[foaf-protocols] foaf+ssl & like for blogs
kidehen at openlinksw.com
Wed May 26 13:10:21 CEST 2010
Story Henry wrote:
> Summary of responses:
> types of liking:
> - nothing: one just wants the person who wrote the blog to know one read it
> (kind of like ping)
> - one would like to give a degree of liking with stars ✩, ✩✩, ✩✩✩, ✩✩✩✩, ✩✩✩✩✩
> (we could have like:✩ like:✩✩ as urls :-) Even like:☹
> - one would like to link and add a category to it.
> (is like:✩ just a type of tag then?)
> - A like is similar to a comment that one writes - as comments are written out now. That is the like resides on the blog server itself.
> + everybody who arrives on the site can see this, and use it
> + it is well understood and the same mechanism can be used for comments
> + the blog user has immediate feedback, which is the problem we wanted to solve
> + one can gain new friends as one discovers other people who like the same thing
> + the blog writer can give the "liker" some privileges, given that he has his webid
> - unless one adds this to one's feed of likes (identi.ca, faviki, smob) the value of that like will be a lot less, as it won't bring in new readers
> - A like event is something one writes to one's own server and one then pings the
> blog post. One can do this as faviki and identica do it, by using a button that is part of one's web browser or blog client
> + This allows one to keep track of one's likes
> + other onlookers of the blog post can verify a like by verifying the linked like (but this may not in practice work out as being that easy to verify)
> + one's followers can read one's "Like" feed and then read the post
> - one needs one's own server space (or hosted one)
> This is similar then to the xmlrpc ping-back mechanism then.
> note: I would not make a link to each "like" directly part of one's foaf (one relation per like). If we put all our info in our foaf then we will have huge foaf file. Better think of likes as a very specialised feed. One could have a blog of one's "likes": ie a feed of things one likes, dislikes, etc.... These would be very similar in a way to microblogs such as twitter.
> "Like", bookmarks, and microblogs:
> In fact a "like" is not that different from a bookmark, which is why the idea of a bookmarklet seems so natural: identi.ca and faviki have their bookmarklets. In turn a tweet is not that different from the above: it is represented as an Atom Entry (or used to be). So one can think of a "like" as a special type of atom entry/SIOC post.
> So in that case the question then becomes: what gets pinged? Sindice? DataSpaces? Faviki? deli.cio.us? Well once one writes a blog one can ping anything. But we were mostly interested in pinging the writer of the blog to show our support for his post, though he in turn is going to be more please if this ping also pings your friend network, as that gets him more readers.
> So given the above reasoning we can in fact see that the "like" buttons already exist on the blogosphere today. We have both comment mechanisms, pingbacks and bookmarking services. So all that foaf+ssl adds, is make the commenting authentication much simpler, more reliable and allow the author to more easily connect with the pinger.
> What we need to see is what further advantages a foaf+ssl authentication can bring that brings at low cost a network scale improvement to blogging.
> On 25 May 2010, at 19:03, Story Henry wrote:
>> During a discussion on the Social Web with Amaru Rance in Oxford, we came across the idea of a new very simple usage of foaf+ssl: a like button.
>> The conversation was turning around how on open blogs one could, but tends not to, leave feedback about having liked reading it. It is a major part of the FB experience, and in Twitter land one has the same with RT. This lack of intentional feedback, makes blogs a lot less personal than they need be.
>> Adding it to blogs would be really simple. A foaf+ssl enabled like button, could allow one to authentify oneself in one easy click. The blog post could then simply add a like relation to the user's WebId to the RSS feed. This could of course also be used for comments. A good blog engine could then parse the foaf and use the information gleaned there to produce a profile of the liking person: what blogs he had (verified if the blog points back to the webid), what SN he is on, etc....
>> Currently it would require leaving a comment with an easily falsifiable email address. Asking people to authentify via email would be cumbersome for something as easy as liking something.
>> Just a thought,
I see Annotations as Bookmarks with benefits. I also see enhanced
Bookmarks++ as expressible in structured form via triples. The act of
pinging should result in:
1. Statements (triples) in the Data Space (a collection Named Graphs at
a location on a Network) of the Pinger
2. A ping notice sent to the Pingee
3. Pingee generating Statements (optionally) in his/her Data Space.
All of these statement ultimately add density to public, private, and
semi-private (ACL controlled) Linked Data Spaces. Basically, we simply
accelerate the self-annotation aspect of the Web of Linked Data, by
embracing and extending a broadly used interaction i.e., pinger URLs.
The data created from these activities doesn't need to be in a FOAF
document, it simply needs to be de-referencable via WebIDs associated
with a FOAF profile. This is basically where SIOC comes into play since
it actually models Data Spaces (which goes beyond the initial goal of
online forums modeling).
In ODS you have SIOC based Data Spaces (which have their own URIs) for a
myriad of data item collections: Calendars, Blogs, Discussions,
Bookmarks, Feeds, Bookmarks etc., each is associated with your ODS
-- a cache from one of my data spaces (note: MyOpenLink instance is in
transit at the current time)
-- an old presentation about Linked Data Spaces .
>> Social Web Architect
> foaf-protocols mailing list
> foaf-protocols at lists.foaf-project.org
President & CEO
More information about the foaf-protocols