SIGIA-L Mail Archives: Re: SIGIA-L: Dynamic documentation (was
Re: SIGIA-L: Dynamic documentation (was Adventures in XML in VISIO) - Heisenberg in action
From: Anders Ramsay (anders_at_andersmatic.com)
Date: Sun Mar 03 2002 - 12:19:32 EST
On 3/3/02 12:38 AM, Ziya Oz at ZiyaOz_at_earthlink.net wrote:
> Anders Ramsay wrote:
>
>> The data is fundamentally still the same; your diagrams may look
>> different from mine, but ultimately we are still talking about objects,
>> actions, and attributes (in part because that is the language of the
>> programmers, for whom the fundamental principles of their work has not
>> changed for quite some time.)
>
> Are you suggesting that IAs should get into the business of telling
> programmers what their objects, methods and class properties ought to be? I
> can't imagine many programmers being happy with that notion, can you?
This might just be an issues of semantics. By objects, I am referring to
any discreet entity that is part of the system design. My work would only
involve the user-visible object layer, such as a login form object (which in
turn would contain other objects, such as a password field.) The definition
and design of these objects is something I would complete *together* with
the programmers and the visual designers. This is essentially the first
pass at objects/actions analysis. And all we--as a team--are defining at
this point is *the what* (eg there will be a login form object on this page
and it will have functionality x, y, and z) while leaving *the how* to the
programmers (eg. functionality x, y, and z, will be implemented in this
way), meaning that the programmers will take this first pass of system
analysis and then use that as a basis for their code design.
And just as the user-visible layer will need to be documented, so will the
code design of the programmers. If object definitions, attributes and
values already exist in one document, I'm sure they would be happy to have
access to that data for use in their own documentation. Another reason for
open-sourcing documentation data.
Incidentally, there is an interesting article in a recent issue of
Interactions, which discusses object/actions analysis at the user-visible
layer:
http://portal.acm.org/citation.cfm?doid=503355.503366
-Anders
This archive was generated by hypermail 2.1.2
: Sun Nov 23 2003 - 22:55:05 EST
|
