I was listening to the Supertop Podcast today (you should too -- in the past, this show has had some great sponsors), and they were discussing Manton Reece's micro.blog Kickstarter, which added and met a stretch goal at $80k before it was all said and done.

I shelled out $10 to see the ebook, mostly because I'm intrigued, but not real sure what the project is about. On a past episode of Supertop, the guys mentioned that they weren't real sure what the project was precisely about themselves (EDIT: Turns out Daniel on Release Notes told Manton this same thing back on January 20th!).

Manton has discussed on his podcast, Core Intuition, several times about how much he's sweated the video for that Kickstarter (come to think of it, I don't think I ever watched!), but his real kicker (har har) is that he never actually precisely described what it was he's doing, so we're left to imagine micro.blog to be whatever we want it to be. It's a book about bringing back blogging, and it has a Twitter-like component. Not only that, it promises to "embrace... the open web". These are all laudable goals. Who wouldn't want to be on board?

But isn't that two services?

Interestingly, the stretch goal was...:

Stretch goal update: if we reach $80,000, I'll hire a community manager for Micro.blog to help build a safe community from the start. See this update on Kickstarter for details on the stretch goal and a new feature called Safe Replies.

A community manager? Wait, what? An open web service with a manager? How do you pull that off, exactly?

There are a number of obvious strategies. One is that the manager lords over a single server which is used to disseminate posts, and that your publishing process lets that server know as you're posting your content. Then others subscribe to that server in typical client-server fashion.

But that kills the open, doesn't it? Someone holds your access to your own subscriptions!

Maybe you could have a service that gives out guids for posts -- or even users -- that it doesn't feel a cultured schmoe would like to see, and your client uses those review services to pre-censor your open content. Nevertheless, management is much different than open.

I think a management/review service is a very different project from this one:

Do you remember how the web used to work? How the web was supposed to work?

In the earlier days of the web, we always published to our own web site. If you weren’t happy with your web host, or they went out of business, you could move your files and your domain name, and nothing would break.

Hopefully what this means is that we have two services. One is an open stack for publishing & consuming and the other is a review service that you can optionally use.

What is micro.blog?

Whether micro.blog is open or managed is precisely what I've wondered from the start. Manton's mentioned using (or replacing) Webmention, a w3c spec, a few times. That's an interesting move.

Webmention is sort of like an anonymously built webring -- remember those? When you'd slap a banner on your site so that anyone clicking could go to the "next, prev, or random" member of the same virtual community? Webrings built some of the first web-based, self-consciously social networks, versus, say, simply chatrooms and newsrooms based on topics, though I guess chatrooms and IRC were close.

Anyhow, webrings are sort of what Medium has effectively replaced. Now your blog has a social component [again]. Before, you had related conversations linked together. Now you link the conversations themselves. Webmention and Medium both accomplish this in one form or another.

What's necessary to have an open conversation?

But micro.blog is apparently more than that; it's also a Twitter replacement. And for that -- high intensity, short-form conversations broadcast publically -- you either have to subscribe to an RSS feed (or equiv) for someone's "micro blog posts" and then follow all reciprocally linked replies, or you'd have to have a centralized server which keeps track of unique names and collects and gives out all of the posts/information.

The first, a feed subscription, is a Webmention (or effectively a webring, though often even webrings had a server providing the "random" linking) like function. The second is a centralized client-server setup that's more susceptible to command and control. Even if there are, in theory, thousands of potential manager servers, it's still not really an open system.

And it's not really the openness that necessarily drives growth. I mean, there are already open Twitter alternatives. Building it doesn't mean that they'll come. There's a reason Twitter is still the 800 lbs gorilla.

Also remember that one of the giveaways for backing micro.blog was, "You can reserve your Micro.blog username even before the book is finished." Username? How is there a unique username on the open web without, say, an ICANN? Nonprofit or not, you're still centralized. Users won't own everything until you're using a mesh/p2p-ish network with the possibility of data collision -- and the algorithms to manage them. (Welcome to the world of public keys...)

In any event, I hope the book makes this stuff clear, and some folks bite on creating infrastructure. Even if Manton's holding the keys to "the" central server, it's an interesting experiment. How much can he offload hosting to keep costs down, but still provide reliable content? And how difficult would it be to make an Android client... ;^)

micro.blog is an interesting pie in the sky, but it's going to be a real pain to pull off and deliver on all the ambiguous promises it's perhaps unintentionally made. I wish him luck.

Labels: , ,