One Person, Many Links
The heart of the issue I wrote about recently comes down to, I think, this question: if I’m writing on the Web and I want to link to a person, how do I do that? What on the Web represents someone? Is it their Facebook page? Their Twitter account? Weblog? LinkedIn?¹
To get into this, I think it’s useful to think about how our personal use of the Web has developed. If I may boldly striate it into eras:
- The original Web, starting with Tim Berners-Lee, through Yahoo!’s human-curated index, Spacejam’s rad website, and so on. People wander out of the darkness of AOL, blink their eyes at the bright future the Web offers, and take to message boards to talk about electronic music, The X-Files, Kirk vs Picard, and how violent video games and movies based on male power fantasies could never wind up having a negative impact on society. Dave Winer carves the first weblog into a stone
- Web 2.0. Weblogs, Flickr, del.icio.us, and the like. People write about George W. Bush, recommend mortgage-backed securities, organize things using tags, and still think comment systems are a pretty neat idea.
- The Social Media Era. LiveJournal gives rise to Friendster, MySpace, then Facebook. People rate dogs on Twitter, evangelize the blockchain, and misunderstood notions about the freedom of speech are put forth to justify white supremacy.
- The Open Social Web. The future… maybe. People do all the stuff they did on Facebook and Twitter, but they’re free to use magic JSON incantations to move from platform to platform whenever the one they’re using gets too full of white supremacists. Users don’t have to agree to forfeit copyrights on whatever they publish to the company running the network. The term “blockchain” comes to conjure similar connotations as “mortgage-backed securities.” (Hey, I said “maybe.”)
Let me broadly sketch out what the Open Social Web era might be, starting with two meanings of the word “open.” First, the Open Social Web is “open” because it’s not owned by any one company. It’s open and not closed. You’re not bound to whatever decisions Facebook or Twitter makes about their platforms. You choose where you write, where you host, what you use, and so on. If a piece of software redesigns and you don’t like it, you take your files and you go elsewhere.
Second, “open” means “not private.” Everything on the Open Social Web is public. If you want to share stuff privately, do it over email or text message or, I don’t know, a phone call I guess. (I actually think that Apple’s iMessage group chats make very good semi-private social networks if you take the step to tap “Hide Alerts” for the thread.) You should always assume everything you write down may become public, and building features that attempt to keep things private means giving your customers false confidence that can get them into trouble.
The “social” part is fairly self-explanatory. You can befriend people and reply to them and Webmentions help see to it that people will actually see your replies, unlike the forgotten comment threads of the 2.0 era. There’s no consensus on if/how to do retweet-style sharing, which I think of as an essential part of Twitter, but it’s more complicated when you talk about the wholesale reposting of copywritten content. I actually have a long piece on the concept of retweets that I wrote a few weeks ago but I’ve been letting it sit while I think it all over some. I’m sure I’ll get back to it soon.
The Open Social Web is true to the Web. If you think about closed networks like Twitter, they’re not truly inter-connected, they’re *intra*nets. Critical to the idea of the Social Web era is that you have to make choices. You have to join Facebook if you want to read people’s Facebook posts. If you’re not on Twitter, you miss out on my tweets. On the Open Social Web, I post on my site, you post on yours, and we can all read each other’s stuff without having to buy into any one network. It’s a web of worldwide, *inter*connected *net*works.
But let’s step back and accept that social media sites do something well: they’re people-focused, not site-focused. The Web 2.0 era was about subscribing to a bunch of RSS feeds in Google Reader, which were a mélange of personal weblogs, links, political blogs, Flickr photos, webcomics, and so on. What Facebook and Twitter have in common is that when you’re looking at someone’s profile, you feel like you are connecting with that person, not a website where that person writes.
Despite this, in the social media era it’s common for someone to be on Facebook, Twitter, Instagram, and Snapchat. Our identities are fragmented because we exist in multiple social circles at once, have varying interests, and so on.² That’s all fine. I started all this a few days ago by talking about the difficulty of linking to someone when they might be on Twitter and micro.blog and elsewhere. The question remains: if I’m all those places, where am I?
The answer is that we’re all of it. I’m my homepage at my domain name and I’m my tweets and I’m my 1940s Superman cartoons. You’re your photos and your podcast and your… whatever Pinterest is. We each get to pick. And there are four magic letters that could tie it all together: OPML.
Behind the scenes of any Open Social Web service, whether it’s something that’s sorta like Twitter, a photo stream sorta like Instagram (but, y’know, not all corporate!), or whatever, is an RSS feed.³ You think you’re writing about how The Last Jedi should have conformed to your misunderstood interpretation of the end of Return of the Jedi, but what you’re really doing is adding an
item to your RSS
channel that your friends will see in their timelines.⁴ So you have your micro.blog, your podcast, your photos, etc., and all of them are backed by their feeds. And how do we organize lists of RSS feeds? OPML.
Now, finally, I come to my proposal: an Open Social Web site should publish an OPML file which it links to in its
head. Customers on their “about me” page list the sites they own (this is my LinkedIn, this is my dating profile, etc), the software verifies ownership via reciprocal microformatted links and grabs the RSS feed for each of those other sites. (More saavy customers could paste in the feeds manually, obviously, but remember that most people don’t know what an RSS feed is and I’d argue a general guideline is that if your design requires them to, you have more work to do.) The site then generates an OPML file listing each of the feeds a particular person is the author of and puts a link to that in its
head. I’d gently suggest something like the following but am open to revisions:
<link rel=“alternate” title=“John Doe’s Feeds” type=“application/xml” href=“[example.com/feeds.opm...](http://example.com/feeds.opml)”>
Someone more knowledgable of these things can weigh in if that’s the right
mime-type. I’m assuming
rel alternate is correct. I insist that the title should be “[name]’s Feeds” instead of “My Feed.”
When you add a friend on an Open Social Web app (RSS reader, Twitter-like RSS reader, podcatcher, etc.), it will look for an OPML file that describes all the sites that person contributes to and, where appropriate, ask if you also want to subscribe to their other sites. A podcast app might see that the OPML file describes other podcasts from the same person and offer those as well. A reporter might list her microblog and also her newspaper’s site where her reporting is published. A cinephile else might have a weblog where he writes about film and also a YouTube feed where he publishes videos. The idea here is that the Open Social Web is people-focused. You’re not subscribing to an XML-formatted file of structured data representing “content,” you’re befriending someone.
The OPML file would be kept up-to-date by whatever services you use, so if you start a new podcast or a new site for a new project, you add it to one of your profile pages and your friends’ apps will notice it (checking once a day or week or whatever seems right) and decide what to do. Maybe you want to automatically subscribe to new sources from your friends, or maybe you get an alert with the option to do so. This is up to the developer, and there’d obviously be a place where you could click on a friend and select/deselect sources if you, say, only want see someone’s photos and not their posts. I think this would start to solve the problem of many sites, one person.
Armed with an app that has knowledge of all the different services you use, the confusion around @usernames I talked about last time becomes a little easier to manage. My advice when referring to someone, then, would be to omit the @ in most cases unless you’re specifically intending to ping the person or replying. Call me David Ely or just davextreme. No need for the silly @.
Returning one last time to my central question of “how do you link to someone on the Web?”, The short answer is that you should find whatever you think is the most “homepage-y” of the available options, but a properly-structured Web will let you find their other stuff easily. What on the Web represents a person? It’s the collection of feeds he/she authors. If the thesis of that piece was “using these @ symbols is complex when you talk about multiple services,” my point today is, “your software should be solving this complexity for you.” Let RSS, Webmention, Websub, and OPML do all the work and just go out there and make friends. <3
Shh! Here’s a secret: there is an answer to this, but it’s no good. Ready? Here it is: *a person is their email address*. This answer’s no good because of spam and privacy, but the email address *is* the piece of glue that ties everything together. We can’t actually build anything around email addresses, and in certain venues it’s the dreaded *personally-identifiable information*, so now I have to spend several paragraphs thinking up something else.
And since I'm tossing out infeasible ideas, here's another one: each government should provide its residents with an official homepage. This page would be a simple profile that the user could keep updated with links to other places they can be found online. It’d also be the keysyone for official identify verification using appropriate cryptographic doohickeys instead of making everyone pretend that Social Security Numbers are that.
I'm talking about *public* identity here. Obviously we'd all still be free to make pseudonymous accounts where that makes sense. But right now we more or less use Google to handle the question of authority. Wikipedia, I guess, if you're famous. It makes more sense to me for there to be an official, verified place to say, "this is me." ↩︎
Confession: for argumentative purposes I'm knowlingly setting aside the biggest reason we fragment our online personæ, which is that we act differently on Facebook where our parents are watching than we do with our friends on Snapchat. Some people use Twitter for work purposes only. Some people have gender identities they don't express everywhere. It's useful to spread ourselves across different services because we're all varied, complicated people. ↩︎
I say "RSS feed" but I really mean JSON Feed with RSS as a backup. I know this ia holy war for some, though, and I preemptively regret writing this footnote lest it derail the rest of my points in the article. ↩︎
I have another piece in mind to write about Open Social Web software. Short version: we need a new generation of RSS reading software that behaves more like the timelines of Facebook or Twitter. It needs to know how to format shorter posts for tweet-like readability and have a "reply" button that knows where to send the reply. And it all needs to be easy enough to use that customers don't ever need to really know what an RSS feed is. They just follow their friends and the posts show up, in the same way that podcast apps are RSS readers but they just feel like you're subscribing to a show. ↩︎