OLike Distribution

[Jonah Peretti]

The main reason publishers use FB Like is for distribution to FB's massive audience.  To succeed, OpenLike needs to provide massive distribution AND public data exhaust.  

One possible solution:

Wikipedia provides distribution for OpenLike.  Whenever you OLike something it adds the URI to the relevant Wikipedia page with the number of people who liked it.  If there is not a relevant Wikipedia page it automatically generates a stub and a like stream.  URIs on each wiki page are ranked by # of likes and personalized by # of likes from your friends (if your identity is known).  Then every wikipedia page would be enhanced with a ranked list of relevant links which would send traffic back to sites that support OLike.  And the Wikimedia Foundation would be the keeper of GPLed public like data for the benefit of all.  

[Elizabeth Stark]

This is an excellent idea. I can bring it up with various Wikimedia board members + Jimmy Wales if need be.

Ultimately it would be the community that would have to decide, but I think this is certainly in the range of possibility.

[Jeff Eddings]

Love this idea.  We should also talk with Craigslist, local and federal government web sites, etc.  Once we have a (semi-)fully baked solution, having an open "like" solution will be appealing to many who don't want to be tied to a commercial entity like Facebook.

Jeff

[Rabbit]

If the concern is "How do we compete with the data Facebook can accumulate?" then WikiLike or CraigsLike might be the answer.

The question is whether the open web response to Facebook Like is going to be a centralized response or a distributed one. I obviously prefer the distributed approach. imho, Facebook has done a wonderful job illustrating the service-centric approach to this technology. The alternative should be a user-centric approach.

=Rabbit

[Jonah Peretti]

Users benefit from having their likes widely distributed, having a socially enhanced wikipedia, having a database of likes that lead to cool apps created by other companies, and having competition between FB and open platforms.  It is like android and iPhone - great for users to have both.

[Rabbit]

I'm certainly not disagreeing with what you are saying in spirit. I am just saying that if the "open" version of what FB is doing is to do the same thing with a different domain (be it Wikipedia or Google or LovableOpenCompany) I just don't see the difference. Might as well use Facebook because it already exists and plenty of people will use it.

I feel like you can have the benefits of a widely distributed "likes" store without having to rely on a single domain to provide it. I will certainly concede to whatever the community wants to do. I just won't find the result very interesting if it's pushing my likes to Wikipedia instead of Facebook.

=Rabbit

[Jeff Eddings]

I don't think pushing likes to Wikipedia is the end game, but merely something that give a nice kick start to the process, gets people to understand why this is interesting and important, and why not giving a single commercial entity control over your likes/reads/whatever is critical.  It's similar to the feedback I gave re: the site likefeature.com that was put together.  I love the idea, but we have to be careful to make that one of the options, and not the (default) option.

Jeff

[Jonah Peretti]

Exactly.  The likes can go lots of places but knowing they will go to wikipedia would get everyone on board.

[sutter]

This is a spectacular idea that must be done if at all possible.

> --
> Subscription settings:http://groups.google.com/group/openlike/subscribe?hl=en

[Nathan]

a perhaps even more spectacular solution is to use the billions or uri's
to identify things all ready out there, if I may nudge your attention
towards dbpedia for instance which is wikipedia, turned in to machine
readable data, for the very purpose you say, with the added benefit that
when you look up a uri for a thing, you get mountains of machine
readable data back about that thing (inc all wikipedia info) and links
to wikipedia, and interlinks between datasets (like geonames and many
many more).

I really have to point out that this is re-inventing the wheel, it's
called "linked data", invented by Tim Berners-Lee (you may recognise the
name), has been proven, deployed on mass, adopted all ready by major
governments and institutions around the world, is the definition or web
3 / web of data, is rolled out in RDFa all over including the next
drupal, has a huge following, OpenGraph is linked data (yes open graph),
and as mentioned there are URI's for virtually everything on the planet
existing and ready to use, + a huge interlinking distributed web of
linked data that is in to the many billions of bits of info now.

Linked Data:

- Use URIs as names for things
- Use HTTP URIs so that people can look up those names.
- When someone looks up a URI, provide useful information, using the
standards (RDF, SPARQL)
- Include links to other URIs. so that they can discover more things.

sound familiar?

Sincerely guys, jump in, what you are describing is all ready a reality,
ready for you whenever you want - every topic thus far on openlike has
pretty much pointed to, or described, linked data.

Best,

Nathan

[NickMikhailovsky]

Cool idea about govt use. I'll talk with DC gov\OCTO Labs.

[Nathan]

It's already implemented, adopted and being used by uk and us govs, and
many local authorities ;) Even the prime minister of the UK talked about
it recently, is being guided by Sir Tim Berners Lee on the matter, and
Obama has pushed for the same thing.

Honestly, this is reinventing what's all ready there and being worked on :)

Best,

Nathan

[bernard]

Hi folks

Lurking on this forum since a couple of days, and jumping in after the
call of Nathan on the linked data forum, to follow-up with his
recommendation.
Each "thing" a Wikipedia page is about is identified by a matching
DBpedia URI, and this URIs is de-referencable in the linked data cloud
providing a wealth of structured data.
So if anything, build on that, yes, by all means. See e.g., how a
semantic bookmarking service such as Faviki builds on top of this.

Re. adding information on Wikipedia pages, this does not seem a good
idea given the instinctive reject of the WP community of anything
added to pages piloted from the outside. DBpedia or other linked data
URIs is the place to put the added value.

Cheers

Bernard
http://twitter.com/hubject/

[sutter]

Nathan, We're talking about like buttons.

[Nathan]

Ahh, forgive me I thought: Openlike was "An open protocol to allow
sharing the things people like in a simple and standard method between
web applications." - I must admit though I did also think it was just a
few like buttons too (earlier).

if it's just like buttons then all of this is way way off topic, as with
most of the discussions at the minute.

what is it then? a few buttons in javascript that somebody can add (ie
another addthis but for "likes" only) or something so much more..

Nathan

[Kingsley Idehen]

sutter wrote:
> Nathan, We're talking about like buttons.
>

What happens when a user clicks the "Like Button" ? I believe something
happens (i.e, data is created and persisted somewhere).

Kingsley

--

Regards,

Kingsley Idehen
President & CEO
OpenLink Software
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen

[Steve Crossan]

Like the idea. I think though that OpenLike should encompass other services. For a publisher, OpenLike should give them FB's distribution plus openness, plus the distribution of other networks.

--
Steve Crossan
+1 650 276 0788