« KmWiki AAR | Main | Mistakes & failures »

July 16, 2005


David Locke

With your mentions of specialist search engines, and backgrounding of identity in conjunction with making distinctions, I'm wondering if you are busting the silo or reinforcing it. Which is it? Both?

David Locke

"A network is essential to extend the reach of your sharing." Well, ever representation has its own network, so you need more than one network. In most of your own knowledge discussions, you mention the social network as well as the technical network. But, there are many more networks than that.

David Locke

Yes, building meaning around a prototype is possible. What might not be possible is explicating that meaning. The point of a prototype is to let implicit/tacit knowledge to determine what impicit/tacit knowledge gets embedded in the artifact.

We can see that the prototype changed, but we won't necessarily know what the change means, nor do we always know what drove the change. The change could be an experiment, so the meaning is yet to be decided.

David Locke

"... the wall map in a busy 'war room'" was never self organizing.

A blackboard is a representation that allows agents to communicate. This claim of self organization comes from the inherent emergence of agents, but that isn't real world.

A blackboard or war room map requires models, views, and controllers. They require an organizational overhead, which is usually missing.

David Locke

In a code is free environment you do not need to elicit group consensus. Eliciting group consesnus is where requirement elicitation goes wrong.

Concept maps would reveal the cultural boundaries of those participating in the annealing. Those boundaries need to be respected and coded to. Those boundaries do not have to be erased by the poltics of consensus and the target destination of those politicians.

David Locke

Sorry Denham, but when you create a controlled vocabulary to communicat with generalists you are not "Making deliberate distinctions to bring forth unique meaning."

You are in fact destroying deliberate distinctions.

David Locke

Developing ontologies is a wrong headed exercise. Silos exist for cultural reasons. Giving up the definitions within the silos for a thin vocabulary of politically constructed generality reduces options for the organization and ultimately kills it. The death will come from the narrowed bandwidth and the communications failures inherent in a language that no longer says what it means.

Efficency in IT has given rise to functionality that doesn't quiet do the job. This is because the requirements elicitors ignore culture. But, this efficency should not be the goal any longer, particularly in an environment where code is free, aka OpenSource, Offshoring, Outsourcing. This efficency created problems with the mappings of UIs to conceptual models even as developers think that is what UML does. No, IT ignores culture.

And, by extention controlled vocabularies and other web techniques continue that tradition. So we end up with effiecent looking controlled vocabularies that help us fail to communicate, because culture is ignored again.

Culture creates its own issue as well, because corporate culture is selected from amoung the many that comprise the organization. The corporate culture is inherently line and predisposed against the functional unit, against the silos, against the generation upon generation of expertise that consitutes silo culture, cultures that travel with the employee because that culture is tied to the employee's identity from the moment they decide to separate themselves from the line people, the "on" discipline generalist, and become the "in" discipline practitioners.

We reengineered to kill our companies. Now we will but the silos to kill our companies.

David Locke

The model-view-controller pattern works for clowns and magicians.

They may look like distillations of practice, but they are really metaphors. Everything I ever learned about computer science turned out to be a metaphor for other applications of the patterns.

I'd almost go as far as saying that the metaphors were constraining practice even before the pattern was recognized. Practice mearly explicates the tacit knowledge that was there the whole time. Patterns are explicit, so pattern writing is explication.

But, how do patterns escape probablistic assertions? How do patterns escape being theories?

David Locke

Why would a groupware vendor know anything or care anything about collaboration? Did they hire the collaboration vendors, so they could pick their brains.

Nothing a software vendor creates is done with the end markets in mind. Desktop publishing software was built without asking typographers anything. That's why you could position things by centimeters and tenths of an inch, easily programmed stuff that really wasn't real world.

Software vendors don't hire product managers until after version 1.0 is sold. Only then does a market identify itself.

And, the collaboration vendors/preachers are not going to site down and write software. They've got too much preaching to do. Geeks do this stuff off the top of their heads even if they know almost nothing about it.

The comments to this entry are closed.