xfn-brainstorming: Difference between revisions

From Microformats Wiki
Jump to navigation Jump to search
(moved Frames section to be a subsection of Identity Consolidation which it is a part of. added notes about workaround frame elements' lack of rel attribute.)
m (Replace <entry-title> with {{DISPLAYTITLE:}})
 
(62 intermediate revisions by 17 users not shown)
Line 1: Line 1:
<h1> XFN Brainstorming </h1>
{{DISPLAYTITLE: XFN Brainstorming }}
{{TOC-right}}
This page is for brainstorming about various uses and details of [[XFN]], as well as collecting input for potential extensions.
This page is for brainstorming about various uses and details of [[XFN]], as well as collecting input for potential extensions.
__TOC__


== Required Reading ==
== Required Reading ==
Before participating in any XFN brainstorming please read and understand the following:
Before participating in any XFN brainstorming please read and understand the following:
* [http://gmpg.org/xfn/intro XFN intro]
* [http://gmpg.org/xfn/intro XFN intro]
Line 11: Line 9:
* [[xfn-faq|XFN FAQ]]  
* [[xfn-faq|XFN FAQ]]  
* [http://gmpg.org/xfn/and XFN and other services]
* [http://gmpg.org/xfn/and XFN and other services]
* [[rel-faq]]
Note that all existing XFN values were based on research that showed real world sites that indicated such relationships explicitly via text and hyperlinks to other sites.  Thus any new semantics or values will be more seriously considered if URLs demonstrating existing text labeling and hyperlinking behavior are provided.


== Contributors ==
== Contributors ==
Line 16: Line 17:
* [[User:CiaranMc |CiaranMc ]]
* [[User:CiaranMc |CiaranMc ]]
* [http://factoryjoe.com/ Chris Messina]
* [http://factoryjoe.com/ Chris Messina]
* [http://kevinmarks.com/ Kevin Marks]


== general ==
* [[xfn-wants]]


== Identity Consolidation ==
== Identity Consolidation ==
See [[rel-me]] and [[identity-consolidation]]. <span id="Frames">&nbsp;</span>


First, read http://gmpg.org/xfn/and/
=== Indicating non-identity ===
I think we need a way to indicate that another page should not be consolidated into your identity. - KevinMarks


Capturing some Q&A and thoughts here before I clean it up for the FAQ and/or a separate [[identity-consolidation]] page (or perhaps a [[rel-me]] page that discusses identity consolidation).
==== common name disambiguation ====
If you have a common name, creating a disambiguation page or pages to indicate which ones aren't you is useful for indexers and people alike. In fact there are examples of people already creating such a page and linking to pages that are not them. Real world use cases that would benefit:
* Kevin Kelly (the former Wired editor) has put together a page to indicate [http://kevinkelly.name|other people also called "Kevin Kelly"].
* [http://vadim.com Vadim Akslerod] has created a page of [http://vadim.com/vadim/ Vadims on the Net] that link to other people called Vadim.


2007-08-17 Q's are paraphrased from [[User:JosephSmarr]] of Plaxo, and A's are written/edited by [[User:Tantek]].
==== accidental or malicious rel-me links ====
If someone accidentally or maliciously links to one of my pages with rel="me" it would be good to be able to actually deny the connection, rather than just passively not link back. See [http://socialgraph-resources.googlecode.com/svn/trunk/samples/findyours.html?q=kevinmarks.com | the social graph API results for kevinmarks.com] for examples. - KevinMarks


'''Q:''' do the rel="me" links need to be in both directions to verify the link?  seems like they do, since otherwise someone could find anything that links to them and "claim" it just by linking to back to it with rel="me".
This appears to be more of a theoretical use case (as compared to the common name disambiguation) because the real world use ''is'' to simply '''not''' link back, and I don't think we should encourage people to add links to things that are not them, nor are they likely to, especially if it is a spammer/stalker/griefer that is linking to them. - [[User:Tantek|Tantek]]


A: Yes, in general rel="me" links need to be in both directions for exactly that reason.
==== notme proposal ====
initial proposal: <code>rel="notme"</code>, though very open to better suggestions.


'''Q:''' But some sites that let you list your homepage on the profile don't use rel="me", so do we have to just get them all to use it before bi-directional claims will work right?
== Extending family relationships ==


A: Not necessarily.  Of course we prefer the [[advocacy]] path to get them to implement rel="me", but for old sites, as documented on http://gmpg.org/xfn/and/ we can check that specific fields on the the profile page are filled in accordingly, with site specific heurstics.
=== grandparent ===
<cite>[http://www.gmpg.org/xfn/background The XFN: Background page]</cite> says:
<blockquote>
We considered adding "grandparent," but in the end dropped the term because it seemed unlikely to be used in the near future. It may appear in future versions of XFN.
</blockquote>


'''Q:''' So either they need to use rel="me" or we can scrape known sites and trust the link anyway?
Presumably the assumption is that a grandparent won't have a website, but:
<div class="discussion">
* I think it's important in some situations to capture the age difference in a relationship in a way that rel="kin" doesn't seem to.
** Additional XFN rel values is the wrong way to represent age information (even relative). Instead, mark up each person's page with an hCard for them that has a <code>bday</code> property, with even just the year if you want - that could then be used to determine an approximate age difference, which is presumably all that is desired. [[User:Tantek|Tantek]]
* ''rel=kin'' with an age difference of, say, 45 years, might represent any of parent, grandparent, uncle, sibling, cousin or more.
** Again, age (and thus differences) should be represented by use of the [[hCard]] 'bday' property, not rel. [[User:Tantek|Tantek]] 19:27, 15 June 2009 (UTC)
</div>
 
==== grandparent examples ====
If people actually find and document non-trivial examples of links to grandparent sites and this section gets too big, perhaps we can move it to [[grandparent-examples]].


A: rel="me" is the standard that scales (so new players "just work") and for "old players" we write a white-list with compat rules to make it work.
To date, no real world examples have been documented of URLs of grandchildren linking to their grandparents, thus, per the microformats [[process]], we should not complicate a format for a ''theoretical'' need.  


'''Q:''' So i guess the idea is i can't insert a rel="me" link into any user-generated content, like comments on someone else's blog?
Anyone that cares to pursue this may find some real world examples to document in the following web searches (note that search links themselves are not examples, but merely a step towards finding real world examples which still need to be individually analyzed, checked against being false positives etc.)
* [http://www.google.co.uk/search?&q=%22my+grandpa%27s+website%22 search for "my grandpa's website"]
* [http://www.google.co.uk/search?&q=%22my+grandfather%27s+website%22 search for "my grandfather's website"]
* [http://www.google.co.uk/search?&q=%22my+grandma%27s+website%22 search for "my grandma's website"]
* [http://www.google.co.uk/search?&q=%22my+grandmother%27s+website%22 search for "my grandmother's website"].


A: Right, and [[rel-nofollow]] (and [[vote-links]] "vote-against" or "vote-abstain" for that matter) should nullify the rel="me".
See also [[genealogy-examples]].


'''Q:''' What about Yelp which uses rel="nofollow" to your home page link?
=== ancestor descendant ===
Even if they don't maintain it themselves (or are deceased) there might be a URL that does a good job of representing a person.


A: Well the way around it is to *only* look for that specific "home page" link, not any link on the page, that's the key for old players, with the assumption being that *only* the user/owner of that profile could change that URL.  In addition Yelp is actually violating the [[rel-nofollow]] spec because that's not a third party link, that's a first party link, by the owner of that user profile, and therefore it MUST NOT have rel="nofollow" on it.  This bug should be reported to them.
Consider a site about one's family tree.  It might have something like:
<source lang="html4strict">
I can trace my family back to <a href="http://en.wikipedia.org/wiki/William_I_of_England">William the conqueror.</a>
</source>
In these cases it would be useful to have <code>rel="ancestor"</code>, and perhaps a corresponding <code>rel="descendant"</code>.
-CiaranMc


'''Q:''' Is two-way links plus transitive closure sufficient?  Because many sites may only link to your homepage which then links out to many other sites and you'd like to be able to "reel those in"?
See also: [[genealogy-brainstorming#Relationships]]


A: 2-way links plus transitive closure is a good start. But there are common cases where you'll have 3-step triangle circuits you need to detect. For instance, say my Plaxo profile page is joseph.myplaxo.com and I want to add my twitter page twitter.com/jsmarr. My twitter page only links to my home page josephsmarr.com, but that page links back to twitter and also to my plaxo profile. So you can prove I'm authoritative for twitter.com/jsmarr even though it doesn't have a two-way link with joseph.myplaxo.com. There may be even more complex cases, but I think the 3-way is common because many sites only let you have one URL link, which will usually be to your home page, so unless you start by telling a site your home page, you will have to crawl from a "spoke" into the "hub" and then back out again. So in general, you'll need to keep all the rel="me" links on all the pages you crawl, then assemble the graph, then detect all the circuits, and then all the nodes in circuits that are connected to the root page you're starting with are verified. I think. :)
* rel="kin" can be considered sufficient until at least some number of non-trivial real world examples are provided.


'''Q:''' How should I crawl rel="me" links then?
==== ancestor descendant examples ====
If this section gets too big, perhaps we can move it to [[ancestor-descendant-examples]].
* Wikipedia examples
** [http://en.wikipedia.org/wiki/Winston_Churchill#Family_and_early_life Winston Churchill, with list if ancestors]
** [http://en.wikipedia.org/wiki/Elizabeth_II_of_the_United_Kingdom#Marriage Queen Elizabeth II]: refers to "Queen Victoria as [her] great-great-grandmother".
* See also [[genealogy-examples]].


A: Do each 2-way one at a time.  e.g. go to a rel="me" destination, look for rel="me" link back to the same page in that source, and *then* enqueue all the remaining rel="me" links for crawling. Enqueue rel="me" relations as you crawl, e.g. you crawl a, you don't enqueue just the destinations of links  b and c, but rather, enqueue the relations a-me-&gt;b, a-me-&gt;c.  And then you crawl the destinations in the queue, and for confirmed rel="me" 2ways, just move those to another list, e.g. when you see b-me-&gt;a you just remove a-me-&gt;b from the queue and put a&lt;-&gt;b into the "me" file, and when you see b-me-&gt;d you just add it to the queue.  Repeat until you have crawled all the destinations in the queue and you're done.
So far only a couple of (representative) real world examples (from just one site, Wikipedia) have been provided, thus, per the microformats [[process]], we should not complicate a format for a need clearly outside of the 80/20.


'''Q:''' There are often multiple equivalent pages, like http://flickr.com/photos/jsmarr , http://www.flickr.com/photos/jsmarr , http://flickr.com/people/jsmarr , http://flickr.com/people/jsmarr/profile Do we need to write equivalence rules or just make people use the same form?
== Simple Groups and Members ==
A very simple extension to XFN could enable decentralized group membership. E.g.
* rel="group". A person could link from their page to the page of a group that they belong to (e.g. their company, school, DJ collective etc.) and thus assert that they belong to that group.
* rel="member"The web pages of groups could link to who they consider members and thus assert that person's membership.


A: Such pages should a) rel="me" link to both "www." and non "www." versions themselves either via the links already on the page (as they often already have), OR add equivalent &lt;link rel="me" href="..." /&gt; tags to the &lt;head&gt; of the document.
The relationships could be required bi-directional in order to confirm group membership, that is, both the individual must link to the group with rel="group" and the group must link to the individiual with rel="member" in order for the membership to be considered "true".


'''Q:''' So when crawling a page for rel="me" links, should I look for BOTH &lt;a rel="me" links in the body AND &lt;link rel="me" links in the head?
See also [[group-brainstorming]] for more on this including documentation of examples.


A: Yes, they're equivalent, so look for both.
== fans and followers ==
It's becoming a common aspect of many social networks (see list below) that you have unreciprocated and non-friend-based connections to people. These connections are particularly noteworthy where you don't really ever expect to have your connections reciprocated, but instead are similar to the idea of "rel-muse".  


Related: [[hcard-supporting-user-profiles]], [[hcard-xfn-supporting-friends-lists]].
=== follower and following ===
The brainstorming on fans and followers and inverses thereof has appeared to settle (for many months - over a year(?)) on "follower" and "following" as mutual inverse relationship terms.


=== Frames ===
Current summary:


Strangely the new [http://www.google.com/s2/sharing/stuff Google Share] site supports [[hcard|hCard]] but in a frame. In order to parse this page, crawlers need some hook to identify the source for the hcard. It is reasonable to consider using <code><frame src="http://example.com/framesrc.html" rel="me" /></code> (as well as the same for <code><iframe></code> and <code><object></code>) in order to accommodate this unconventional source of profile data. Likewise linking back to the page containing the frame using rel-me would be necessary to produce a valid claim.
* rel="follower" - links to someone who is a follower, that is they are following you.
* rel="following" - links to someone who is one of your "followings", i.e. someone you are following.


Unfortunatly none of <code><frame></code>, <code><iframe></code>, <code><object></code> have a <code>rel</code> attribute in HTML4, and thus we must make use of an alternate solution.
Summary discussion:
<div class="discussion">
* I'd say they're ready for experimentation on websites in the wild. Once we have examples in the wild, we can promote this to a draft of [[rel-follower]]. [[User:Tantek|Tantek]] 00:13, 21 October 2010 (UTC)
* We've had consensus on this for a while now - consider this an invitation for anyone to start stubbing out minimal descriptions for [[rel-follower]] and [[rel-following]], and then we can add them as brainstorming values as well as register them for use in HTML5 in the [[existing-rel-values]] page. [[User:Tantek|Tantek]] 16:55, 30 June 2011 (UTC)
</div>


HTML4 does have a <code>[http://www.w3.org/TR/html4/present/frames.html#h-16.4.1 <noframes>]</code> element which web authors often use to hyperlink to the frame contents so they are accessible in browsers that don't support frames (like most of those on mobile phones for example).
FAQ:
<div class="discussion">
* Q: Does <code>rel=following</code> mean that I read all their content daily/hourly/in real time?[http://indiewebcamp.com/irc/2013-08-28/line/1377706665]
** A: <code>rel=following</code> just means reading some amount, no implications about timeliness, nor comprehensiveness.[http://indiewebcamp.com/irc/2013-08-28/line/1377706765]
</div>


That hyperlink inside the <code><noframes></code> element should have <code>rel="me"</code> on it in order to consolidate the framed portion of the profile with the profile page itself.  Similarly, on the framed portion, there should be a <code><noframes></code> element that links back to the parent frame/document with <code>rel="me"</code> in order to establish the bidirectional identity consolidation relationship.
See below for the analysis and brainstorming that led to this outcome and for specific discussions.


For <code><object></code> embeds, one can simply put an inline hyperlink inside the <code><object>...</object></code> contents as fallback content for browsers that don't render embedded objects, thus providing both a visible method to browse to the embed, and a hyperlink on which to place the <code>rel="me"</code>.
=== fan follower examples ===
Sites that have the semantic / implied-schema of "fan" or "follower")
* http://crowdvine.com/ (has fans and mutual fans and want to meet)
* http://twitter.com/ (has followers vs friends)
* http://upcoming.org/ (has fans vs friends)
* http://www.sugarloot.com (has Fans vs Favorites)


==== IndieWeb ====
* http://hmans.io/following


== Extending family relationships ==
=== possible inverse of fan terms ===
<cite>[http://www.gmpg.org/xfn/background The XFN: Background page]</cite> says:
Brainstorm list of possible terms (for an inverse of fan/follower):
<blockquote>
<div class="discussion">
We considered adding "grandparent," but in the end dropped the term because it seemed unlikely to be used in the near future. It may appear in future versions of XFN.
* source
</blockquote>
** +1 I like this as a neutral term, <s>except that it would be better kept as a rel value for atom:source in [[hAtom]]</s> (on re-reading the Atom spec, atom:source should be a class on a containing element) [[User:Kevin Marks|Kevin Marks]]
* influence
* influencer
** 0 [[User:Tantek|Tantek]] - I like this one, it seems to represent the implied semantics.  If I follow someone, then they are an "influencer" to me. However, the point made by [[User:Donohoe]] below is a good one thus I've changed my opinion from +1 to +0 on this option.
** -1 Sounds like "influenza", and just because I follow someone, doesn't mean they influence me. [[User:Gazza|Gazza]]
** -1 - influence can be negative as well as positive: "The child only shoplifted because he was under the influence of older boys". [[User:AndyMabbett|Andy Mabbett]] 01:04, 26 Jan 2008 (PST)
*** -1 Argument from theoretical example. [[User:Tantek|Tantek]] 21:46, 15 June 2009 (UTC)
** -1 This seems too strong; the relationship is not meant to imply an endorsement; we have [[vote-links]] for that [[User:Kevin Marks|Kevin Marks]] Donohoe's reasoning below is good and I prefer 'following'
** -1 This value is assigned one of 2 ways; arbitrarily by app, or by the user. I don't trust an app to make a decision on what constitutes an ''influencer'' and I don't see a user choosing that in its current wording --[[User:Donohoe|Donohoe]] 19:55, 5 Aug 2008 (PDT)
*** I tend to agree with [[User:Donohoe|Donohoe]]'s reasoning. [[User:Tantek|Tantek]] 21:46, 15 June 2009 (UTC)
* leader
* star
* hero
* favorite
* guider
* role-model
* idol
** -1 [[User:Tantek|Tantek]] These all seem too strong and sometimes imply more subordination than most people want.
* subscription
** +1 Tony Stubblebine
** +1 this one seems most neutral to me (a good thing) [[User:STHayden|STHayden]] 08:35, 3 Mar 2008 (PST)
** -1 this makes sense in a feed-reader context, but not necessarily in a blogroll one; it implies the content rather than the person [[User:Kevin Marks|Kevin Marks]] 14:30, 24 Mar 2008 (PDT)
** +1 Was neutral, now in favor --[[User:Donohoe|Donohoe]] 19:11, 6 Aug 2008 (PDT)
** -1 against for same reasons as Kevin Marks. And "subscription" doesn't sound like a term for a person. --  [[User:Tantek|Tantek]]
* of-interest
* interest
* followee
* focus
* <span id="following">following</span> - as the singular of "followings". If someone is one of your "followings", then you are following that someone, as opposed to if someone is one of your "followers", then that someone is following you! It may sound a bit clumsy/awkward as a term at first, but I think that's actually a sign of novel usage, which has some appeal because then it means the term may be available for us to fill it with this meaning - or I should say, amplify this meaning as it [http://search.twitter.com/search?q=followings has been given by existing Twitter users].
** +1 [[User:Tantek|Tantek]] '''following''' is the best term I have seen for the inverse of fan/follower, and paired with '''follower''', provides a good enough complementary pair of relationships to express both directions.
** Note that Google has now [http://buzz.blogger.com/2008/08/show-off-your-followers.html adopted the "follow" terminology]. -- [http://factoryjoe.com/ Chris Messina]
** +1 'following'is widely adopted. Lets just decide on this and get it implemented.
** +1 I think 'following' is now well understood thanks to both Twitter and Blogger using it, and worth codifying here [[User:Kevin Marks|Kevin Marks]]
* follows
** proposed by [http://schema.org/Person unknown authors at schema.org] I assume this meant to be a verb (like 'knows' ) rather than a plural adjective.
</div>


Presumably the assumption is that a grandparent won't have a website, but:
=== rejected fan follower terms ===
==== rel-fan ====
<div class="discussion">
I would propose adding "rel-fan" or "rel-follower" to the collection of XFN values -- as being something like a "contact" or a "muse" but having a different purpose within the realm of social networking. Again, given that this is showing up in social networks like Pownce (fan), Twitter (follower), and that these words are becoming common, I wonder if it wouldn't make sense to even ditch rel-muse in favor of rel-follower and rel-fan (the former implying some kind of positive social stalking and the latter a kind of amiable appreciation for someone's work).


* I think it's important in some situations to capture the age difference in a relationship in a way that @rel="kin" doesn't seem to.
This fan/follower designation seems sorely lacking from Flickr where not everyone falls into either contact, friend or family... but in many cases, you just like someone's photos and want to be able to check in on them every now and again, similar to the way that people "follow" or "subscribe" to blog feeds... Someone who reads my blog feed could be considered a "follower" -- as in, "someone who follows my blog".
* Even if they don't maintain it themselves (or are deceased) there might be a URL that does a good job of representing a person.


Consider a site about one's family tree.  It might have something like:
- Chris Messina


I can trace my family back to  
Chris, rel-fan or rel-follower would work for you to point to people who follow you, but the converse case is probably more useful, saying who you follow (eg blogroll case too). We need a good noun for that relationship that describes how you view them, that is less coloured than 'muse' currently is by being classified as romantic. rel-source or rel-influence maybe? Very hard to come up with a good noun.  
<a href="http://en.wikipedia.org/wiki/William_I_of_England/">
  William the conqueror.
</a>


In these cases it would be useful to have @rel="ancestor", and perhaps a corresponding @rel="descendant".
[[User:Kevin Marks|Kevin Marks]] 12:57, 17 Aug 2007 (PDT)
-CiaranMc


== Simple Groups and Members ==
* Hence the existence of the HTML "rev" attribute. [[User:TobyInk|TobyInk]] 00:12, 25 Feb 2008 (PST)
** The [[rel-faq#Should_rev_even_be_used|use of rev has been deprecated]] since authors nearly always get it wrong. [[User:Tantek|Tantek]] 10:32, 6 Aug 2008 (PDT)
*** I'm not suggesting that we create a new microformat using rev — I'm pointing out that if we define <code>rel=fan</code> (for linking to your fans) then <code>rev=fan</code> "automatically pops into meaning". (Besides which the usually quoted evidence of rev being "confusing" to authors (i.e. Google's analysis) is flaky at best.) [[User:TobyInk|TobyInk]] 15:26, 6 Aug 2008 (PDT)
**** Suggesting the use of <code>rev=fan</code> (as you just did) *is* creating a new microformat using rev, so yes, you are suggesting it.  And Google's analysis backed up many years of my (and others') personal anecdotal experience with web developers where &gt;90% did not understand the difference between rel and rev. [[User:Tantek|Tantek]] 17:53, 6 Aug 2008 (PDT)
***** My point is that as per the [http://www.w3.org/TR/REC-html40/struct/links.html#adef-rev HTML 4 definition of <code>rev</code>] any time anyone defines a new link type <code>rel=''X''</code>, the meaning of <code>rev=''X''</code> becomes defined too. This is unavoidable. We should define new link types in terms of <code>rel</code> rather than <code>rev</code> for clarity, but we can't prevent people from taking advantage of the facilities inherent in (X)HTML to mark up the inverse meaning. (e.g. [http://changelog.ca/log/2005/09/12/proposed-microformats-for-reputation-and-trust-metrics <code>rev=tag</code> mention], [http://singpolyma.net/2007/05/the-state-of-distributed-social-networking/ and again], [http://styrheim.weblogg.no/311005201154_revtags.html and again]) [[User:TobyInk|TobyInk]] 01:17, 7 Aug 2008 (PDT)
****** We can avoid use of 'rev' by (1) specifically recommending against use of it as the [[rel-faq#Should_rev_even_be_used|rel FAQ does]], and (2) by providing an inverse term that can be used with 'rel' so there is no need to use 'rev'. [[User:Tantek|Tantek]] 19:27, 15 June 2009 (UTC)
</div>


A very simple extension to XFN could enable decentralized group membership. E.g.
==== rel-reader ====
* rel="group".  A person could link from their page to the page of a group that they belong to (e.g. their company, school, DJ collective etc.) and thus assert that they belong to that group.
For blogrolls at least, what about rel-read / rel-reader? -- [[User:SteveIvy|Steve Ivy]]
* rel="member".  The web pages of groups could link to who they consider members and thus assert that person's membership.


The relationships could be required bi-directional in order to confirm group membership, that is, both the individual must link to the group with rel="group" and the group must link to the individiual with rel="member" in order for the membership to be considered "true".
Steve, "read" is still a verb (and [[rel-values]] need to be nouns), and "reader" would still imply that they are a "reader" of yours, rather than vice versa. -- [[User:Tantek|Tantek]]


See also [[group-brainstorming]].
How about "rel-reader" instead of "rel-follower" and "rel-reading" instead of "rel-following" ? It has the more 'neutral' and 'real world' connotation of actually reading and not 'following' people around. Also refers to the traditional word for such things in the book-world as in "Neil Gaiman's readers" and "I am reading Neil Gaiman" -- [[User:KP|Kartik]]


-Tantek
On further thought and discussion with [[User:Tantek|Tantek]] (see: [http://logs.glob.uno/?c=freenode%23microformats&s=1+Jul+2014&e=1+Jul+2014#c72761 IRC logs reference] ), reader/reading denote textual content which seems too narrow for the Web.-- [[User:KP|Kartik]]


== Fans and followers ==
==== rel-subscriber ====
What about ''subscriber'' instead - also very neutral and more akin to others in the list? --[[User:Donohoe|Donohoe]] 19:55, 5 Aug 2008 (PDT)


It's becoming a common aspect of many social networks that you have unreciprocated and non-friend-based connections to people. These connections are particularly noteworthy where you don't really ever expect to have your connections reciprocated, but instead are similar to the idea of "rel-muse".  
"subscriber" is no different from "reader" in this respect, and means the *opposite* of what we are looking for.  I.e. a rel="subscriber" link to someone would mean that someone is a subscriber of my content/feed. -- [[User:Tantek|Tantek]]


I would propose adding "rel-fan" or "rel-follower" to the collection of XFN values -- as being something like a "contact" or a "muse" but having a different purpose within the realm of social networking. Again, given that this is showing up in social networks like Pownce (fan), Twitter (follower), and that these words are becoming common, I wonder if it wouldn't make sense to even ditch rel-muse in favor of rel-follower and rel-fan (the former implying some kind of positive social stalking and the latter a kind of amiable appreciation for someone's work).  
=== past examples ===
Sites that previously provided real world examples of fans vs friends:
* http://pownce.com/ (has fans vs friends) - until 2008-12-15 when the site moved to read-only mode.


This fan/follower designation seems sorely lacking from Flickr where everyone falls into either contact, friend or family... but in many cases, you just like someone's photos and want to be able to check in on them every now and again, similar to the way that people "follow" or "subscribe" to blog feeds... Someone who reads my blog feed could be considered a "follower" -- as in, "someone who follows my blog".
== favorites ==
Another possibility is perhaps 'favorite', that is, people do link to favorite bands for example, which is similar to saying they are a fan of the band. Thus you could add <code>rel="favorite"</code> to such hyperlinks to indicate that that music band over there is a favorite of yours. [[User:Tantek|Tantek]] 17:57, 2 Oct 2007 (PDT) based on a question raised by DanBri.


- Chris Messina
== mentors and mentees ==
Though seemingly rare, I personally have found use for <code>rel="mentor"</code> and inverse <code>rel="mentee"</code> (see Wiktionary definitions: [http://en.wiktionary.org/wiki/mentor mentor], [http://en.wiktionary.org/wiki/mentee mentee]).  I don't have sufficient evidence to even consider proposing adding these to XFN, but I wanted to capture them here as a brainstorm while I look into using them personally and research examples in the wild.  I may just use them as [[POSH]] myself.


Examples (sites that have the concept of "fan" or "follower")
* http://twitter.com/ (has followers vs friends)
* http://upcoming.org/ (has fans vs friends)
* http://pownce.com/ (has fans vs friends)
[[User:Tantek|Tantek]] 02:34, 8 Jul 2007 (PDT)
[[User:Tantek|Tantek]] 02:34, 8 Jul 2007 (PDT)


Chris, rel-fan or re-follower would work for you to point to people who follow you, but the converse case is probably more useful, saying who you follow (eg blogroll case too). We need a good noun for that relationship that describes how you view them, that is less coloured than 'muse' currently is by being classified as romantic. rel-source or rel-influence maybe? Very hard to come up with a good noun. [[User:Kevin Marks|Kevin Marks]] 12:57, 17 Aug 2007 (PDT)
* +1 Erica OGrady
* +1 needed by [https://github.com/open-app/core Enspiral] in their [http://www.enspiral.com/ Open App Ecosystem] as explained in [https://github.com/hackers4peace/plp-docs/issues/12#issuecomment-78043877 this github issue] &mdash [[User:Elf Pavlik|Elf Pavlik]] 22:49, 3 May 2015 (UTC)
 
== advisor and advisee ==
Similarly, I have seen folks reference someone as an advisor, or note that they are advising someone.  We could consider <code>rel="advisor"</code> and <code>rel="advisee"</code> to capture and represent these semantics.
 
Note that the relationship of advising a company or organization would be much better captured by noting "advisor" as a "role" property value in an hCard listing that organization, e.g.:
<pre><nowiki>
<span class="vcard">
<span class="fn">Tantek Çelik</span>
<span class="org">Citizen Agency</span>
<span class="role">advisor</span>
</span>
</nowiki></pre>
 
Again, I think [[POSH]] usage of these terms would make a good experiment to see if there is sufficient use to formalize them.
 
[[User:Tantek|Tantek]] 07:45, 29 Dec 2007 (PST)
 
== Influence out and influenced in ==
 
Needing to provide more options to describe the flow of influence I've been considering the following relationship terms that fall into two predicate groups, influence out(applied) and influence in(received).
 
Influence out:
* 'follower'
* 'student'
* 'subscriber'
* 'listener'
* 'reader'
* 'viewer'
* 'supporter'
* 'collaborator'
 
Influence in:
* 'inspiration'
* 'favorite'
* 'teacher'
* 'mentor'
* 'adviser'
* 'influence'
* 'source'
* 'collaborator'
 
Inline with xfn convention the terms are all nouns and refer to the far side of the link they are used in.
 
[[User:James Tindall|James Tindall]] 14:23, 31 Aug 2008 (BST)
 
== scholarly ==
A possibility that might be added is scholarly definitions.  For example:
* <code>rel="mentor"</code>[Student's Teacher] (similar to above)
** or perhaps <code>rel="teacher"</code> to indicate the distinct scholarly semantic. - [[User:Tantek|Tantek]]
* <code>rel="student"</code>[Teacher's Student] (inverse of mentor)
* <code>rel="classmate"</code>[Student's classmate]. 
Possible uses in Social Networking sites that involve adding your school, grading your teachers, etc. Teacher &lt;-&gt; Teacher would be specified in with the existing XFN <code>rel="co-worker"</code> or <code>rel="colleague"</code> values.
 
[[User:Navarr|Navarr]] 05:56, 13 Dec 2007 (CST)
 
== professional ==
* [[xpn]] / [[xpn-examples]] - proposal to extend (or replicate) XFN for business (or professional - hence X'''P'''N) relationships
 
== business to business ==
There has been some discussion (on the microformats-discuss mailing list) about wanting to markup business to business relationships.


== Mentors and Mentees ==
While this is out of scope for XFN's person to person links, if a page/site does represent a company (e.g. if the [[representative hCard]] of the page was an organization) and links to other company sites, it could be useful.
Though seemingly rare, I personally have found use for <code>rel="mentor"</code> and inverse <code>rel="mentee"</code> (see Wiktionary definitions: [http://en.wiktionary.org/wiki/mentor mentor], [http://en.wiktionary.org/wiki/mentee mentee]).  I don't have sufficient evidence to even consider proposing adding these to XFN, but I wanted to capture them here as a brainstorm while I look into using them personally and research examples in the wild.


[[User:Tantek|Tantek]] 02:34, 8 Jul 2007 (PDT)
In particular <code>rel</code> values of:
* <code>subsidiary</code>. designates a link to a subsidiary company of a controlling/parent company
* <code>controlling</code>. designates a link to the controlling/parent company of a subsidiary company
* ...
 
Currently this is a theoretical example as no real world example pages of companies linking to other such company pages have been provided, and as such, there is insufficient research/usage/experience to merit consideration as a microformat.
 
Those wishing to experiment with these rel values, should consider doing so as [[POSH]] and document their experience here so that we have some real world examples!
 
[[User:Tantek|Tantek]] 19:27, 15 June 2009 (UTC)
 
== met and wants to meet ==
 
[http://lanyrd.com/ Lanyrd] allows to mark people as "Want to meet". Also some people like to keep track on who they have met. Based on [http://indiewebcamp.com/irc/2015-05-03/line/1430691481869 discussion on #indiewebcamp IRC]
 
[[User:Elf Pavlik|Elf Pavlik]] 23:01, 3 May 2015 (UTC)
 
<div class="discussion">
* Note: [[XFN]] 1.1 already has [[rel-met|rel=met]]. [[User:Tantek|Tantek]] 23:01, 25 August 2015 (UTC)
</div>


== See Also ==
== see also ==
* [[xfn]]
{{xfn-related-pages}}
* [[hcard|hCard]]
* [[hcard|hCard]]
* [[hcard-brainstorming|hCard brainstorming]]
* [[hcard-brainstorming|hCard brainstorming]]
* [[social-network-portability]]
* [http://xmlns.com/foaf/spec/ FOAF Vocabulary]

Latest revision as of 16:35, 18 July 2020

This page is for brainstorming about various uses and details of XFN, as well as collecting input for potential extensions.

Required Reading

Before participating in any XFN brainstorming please read and understand the following:

Note that all existing XFN values were based on research that showed real world sites that indicated such relationships explicitly via text and hyperlinks to other sites. Thus any new semantics or values will be more seriously considered if URLs demonstrating existing text labeling and hyperlinking behavior are provided.

Contributors

general

Identity Consolidation

See rel-me and identity-consolidation.  

Indicating non-identity

I think we need a way to indicate that another page should not be consolidated into your identity. - KevinMarks

common name disambiguation

If you have a common name, creating a disambiguation page or pages to indicate which ones aren't you is useful for indexers and people alike. In fact there are examples of people already creating such a page and linking to pages that are not them. Real world use cases that would benefit:

accidental or malicious rel-me links

If someone accidentally or maliciously links to one of my pages with rel="me" it would be good to be able to actually deny the connection, rather than just passively not link back. See | the social graph API results for kevinmarks.com for examples. - KevinMarks

This appears to be more of a theoretical use case (as compared to the common name disambiguation) because the real world use is to simply not link back, and I don't think we should encourage people to add links to things that are not them, nor are they likely to, especially if it is a spammer/stalker/griefer that is linking to them. - Tantek

notme proposal

initial proposal: rel="notme", though very open to better suggestions.

Extending family relationships

grandparent

The XFN: Background page says:

We considered adding "grandparent," but in the end dropped the term because it seemed unlikely to be used in the near future. It may appear in future versions of XFN.

Presumably the assumption is that a grandparent won't have a website, but:

  • I think it's important in some situations to capture the age difference in a relationship in a way that rel="kin" doesn't seem to.
    • Additional XFN rel values is the wrong way to represent age information (even relative). Instead, mark up each person's page with an hCard for them that has a bday property, with even just the year if you want - that could then be used to determine an approximate age difference, which is presumably all that is desired. Tantek
  • rel=kin with an age difference of, say, 45 years, might represent any of parent, grandparent, uncle, sibling, cousin or more.
    • Again, age (and thus differences) should be represented by use of the hCard 'bday' property, not rel. Tantek 19:27, 15 June 2009 (UTC)

grandparent examples

If people actually find and document non-trivial examples of links to grandparent sites and this section gets too big, perhaps we can move it to grandparent-examples.

To date, no real world examples have been documented of URLs of grandchildren linking to their grandparents, thus, per the microformats process, we should not complicate a format for a theoretical need.

Anyone that cares to pursue this may find some real world examples to document in the following web searches (note that search links themselves are not examples, but merely a step towards finding real world examples which still need to be individually analyzed, checked against being false positives etc.)

See also genealogy-examples.

ancestor descendant

Even if they don't maintain it themselves (or are deceased) there might be a URL that does a good job of representing a person.

Consider a site about one's family tree. It might have something like:

I can trace my family back to <a href="http://en.wikipedia.org/wiki/William_I_of_England">William the conqueror.</a>

In these cases it would be useful to have rel="ancestor", and perhaps a corresponding rel="descendant". -CiaranMc

See also: genealogy-brainstorming#Relationships

  • rel="kin" can be considered sufficient until at least some number of non-trivial real world examples are provided.

ancestor descendant examples

If this section gets too big, perhaps we can move it to ancestor-descendant-examples.

So far only a couple of (representative) real world examples (from just one site, Wikipedia) have been provided, thus, per the microformats process, we should not complicate a format for a need clearly outside of the 80/20.

Simple Groups and Members

A very simple extension to XFN could enable decentralized group membership. E.g.

  • rel="group". A person could link from their page to the page of a group that they belong to (e.g. their company, school, DJ collective etc.) and thus assert that they belong to that group.
  • rel="member". The web pages of groups could link to who they consider members and thus assert that person's membership.

The relationships could be required bi-directional in order to confirm group membership, that is, both the individual must link to the group with rel="group" and the group must link to the individiual with rel="member" in order for the membership to be considered "true".

See also group-brainstorming for more on this including documentation of examples.

fans and followers

It's becoming a common aspect of many social networks (see list below) that you have unreciprocated and non-friend-based connections to people. These connections are particularly noteworthy where you don't really ever expect to have your connections reciprocated, but instead are similar to the idea of "rel-muse".

follower and following

The brainstorming on fans and followers and inverses thereof has appeared to settle (for many months - over a year(?)) on "follower" and "following" as mutual inverse relationship terms.

Current summary:

  • rel="follower" - links to someone who is a follower, that is they are following you.
  • rel="following" - links to someone who is one of your "followings", i.e. someone you are following.

Summary discussion:

  • I'd say they're ready for experimentation on websites in the wild. Once we have examples in the wild, we can promote this to a draft of rel-follower. Tantek 00:13, 21 October 2010 (UTC)
  • We've had consensus on this for a while now - consider this an invitation for anyone to start stubbing out minimal descriptions for rel-follower and rel-following, and then we can add them as brainstorming values as well as register them for use in HTML5 in the existing-rel-values page. Tantek 16:55, 30 June 2011 (UTC)

FAQ:

  • Q: Does rel=following mean that I read all their content daily/hourly/in real time?[1]
    • A: rel=following just means reading some amount, no implications about timeliness, nor comprehensiveness.[2]

See below for the analysis and brainstorming that led to this outcome and for specific discussions.

fan follower examples

Sites that have the semantic / implied-schema of "fan" or "follower")

IndieWeb

possible inverse of fan terms

Brainstorm list of possible terms (for an inverse of fan/follower):

  • source
    • +1 I like this as a neutral term, except that it would be better kept as a rel value for atom:source in hAtom (on re-reading the Atom spec, atom:source should be a class on a containing element) Kevin Marks
  • influence
  • influencer
    • 0 Tantek - I like this one, it seems to represent the implied semantics. If I follow someone, then they are an "influencer" to me. However, the point made by User:Donohoe below is a good one thus I've changed my opinion from +1 to +0 on this option.
    • -1 Sounds like "influenza", and just because I follow someone, doesn't mean they influence me. Gazza
    • -1 - influence can be negative as well as positive: "The child only shoplifted because he was under the influence of older boys". Andy Mabbett 01:04, 26 Jan 2008 (PST)
      • -1 Argument from theoretical example. Tantek 21:46, 15 June 2009 (UTC)
    • -1 This seems too strong; the relationship is not meant to imply an endorsement; we have vote-links for that Kevin Marks Donohoe's reasoning below is good and I prefer 'following'
    • -1 This value is assigned one of 2 ways; arbitrarily by app, or by the user. I don't trust an app to make a decision on what constitutes an influencer and I don't see a user choosing that in its current wording --Donohoe 19:55, 5 Aug 2008 (PDT)
      • I tend to agree with Donohoe's reasoning. Tantek 21:46, 15 June 2009 (UTC)
  • leader
  • star
  • hero
  • favorite
  • guider
  • role-model
  • idol
    • -1 Tantek These all seem too strong and sometimes imply more subordination than most people want.
  • subscription
    • +1 Tony Stubblebine
    • +1 this one seems most neutral to me (a good thing) STHayden 08:35, 3 Mar 2008 (PST)
    • -1 this makes sense in a feed-reader context, but not necessarily in a blogroll one; it implies the content rather than the person Kevin Marks 14:30, 24 Mar 2008 (PDT)
    • +1 Was neutral, now in favor --Donohoe 19:11, 6 Aug 2008 (PDT)
    • -1 against for same reasons as Kevin Marks. And "subscription" doesn't sound like a term for a person. -- Tantek
  • of-interest
  • interest
  • followee
  • focus
  • following - as the singular of "followings". If someone is one of your "followings", then you are following that someone, as opposed to if someone is one of your "followers", then that someone is following you! It may sound a bit clumsy/awkward as a term at first, but I think that's actually a sign of novel usage, which has some appeal because then it means the term may be available for us to fill it with this meaning - or I should say, amplify this meaning as it has been given by existing Twitter users.
    • +1 Tantek following is the best term I have seen for the inverse of fan/follower, and paired with follower, provides a good enough complementary pair of relationships to express both directions.
    • Note that Google has now adopted the "follow" terminology. -- Chris Messina
    • +1 'following'is widely adopted. Lets just decide on this and get it implemented.
    • +1 I think 'following' is now well understood thanks to both Twitter and Blogger using it, and worth codifying here Kevin Marks
  • follows

rejected fan follower terms

rel-fan

I would propose adding "rel-fan" or "rel-follower" to the collection of XFN values -- as being something like a "contact" or a "muse" but having a different purpose within the realm of social networking. Again, given that this is showing up in social networks like Pownce (fan), Twitter (follower), and that these words are becoming common, I wonder if it wouldn't make sense to even ditch rel-muse in favor of rel-follower and rel-fan (the former implying some kind of positive social stalking and the latter a kind of amiable appreciation for someone's work).

This fan/follower designation seems sorely lacking from Flickr where not everyone falls into either contact, friend or family... but in many cases, you just like someone's photos and want to be able to check in on them every now and again, similar to the way that people "follow" or "subscribe" to blog feeds... Someone who reads my blog feed could be considered a "follower" -- as in, "someone who follows my blog".

- Chris Messina

Chris, rel-fan or rel-follower would work for you to point to people who follow you, but the converse case is probably more useful, saying who you follow (eg blogroll case too). We need a good noun for that relationship that describes how you view them, that is less coloured than 'muse' currently is by being classified as romantic. rel-source or rel-influence maybe? Very hard to come up with a good noun.

Kevin Marks 12:57, 17 Aug 2007 (PDT)

  • Hence the existence of the HTML "rev" attribute. TobyInk 00:12, 25 Feb 2008 (PST)
    • The use of rev has been deprecated since authors nearly always get it wrong. Tantek 10:32, 6 Aug 2008 (PDT)
      • I'm not suggesting that we create a new microformat using rev — I'm pointing out that if we define rel=fan (for linking to your fans) then rev=fan "automatically pops into meaning". (Besides which the usually quoted evidence of rev being "confusing" to authors (i.e. Google's analysis) is flaky at best.) TobyInk 15:26, 6 Aug 2008 (PDT)
        • Suggesting the use of rev=fan (as you just did) *is* creating a new microformat using rev, so yes, you are suggesting it. And Google's analysis backed up many years of my (and others') personal anecdotal experience with web developers where >90% did not understand the difference between rel and rev. Tantek 17:53, 6 Aug 2008 (PDT)
          • My point is that as per the HTML 4 definition of rev any time anyone defines a new link type rel=X, the meaning of rev=X becomes defined too. This is unavoidable. We should define new link types in terms of rel rather than rev for clarity, but we can't prevent people from taking advantage of the facilities inherent in (X)HTML to mark up the inverse meaning. (e.g. rev=tag mention, and again, and again) TobyInk 01:17, 7 Aug 2008 (PDT)
            • We can avoid use of 'rev' by (1) specifically recommending against use of it as the rel FAQ does, and (2) by providing an inverse term that can be used with 'rel' so there is no need to use 'rev'. Tantek 19:27, 15 June 2009 (UTC)

rel-reader

For blogrolls at least, what about rel-read / rel-reader? -- Steve Ivy

Steve, "read" is still a verb (and rel-values need to be nouns), and "reader" would still imply that they are a "reader" of yours, rather than vice versa. -- Tantek

How about "rel-reader" instead of "rel-follower" and "rel-reading" instead of "rel-following" ? It has the more 'neutral' and 'real world' connotation of actually reading and not 'following' people around. Also refers to the traditional word for such things in the book-world as in "Neil Gaiman's readers" and "I am reading Neil Gaiman" -- Kartik

On further thought and discussion with Tantek (see: IRC logs reference ), reader/reading denote textual content which seems too narrow for the Web.-- Kartik

rel-subscriber

What about subscriber instead - also very neutral and more akin to others in the list? --Donohoe 19:55, 5 Aug 2008 (PDT)

"subscriber" is no different from "reader" in this respect, and means the *opposite* of what we are looking for. I.e. a rel="subscriber" link to someone would mean that someone is a subscriber of my content/feed. -- Tantek

past examples

Sites that previously provided real world examples of fans vs friends:

  • http://pownce.com/ (has fans vs friends) - until 2008-12-15 when the site moved to read-only mode.

favorites

Another possibility is perhaps 'favorite', that is, people do link to favorite bands for example, which is similar to saying they are a fan of the band. Thus you could add rel="favorite" to such hyperlinks to indicate that that music band over there is a favorite of yours. Tantek 17:57, 2 Oct 2007 (PDT) based on a question raised by DanBri.

mentors and mentees

Though seemingly rare, I personally have found use for rel="mentor" and inverse rel="mentee" (see Wiktionary definitions: mentor, mentee). I don't have sufficient evidence to even consider proposing adding these to XFN, but I wanted to capture them here as a brainstorm while I look into using them personally and research examples in the wild. I may just use them as POSH myself.

Tantek 02:34, 8 Jul 2007 (PDT)

advisor and advisee

Similarly, I have seen folks reference someone as an advisor, or note that they are advising someone. We could consider rel="advisor" and rel="advisee" to capture and represent these semantics.

Note that the relationship of advising a company or organization would be much better captured by noting "advisor" as a "role" property value in an hCard listing that organization, e.g.:

<span class="vcard">
 <span class="fn">Tantek Çelik</span>
 <span class="org">Citizen Agency</span>
 <span class="role">advisor</span>
</span>

Again, I think POSH usage of these terms would make a good experiment to see if there is sufficient use to formalize them.

Tantek 07:45, 29 Dec 2007 (PST)

Influence out and influenced in

Needing to provide more options to describe the flow of influence I've been considering the following relationship terms that fall into two predicate groups, influence out(applied) and influence in(received).

Influence out:

  • 'follower'
  • 'student'
  • 'subscriber'
  • 'listener'
  • 'reader'
  • 'viewer'
  • 'supporter'
  • 'collaborator'

Influence in:

  • 'inspiration'
  • 'favorite'
  • 'teacher'
  • 'mentor'
  • 'adviser'
  • 'influence'
  • 'source'
  • 'collaborator'

Inline with xfn convention the terms are all nouns and refer to the far side of the link they are used in.

James Tindall 14:23, 31 Aug 2008 (BST)

scholarly

A possibility that might be added is scholarly definitions. For example:

  • rel="mentor"[Student's Teacher] (similar to above)
    • or perhaps rel="teacher" to indicate the distinct scholarly semantic. - Tantek
  • rel="student"[Teacher's Student] (inverse of mentor)
  • rel="classmate"[Student's classmate].

Possible uses in Social Networking sites that involve adding your school, grading your teachers, etc. Teacher <-> Teacher would be specified in with the existing XFN rel="co-worker" or rel="colleague" values.

Navarr 05:56, 13 Dec 2007 (CST)

professional

  • xpn / xpn-examples - proposal to extend (or replicate) XFN for business (or professional - hence XPN) relationships

business to business

There has been some discussion (on the microformats-discuss mailing list) about wanting to markup business to business relationships.

While this is out of scope for XFN's person to person links, if a page/site does represent a company (e.g. if the representative hCard of the page was an organization) and links to other company sites, it could be useful.

In particular rel values of:

  • subsidiary. designates a link to a subsidiary company of a controlling/parent company
  • controlling. designates a link to the controlling/parent company of a subsidiary company
  • ...

Currently this is a theoretical example as no real world example pages of companies linking to other such company pages have been provided, and as such, there is insufficient research/usage/experience to merit consideration as a microformat.

Those wishing to experiment with these rel values, should consider doing so as POSH and document their experience here so that we have some real world examples!

Tantek 19:27, 15 June 2009 (UTC)

met and wants to meet

Lanyrd allows to mark people as "Want to meet". Also some people like to keep track on who they have met. Based on discussion on #indiewebcamp IRC

Elf Pavlik 23:01, 3 May 2015 (UTC)

see also