Ascription is an Anathema to any Enthusiasm

This paper by Steve Johnson is wonderful. I spend a lot of calories thinking about how groups form, but also about how groups create shape the membrane around them. I’m less interested in the organizational problems inside the cell, in part because there is so much liturature about that. Using the Dean campagn as a case study this paper notices that clustering is not enough. At some point a group will need to pile on the means to cope. Coping is not the same as clustering.

The paper suggests a kind of race condition developed in the Dean campaign. It’s clustering drivers ran real fast; too fast for it’s coping skills to build out. That’s not unlike a syndrome we see in Internet systems that catch fire. Friendster for example grew faster than it’s owners could cope with. Slashdotted, or the more venerable flash-crowd, is another name for the syndrome of a group forming event that blows up thru fad, crowd, and into riot.

There are a number of great one liners in the paper.

“…One of the funny things about the literature of emergence is that it is strangely obsessed with slime. Slime mold, to be precise…”

When I speak to an audience about Open Source I’m often asked by a middle aged quiet guy in the audience; “but what about managers.” I love that one liner because it is asking just that question. I might begin to reply by talking about the kind of coping methods you find in these groups.

Clay has an essay about the inevitable constitutional crisses that comes upon groups as they mature. Some organizers try to put the cart before the horse; they write the constitution before they have the revolution. Constitutions are distilled coping skill. It’s a kind of cargo cult confusion of cause and effect. Surely, they think, many groups have constitutions so constitutions must create groups. We can reframe that idea using the insights of this paper; sure it can be fatal for a group to lack coping skils, but first they need to have a driver to form them, e.g. clustering skills.

In the terms of my preferred three legs that a community stands on (common cause, common ritual/narative, and loyality) I tend to emphasis the aspects that drive the clustering. I leave the coping skills are packed up inside the common rituals and loyality. I’ve tended to think of the clustering as more interesting; mostly because the Internet keeps enabling more of that - more gathering, more rondevous, more group forming.

Like “coordination” or “membrane”, the word “coping” provides a nice addition to the vocabulary around groups.

Tim Oren appears to be suffering from sympathy. An unusual affliction for a VC. A handful of reasons why, Mr. Entrepreneur, your amazingly cool new innovation isn’t going to get funding:

1. Cool demo! Great price point, what a leap forward! Man making stuff like this dependable is really hard work. You probably can’t see that, having just moved so fast. You know the incumbent is always improving. Sorry, I think we’ll pass.
2. Cool! - oh - It’s so cool that pretty soon it will be a ‘must have’ and the platform vendor will swallow you. No durable business here. Next.
3. Neat! - oh - Sadly your customers are too smart. Even if you get past the damn designers (NIH etc.) you won’t get by the risk adversion of the project manager.
4. Your customer’s too powerful and too slow. They take years to release a new product and they never leave any real money on the table for the component providers.
5. Customers would love that! - oh - The channel’s doesn’t care. The channel sells by subscription: i.e. they want zero up front cost, sticky features, and value delivered in dribble. You don’t fit that model; your toast.
6. Ha, that’s clever. The early adopters will love it. But, I just can’t see how the late adopters are ever going to see that as anything but added complexity.

Needless to say Tim put’s this all in a much more professional and accurate way.

For any number of reasons I’m interested in middlemen and brokerage.

Here are four ways that a broker might bring value to another group.

• Inform either side of interests or difficulties of the other.
• Import useful techniques, practices, tools.
• Draw analogies that the other group is blind to.
• Synthesis of techniques, practies, tools, ideas thru merging.

I very much doubt this is a usefully complete list. For example it doesn’t even include trading goods between the groups. But it’s a start.

I underestimated how sticky Moveable Type is.

Vendors love things that make their product sticky. If developers really appreciated this software products would be even more sticky. Instead developers hate sticky; they call it things like “backward compatiblity” or “legacy.” Maintaining the sticky bits is a pain. Platform developers have the worst of it because the software that stands on the platforms was written by very very clever dudes who find and depend on every curiosity of your API. The software those clever guys write is extremely brittle. The platform’s vendor has to work very very hard to maintain every bizzare detail.

When I switched to Word Press my web server logs suddenly blossomed in a torrent of broken links. I’d arranged to reroute the obvious links before the switch over. But, it turns out that my site’s users are as devious about finding interfaces into my blog as platform developers. It looks a bit like every URL that you could possible generate for reaching into the blog was used by somebody. This was particularly hairy for the various subscription feeds. I notice that a lot of subscription readers aren’t particularly interested in paying attention when my server notifies them that a resource has moved. So now I’m serving up the subscription feeds from the old locations. I wonder how many subscribers I lost during the service interuption?

The backward compatiblity breakage that I didn’t see comming was with Google. All my page names changed to something new and so Google’s model of what’s on my pages evaporated. All my Google ads suddenly became extremely lame. Imagine how sticky things would be if you depended on the revenue from such ads.

The third interface where backward compatiblity is turning out to be very rough is the blog author user interface. I don’t mind switching to a new user interface. Somethings are better, some are worse. But what about the other folks? The folks who’s blogs I host. I think they are going to hate it. In general they get to suffer the cost of changing but for them the benefit of the switch over is pretty obscure. In particular the photo upload in Word Press is much more tedious. Imagine if those people were paying me for their blog hosting?

I used to think that the #1 thing to worry about in buying software was that I would be able to rescue my data, retire the software, and adopt something else. Apparently in the modern world software embedded in the marvalously messy open platform that is the Internet it’s much more complex. We are all platform vendors now.

I like it because it saves typing. There was a good example of a 20-line Java program that was reduced to 3 lines in Groovy. I’ve lost the link though. — Script Musings

That’s the problem. Tiny programs are so easy to misplace. The entire corpus of APL programs was practically whipped out back in the 1970s by a high wind. Don’t want to loose things? XML!