End of Comments II: Browsers and Identity

2011-02-17

I think it’s pertinent to note first that my thoughts on this thread and some others comes from a place of deep interest in technology, but basic technical laity. I have very little understanding of what’s under the floorboards of the internet. I mostly only see the levers and buttons that are exposed to me, and can imagine many lever/button layouts that would potentially make my life easier and better without first being hobbled by the foreknowledge of the many (I’m sure) technical limitations. I will try to discuss the technical aspects as far as I am able, but be aware that that’s not very far.

Dave Winer, father of two of my favorite things, RSS and podcasting, as well as huge influence on how I think about the internet, stopped by and commented on my last blog post. And then made the comment into a post on his own blog, Scripting News. Which is a great illustration of the point of my last post. This is unnecessary duplication and broken connection. The way I wish it would work, and the way that I was trying to describe in my last post is this: when Dave comments on my blog post, the comment would also automatically appear on his blog, with a link to my post. Or maybe more accurately, Dave would write a blog post (which would naturally appear on his blog) with a reference to my blog post and then my blog post would aggregate all responses to it (including Dave’s) and publish them underneath the text of the my post (with links back to all of them).

I think the difference in mindset is still basically a Web 1.0/2.0 one of read-only web vs. read-write web. When you stop making the distinction between a user and something-above-a-user (publisher, content-creator, aggregator, editor, administrator) you start to realize that it’s even flatter than we thought. Comments as they are now are like a bunch of micro-blogs that everyone creates on everyone else’s sites. Disqus starts to connect them so that all of a user’s comments made through Disqus are a more like a single micro-blog. My point is that microblogs are just a subset of blogs and all the contributions you make to the internet should be connected. It should all be part of your blog.

So I intended for this post to be just an elaboration on how I think the browser should encourage richer interaction, but the comments on my last post got me thinking more and more about how that’s really a subset of the problem of online identity (or lack thereof). So I’ll talk about the browser thing first and then talk about identity at large afterward.

Browser The thing about the internet is that you can be a producer of content very easily (and more easily as tools get better). You can be and are on the verge of contributing to the internet at all times. But our tools can better expose this potential to us as users, as well as encourage a richer contribution.

The components of my vision:

A browser.
An automatically created personal blog for every user.
A TrackBack-like protocol that aggregates all posts with reference to a post.
A set of filters for those TrackBacks.
Alternate views. I’ll elaborate on these in order:
I think it’s pretty obvious how the browser fits into the scheme but I also think that “browser” is no longer an adequate term for what these applications do now. “Browsing” is very low on the list of things I use my browser for, and already we’re linguistically confining ourselves to a passive role on the internet. “User agent” is somewhat clunky (plus I don’t really know precisely what it technically refers to), but I’ll go with that (or UA) until I think of something better.
The UA would also hook directly into a user’s personal blog when she logs into the UA (and ideally would take over most of the use-cases for most of other web services we use, especially simple ones like micro-blogs). Whenever the user feels the inclination, she may hit a button to invoke a horizontal or vertical split in the UA’s window between whatever content she was viewing and a large, expansive, non-modal WYSIWYG rich text field. (Alternatively, it could open a new window for the text editor.) Optional title field, optional body field. Optional tag field. Any post made from the UA would be published on the personal blog. Dave Winer’s Minimal Blogging Tool comes from a similar line of thought, I think (the collapse in hierarchy between input and output by users).
As far as I can tell, that last part could be implemented now on the application level, basically in a vaccuum. But I think the TrackBack part would need some sort of open standard protocol to work in a decentralized way. What it would look like for a user would be that at the bottom of any web page, or maybe by invoking another hard browser button, you could see all the UA-level blog post responses by anyone. Google actually implemented a very similar feature into the Google toolbar called Sidewiki. It failed as far as I can tell, and was panned by users and site owners alike, probably for good reason. My bet is that this was because of execution and lack of filtering, which I’ll talk about later. The bigger problem to me is that I think this used Google infrastructure to store and maintain the whole thing (corporate silo). I suppose the current TrackBack system or something like it would work, but it depends on 1. the original post to be able to receive communication from referring posts and maintain a list of them; and 2. the referring posts to notify the original post that they exist. I guess there’s not really any way around it needing some incarnation of both of those parts to work, which would need some sort of standarization of implementation of blogging platforms.
A user should be able to filter these responses/TrackBacks in any number of ways. I suppose the default would be a chronological firehose, but there should be a way to promote responses from different contact lists to the top of the pile. There should also be a way to filter by community/distributed collaborative filtering. This is more or less what Digg, Reddit, Slashdot, and Hacker News already are. They share stories, comment on them, then crowd-source the ranking of both stories and comments. The model of the Atlantic Monthly blogs’ comment spaces of well-moderated, directed discussion could also pretty easily be converted this way, I think.
The UA should have the ability to look at more than a single page at a time. The particular application for this that I have in mind is a view of an entire conversation thread. Tweetdeck for Android has a neat feature with which you can drill down into an @reply tweet and hit “conversation” and it’ll show you, where applicable, the entire conversation as far back as the back and forth @replies go. I like this because context is always good and it exposes the conversational, derivative, shoulders-of-giants, re-mix nature of ideas in general. Imagine reading an especially compelling post, then being able to hit a button that extends it up and down, showing the conversational context and thought process that led to it, as well as any number of responses.[^1]

This vision is only additive (and barely that). One could still navigate every part of this with a regular old web browser, because it’s still really just web pages and hyperlinks. None of the technology is new. It’s simply a different way of presenting options and artifacts that already exist in order to encourage better, more considered contribution to the community/communities of ideas.

It’s also really just another take on this idea of several posts ago (all content is the same, everything is streams and filters).[^2]

Identity

Let me now apply my lens of elevating the user/collapsing the distinction between user and producer to identity on the internet. I’ve already spent a bit of time thinking about this, and I still have mostly the same position I laid out in this post about persistent online identity. The problem I see in all the current projects like OpenID and OAuth, are the limited scope. OpenID allows me to not have another password to memorize, and OAuth gives me the ability to allow granular access to my data from one service to another, but neither of these do anything to connect the services that are associated with an OpenID or OAuth.

Here are the components of online identity as I see it:

A feed/feeds. Contact list/lists.[^3] Home page/Profile/public firehose/private feeds. You can see how this is mostly already described in my discussion of the User Agent.

Any User Agent home page or dashboard should contain not just incoming feeds and subscriptions, etc. (as described here), but also provide a field for output. A text entry field. And not small, but big, or barring that, at least one that resizes as your post grows longer (expands with your ideas).

It seems to me that the best chance of all of this happening is for Google to do it.[^4] Google already owns every necessary component: Chrome, Blogspot, Buzz, Contacts, and Profiles. And maybe even a protocol to connect all of it, similarly to how I describe: PubSubHubbub. And, indeed, they seem to be at the beginning of a slow process of rolling out some sort of identity service, either in the top bar of different Google services, or in the form of a toolbar.

The only thing really missing is a decentralized/federated way for people to host their own implementation and data, which isn’t necessarily out of the question, either, considering Google’s history.

Notes

[^1]: To take this in a slightly different direction: Currently, looking at web pages one at a time is the primary activity afforded by browsers. Reading feeds (for me in Google Reader) is sort of a secondary, virtualized (because a web-app) activity that’s not as robust or as fast, presumably because it’s not native (and none of the native feed readers I’ve tried have been particularly compelling). I kind of wish it were the other way around, because feeds are now primarily how I interact with the web.

[^2]: A distinction worth noting that I should add to the public-private spectrum: Public (or published) vs. Broadcast. There are things that I might want to put in front of everyone I know. There are other things that I don’t feel the need to hide from public view, but I don’t necessarily want to shove in front of everyone’s faces. I.e. I may want to have an exchange with someone that I don’t care is public, but I don’t necessarily want to clutter up people’s feeds.

[^3]: There is a tension between people’s desire for granular control and their hatred of granular management. This seems like kind of an intractable problem to a degree, but hopefully once identity and contact management are standardized and portable, you’ll only have to create lists once, and then the ongoing new contact sort would remain relatively manageable after that.

[^4]: It looks like Mozilla was working on making the browser the center for identity, too, but that post is pretty old and it doesn’t seem like it’s gone anywhere. Edit: I forgot about this.