[Sigia-l] We could just use whiteboards instead.

[Listera]

"frank.shepard at att.net" wrote:

As usual, Frank goes right to the heart of the issue.

> Here's a new question: Is it really possible to disengage the "end product of
> an AI" from the "end product of development?"

Yes. Even if it were possible, why would you want to chop up the (overall)
design process into disparate, disjointed processes? Especially for smaller
shops/projects that's not cost-efficient, to say nothing of quality control.

> Perhaps this is akin to questioning the status of IA as an independent unit?
> What is the IA's role in the transition from his/her "end product" to the
> development stage?

I'm an advocate of the Japanese-garden approach to design: it's not finished
until there's nothing else you can take out. I view the whole design process
as one of disambiguation.

Color post-its, reams of paper docs, scenarios, personas, etc., do not
sufficiently reduce ambiguity. (Remember, our final products are 100%
digital/binary artifacts.)

As a developer, I don't want, say, my objects to be described by a garden
variety IA, especially one stuck in the post-it era. Most of those UML stuff
created by the same people are most likely not precise enough to be truly
useful to developers; a lot of the complex, abstract wireframe stuff usually
go over the heads of business line people. So much of the legacy
"deliverables" tailored to "IA as a distinct unit" Frank alludes to can be
problematic.

What disambiguates much better is a functional, digital prototype. When in
doubt, click and see for yourself.

Now, as I said before, the current state of prototyping tools is not what it
should be. The trend, however, is positive and we need to look into the
future, not into the past of cellulose madness.

If you are not happy with your digital tools, by all means, raise hell with
the vendor. Take the time to tell them what they are doing wrong, or not
doing at all. If you don't do it, who will?

Ziya
Nullius in Verba