I think that model navigation is *very* important, and the standard for
it should be "so abundantly easy that I can find any needed thing, even
when I'm brain-dead and have only some keywords to start with."  The
question is, how do you trace a feature of the system to the parts of
the model?

It is rare that anybody gets to manage a complex model through several
lifecycles, so there aren't likely to be many takers on that part of
your question.  Organizations move people, hire underqualified staff
whether they like it or not, and suffer from employees leaving.  In
these circumstances, it is best to assume that your model will be
(ab)used by people who have little chance of understanding it if you
don't keep it desperately simple to access.

There are things, like layered architecture, that you can do to resist
entropy, but you probably want cross-referencing and such that will
require a large expenditure of energy.  In software development,
translate energy to time and cost.  Now, figure out how you are going to
cost-justify the work to a management that goes by the Dilbert
management principle that "If I don't know how to do it, it must be
easy."

I conclude that most managements won't let you write this kind of
traceability into company standards if they understand it, and won't let
you actually do it in a project if they catch you, particularly when
they are looking at project deadlines and costs.

The problem doesn't just lie with the tools.  Fact is, we intentionally
hide interests in one (use case or class) diagram from interests in
another, and for good reason.  Breaking in a newbie is easier if what
was hidden can be easily revealed.

This is most definitely a place where the state of the art urgently
needs improvement.  Are you listening, Rational Corp?

-Eric

"Williamson, Rusty" wrote:
> 
> Hello,
> 
> On thing that has come up during the development of the Modeling Guidelines
> is the question of how much effort and attention should be given to model
> navigation: specifically making it easy to know what diagrams exist and
> getting to them.
> 
> My thoughts are that:
> a) The models can become quite large and some diagrams are buried deep
> within the models -- it would require expanding several levels down and then
> across packages in the browser to get to some diagrams and without knowing
> what exists significant searching around just to see what exists.
> b) Many stakeholders unfamiliar with the project (first time viewers or
> those joining for the development of the next version) will need to view
> and/or work with the models -- many unfamiliar with what diagrams do and/or
> 'can' exist, where to look for them and with varying proficiency with Rose.
> c) Making model navigation (drill- up and down) easy and intuitive as well
> as making all models visible can be done with minimal effort.
> 
> Yes Rose has 'find' and 'browse' but in both cases you have to know what
> your looking for.  I believe that this issue applies to all of the models
> but especially to the use-case model.
> 
> Others argue that all this is simply unimportant.  Maybe their right.  I
> wanted to get the thoughts of others who had been through the process of
> managing complex models (and ours will be complex) through several life
> cycles.  Your thoughts?
> 
> Thanks,
> Rusty
> ------------------------------------------------------------
> Rusty Williamson
> > Sr. Systems Architect
> GERS Retail Systems
> http://www.gers.com/
> The Object Workshop
> http://home.san.rr.com/williamson/
> Home Page
> http://www.znet.com/~rusty/
> ------------------------------------------------------------
> 
> ************************************************************************
> * Rose Forum is a public venue for ideas and discussions.
> * For technical support, visit http://www.rational.com/support
> *
> * Admin.Subscription Requests: [EMAIL PROTECTED]
> * Archive of messages: 
>http://www.rational.com/products/rose/usergroups/rose_forum.jtmpl
> * Other Requests: [EMAIL PROTECTED]
> *
> * To unsubscribe from the list, please send email
> *
> * To: [EMAIL PROTECTED]
> * Subject:<BLANK>
> * Body: unsubscribe rose_forum
> *
> *************************************************************************
************************************************************************
* Rose Forum is a public venue for ideas and discussions.
* For technical support, visit http://www.rational.com/support
*
* Admin.Subscription Requests: [EMAIL PROTECTED]
* Archive of messages: 
http://www.rational.com/products/rose/usergroups/rose_forum.jtmpl
* Other Requests: [EMAIL PROTECTED]
*
* To unsubscribe from the list, please send email
*
* To: [EMAIL PROTECTED]
* Subject:<BLANK>
* Body: unsubscribe rose_forum
*
*************************************************************************

Reply via email to