Re: [Fle3-dev] Create new knowledge type set: Software Lifecycle

[Cormac McClean]

Hi Tarmo,

Thanks for the feeback.

> >Isn't the waterfall model also rather restrictive? In that sense some
> >kind of rules could make a lot of sense in here.
> 
> Also, the waterfall is quite old, and in most software development
> situations doesn't work very well. It demands that the customer know
> exactly what they want and that the developers know the technical
> platform like the backs of their hands. If the platform is new or
> unfamiliar, the waterfall will break. And if the customer can't define
> exactly what they want, they'll want changes later on, and the
> waterfall will break.
I agree with you.  In fact, part of our course (CSE2 Software Systems)
is to investigate and evaluate the various software development
models, which includes a discussion of the failings of the waterfall
model and the improvements offered by other approaches - e.g. Rapid
Prototyping model and the Incremental model.

> >Is anyone interested in to create typeset for "extreme programming"
> >(http://www.extremeprogramming.org/)? At least the user stories,
> >elaboration and review of them,  release planning
> >(http://c2.com/cgi/wiki?PlanningGame) etc. could be ran and this way
> >teacher with help of knowledge building.
> I could do it, I guess, but I'm not convinced that using a lifespan
> model as knowledge types really brings any benefits. But that may be
> just that I don't understand how they would be used... But with the
> waterfall model I personally would create one course context for each
> phase and use knowledge type sets like this:
> 
> - requirement analysis: design thinking types
> - analysis: design thinking types
> - design: design thinking types
> - implementation: progressive inquiry
> - testing: progressive inquiry
> - maintenance: progressive inquiry
> 
> Or something like that. Of course, to preserve the waterfall, each
> context would be created only after the previous has been finished.
I see how this approach would work. In fact, with another course group
(EC2-PCE Computer Programming) we actually started this way, using the
built-in KTSs for each phase, but found that the note types and check
lists weren't specific enough for our purpose. I thought that a
custom-built KTS would provide a "one-stop-shop" for the development
lifecycle, with specific note types and checklists, and also having
the benefit of allowing the students to see how the different phases
of each model relate to each other (warterfall model being very
linear, incremental model taking a different route) and this is what
I'm using with the CSE2 Software Systems group.

My students and I are really experimenting with FLE3 and trying to
make most effective use of it in our course. I will compare the two
approaches:
1. EC2-PCE Computer Programming using the built-in KTSs, one for each
phase (so far we have used the design thinking types for the idea,
requirements analysis and design)
2. CSE2 Software Systems using the custom-built software lifecycle KTSs.

Thanks for the great feedback.

Best wishes,
Cormac.
-- 
Cormac McClean.
Lecturer, Athlone IT Software Engineering
Free/Open Source Software Enthusiast

On Tue, 1 Feb 2005 13:51:58 +0200, Tarmo Toikkanen
<address@hidden> wrote:
> >Isn't the waterfall model also rather restrictive? In that sense some
> >kind of rules could make a lot of sense in here.
> 
> Also, the waterfall is quite old, and in most software development
> situations doesn't work very well. It demands that the customer know
> exactly what they want and that the developers know the technical
> platform like the backs of their hands. If the platform is new or
> unfamiliar, the waterfall will break. And if the customer can't define
> exactly what they want, they'll want changes later on, and the
> waterfall will break.
> 
> >>I'm also planning the following:
> >>1. Knowledge Type Set: Incremental Model
> >
> >Don't' know this model. Google told me that it is also a software
> >development model. Interesting, tooo.
> 
> Incremental (or cyclic) models are a generic name for most models that
> break away from the traditional waterfall. Extreme programming is an
> incremental model, so is the Rational Unified Process and countless
> others. Basically all agile methods are incremental, also.
> 
> >Is anyone interested in to create typeset for "extreme programming"
> >(http://www.extremeprogramming.org/)? At least the user stories,
> >elaboration and review of them,  release planning
> >(http://c2.com/cgi/wiki?PlanningGame) etc. could be ran and this way
> >teacher with help of knowledge building.
> 
> I could do it, I guess, but I'm not convinced that using a lifespan
> model as knowledge types really brings any benefits. But that may be
> just that I don't understand how they would be used... But with the
> waterfall model I personally would create one course context for each
> phase and use knowledge type sets like this:
> 
> - requirement analysis: design thinking types
> - analysis: design thinking types
> - design: design thinking types
> - implementation: progressive inquiry
> - testing: progressive inquiry
> - maintenance: progressive inquiry
> 
> Or something like that. Of course, to preserve the waterfall, each
> context would be created only after the previous has been finished.
> 
> >http://lists.gnu.org/archive/html/fle3-dev/2004-11/msg00015.html
> >
> >http://lists.gnu.org/archive/html/fle3-dev/2003-01/msg00000.html
> >
> >Tarmo knows if there is already something done with this - I guess not.
> 
> Nope.
> 
> --
> Tarmo Toikkanen
> http://www.iki.fi/tarmo/
> The Fle3 Team
> http://fle3.uiah.fi
> 
> _______________________________________________
> Fle3-dev mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/fle3-dev
> 


-- 
Cormac McClean.
Lecturer, Athlone IT Software Engineering
Free/Open Source Software Enthusiast