Re: [soa-rm-ra] How to Refer to Various Uses of the word "machine"

[Rex Brooks <rexb@starbourne.com>]

I agree.

How about declining to use "machine?" By the time we finish the job 
of satisfactorily defining our particular usage, and enforcing it 
across the specification, we are liable to spend a lot of time on it.

Working on the Security Threat Model, I could be happy with 
discussing the "mechanics or operational processes involved in 
achieving...."

Metaphors are almost unavoidable because most human understanding 
requires them. Apparently, that is the way our brains prefer to work. 
However, we need to be careful about how we do that. There's a whole 
world of tropes out there, just waiting to be used incorrectly and 
trip us up, such as identifying a metaphor, which states that 
something "is" something else to show how the first something 
partakes of the attributes of the second something, with simile, 
which says something is "like" something else.

Sorry for being pedantic, but metaphor is the second most badly 
abused term in the computing lexicon, with the first being 
"paradigm." I have given up trying to get people to understand that 
paradigm is an example of a principle, not the principle itself, and 
I now just accept that is forever going to be used in place of 
principle and often at the same time. However, there is still hope 
for metaphor. ;-)

Pedantically Yours,
Rex

At 9:11 AM -0700 6/1/06, Danny Thornton wrote:
>One opinion for machine:
>
>When I use the term machine I use it as a metaphor to
>convey something meaningful being built with many
>working parts.  For me, the term machine does not take
>on the formality of an embedded part of the design of
>a software architecture.  In other words, the term
>machine sounds good for slide presenations, marketing
>purposes, and to gain interest but it does not look
>good as a term in a formal software reference
>architecture. 
>
>The term "machine" does show up twice in the Goals,
>Critical Success Factors, and Requirements.  The first
>usage is consistent with using machine as a metaphor:
>
>A reference architecture is like an abstract machine.
>
>The second usage could raise a question about what
>kind of machine:
>
>Any machine is necesssarily broad in its applicability
>and is often under-constrained.
>
>Many engineers might think of several kinds of
>machines that are not broad in their applicability and
>are not under-constrained. 
>
>Other opinions?
>
>Danny
>
>
>
>--- Michael Stiefel <development@reliablesoftware.com>
>wrote:
>
>>  As I start to write my sections for the RA,  I find
>>  myself using the word
>>  machine in several contexts and I wonder if we can
>>  come up with a way to
>>  use these words that will be consistent through out
>>  our document.
>>
>>  One use of the word machine comes from our current
>>  use of the machine
>>  metaphor for describing the RA.  I find phrases as
>>  the "SOA RA machine" to
>>  be quite awkward. Another use of the word machine is
>>  a device such as a
>>  computer that might be using the services of a
>>  particular SOA.  If the term
>>  machine is used unmodified it is often unclear to
>>  which machine one is
>>  referring.
>>
>>  I am sure we will find others.
>>
>>  Michael
>>
>>
>>
>
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around
>http://mail.yahoo.com


-- 
Rex Brooks
President, CEO
Starbourne Communications Design
GeoAddress: 1361-A Addison
Berkeley, CA 94702
Tel: 510-849-2309