[aosd-discuss] How determine what is crosscuting in requierements level

Awais Rashid marash at comp.lancs.ac.uk
Fri Apr 11 00:07:29 EDT 2003


Anderson,

Here is my take on the issues you raised:

> Reading these papers, I didn't realize the difference between a
> concern and
> a requirement. Some papers consider the same (eg. "Towards a Composition
> Process for Aspect-Oriented Requierements" ) and others not (eg. Early
> Aspects: A Model for Aspect-Oriented Requirements Engineering).

As for the term "concern", this has been used differently in different
methods. There is the notion of the general "concern" in a system. This
could be anything e.g. a viewpoint, use case, aspect, etc. Then there is the
more restricted notion of concerns which are collections of broadly scoped,
non-functional requirements. Cosmos, uses the term concern in the former
sense of while our work employs it in the latter sense of the word. This is
not to cause confusion or provide an alternative definition of the term. It
has been used by earlier work on requirements engineering, the PREView
method, which we use as a basis to identify broadly scoped properties. So we
are not contradicting the general use of the term. We are simply using it as
it is employed in PREView to assist in elicitation of broadly scoped
properties.

>
> In this discussion list an explication about this question was done like
> "security is a concern for each system there might be different set of
> requirements with respect to the security concern", but Chung`s
> work (very
> considered in requirents community) has classified security as a
> non-functional requirement (Dealing with Non-Functional
> Requirements: Three
> Experimental Studies of a Process-Oriented Approach).

Non-functional requirements are still a concern in the general sense of the
word (and even in the more restricted sense used by PREView and ARCADE).

>
> Other open question, in my opinion, is determine a crosscutting
> concern by
> use case used by other use cases. By this analysis will be
> possible identify
> candidate aspects? In this case, the use case, wouldn`t be only
> related with
> each other?

If a use case has a broadly scoped effect on other use cases then it should
be considered as a candidate aspect. Let me also add that identification of
candidate aspects is a challenging task. We (developers and software
engineers) find it easier to identify use cases, viewpoints, objects, etc.
as we have more experience of these compared to identification of aspects.
So we have some intuitive understanding of what an object or a use case
might be. Such intuitive understanding for aspect identification (at the RE
level or later stages) is going to be developed as we use AO techniques more
regularly. In fact, identifying RE level aspects can assist in
identification of aspects at later development stages if traceability can be
maintained.

Kind regards,

Awais.

>
> Kind regards,
>
> Anderson Davis
>
>
> >From: "Awais Rashid" <marash at comp.lancs.ac.uk>
> >Reply-To: <marash at comp.lancs.ac.uk>
> >To: "Anderson Davis" <aosd at hotmail.com>,<discuss at aosd.net>
> >Subject: RE: [aosd-discuss] How determine what is crosscuting in
> >requierements level
> >Date: Wed, 9 Apr 2003 23:27:47 +0100
> >
> >Anderson,
> >
> >The early aspects workshops at AOSD 2002 and 2003 are good
> starting points:
> >
> >http://trese.cs.utwente.nl/AOSD-EarlyAspectsWS/
> >
> >http://www.cs.bilkent.edu.tr/AOSD-EarlyAspects/
> >
> >The following papers also provide some insight into this:
> >
> >S. M. Sutton and I. Rouvellou, "Modeling of Software Concerns in Cosmos",
> >International Conference on Aspect-Oriented Software Development, 2002,
> >ACM,
> >pp. 127-133.
> >
> >J. Grundy, "Aspect-Oriented Requirements Engineering for Component-based
> >Software Systems", 4th IEEE International Sympsium on RE, 1999, IEEE
> >Computer Society Press, pp. 84-91.
> >
> >A. Rashid, A. Moreira, and J. Araujo, "Modularisation and Composition of
> >Aspectual Requirements", 2nd International Conference on Aspect-Oriented
> >Software Development, 2003, ACM, pp. 11-20.
> >
> >A. Rashid, P. Sawyer, A. Moreira, and J. Araujo, "Early Aspects: A Model
> >for
> >Aspect-Oriented Requirements Engineering", IEEE Joint International
> >Conference on Requirements Engineering, 2002, IEEE Computer
> Society Press,
> >pp. 199-202.
> >
> >Kind regards,
> >
> >Awais.
> >
> > > -----Original Message-----
> > > From: discuss-admin at aosd.net
> [mailto:discuss-admin at aosd.net]On Behalf Of
> > > Anderson Davis
> > > Sent: 09 April 2003 21:20
> > > To: discuss at aosd.net
> > > Subject: [aosd-discuss] How determine what is crosscuting in
> > > requierements level
> > >
> > >
> > > Dear Friends,
> > >
> > > I have some doubts about how we can think about aspects in
> requirements
> > > level. In my opinion, aspects are part of solution domain and not
> > > about the
> > > problem domain. One of my questions is:
> > >
> > > Considering crosscutting concerns as concerns that can't be
> > > encapsulated in
> > > a only one entity and that generally is scattering around several
> > > entities,
> > > how is possible to determine if a concern in requirements level is
> > > crosscutting or not?
> > >
> > > I would be very grateful if some of you could answer these questions.
> > >
> > > Thankful for your attention,
> > >
> > > Anderson Davis
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > _________________________________________________________________
> > > MSN Hotmail, o maior webmail do Brasil.  http://www.hotmail.com
> > >
> > > __________________________________________________
> > > AOSD Discuss mailing list    -    discuss at aosd.net
> > > To be removed send mail to  discuss-admin at aosd.net
> > > or visit http://aosd.net
> >
>
>
> _________________________________________________________________
> MSN Hotmail, o maior webmail do Brasil.  http://www.hotmail.com




More information about the discuss mailing list