get it” now. Andrea_r also has a good point though that WPMU is distributed in other ways that are meaningful and powerful. I think the author’s comment was just speculation, and limited to a project called “DiSo” that was apparently designed to integrate with WordPress; not sure if DiSo worked at all with MU.

The part about that quote that resonated with me is just the idea of “WP wasn’t designed from the ground up to host BP.” When I’m writing something for BP, I always feel like I’m gaming WordPress to do something it was not really designed to do. So I’m writing code that somehow has to play nice with the host, WordPress, which is awesome in its own right – but what I’m really trying to do is build a social network. So it would be more logical to me to say, “OK, here’s my BuddyPress network. WP is really good at XYZ, so how can I adapt *WordPress* to meet my BP needs?”

Boone is correct. A distributed (sometimes also called decentralized) network is when data are kept in different places, places that each user (be that an individual, group, or company) controls. WPMU, Twitter, Facebook, etcetera control the users’ data and experience. Although the data may be spread throughout a series of servers or even distributed server clusters, the data is controlled by one authority and is often considered locked into a closed data silo–even if that data silo is hosted in the cloud.

For more details on distributed social networking (with a Semantic Web twist) see my article, A Flock of Twitters: Decentralized Semantic Microblogging.

I was thinking about how to tackle that problem. What I envisioned would be that each member would have a “virtual calling card”. When they connect with someone in the decentralized network the person they connect with would store this (and only this) card in their database.

Essentially what would be included would be their basic, public information, including an rss of their news-stream. The effect would be as if you went to your facebook page. Just as in FB when you click on an album you go to that page, this would work the same way.

No one authority would need to manage this system.

To solve to WP/BP issues above I think that it would be really effective to leverage the is_user_logged_in() and current_user_can(‘administrator’) functions:
If a user is not logged in, it shows a regular theme, like any other blog.
If user is logged in it show the BP interface.
If user is admin it shows the BP dashboard.
This would effectively separate between wordpress and buddypress dashboards, as well as offer a person the ability to have a one-stop shop for their social and professional needs.

For the DB issues, bbpress requires it’s own table but shares the user table with wordpress, would this be an option?

Just read your article @Jeff_Sayre looks like you were thinking what I posted above already… =)

Who would I contact about getting involved in such a project?

I am game for getting involved in such a project. What frustrates me about Facebook is that they had the opportunity to link into the LOD cloud and did not take it. Their “Open” graph API pirates the concept of microformats, RDF, and schemas that have been around for years. They pretty much just took the ideas and made their own version. Bad form, I say!

What is missing from this conversation, though, is the idea that a distributed social network is not as much about where the data is and who controls it as it is about how the data is defined or “marked up.” So, yes, each person has that “virtual calling card” you define above, but it is in a standard XML file like RDF using a specific schema or “vocabulary” common to people who use a particular service. That RDF file, though, belongs to me, the user, and I can choose to add my affiliation with a specific network to that file or not.

Seems to me that what Mu (or the very excellent WordPress 3.0) needs is its own “vocabulary” within the LInked Open Data cloud to be a truly “open” distributed network.

Just my 2 cents…