swamp! alligators, it's hard to remember muzzling the ... slides.pdf · slides for embassy...
TRANSCRIPT
Page 1 31/12/99
Muzzling the Alligators
When you are up to your neck in alligators, it's hard to remember that the objective is to drain the
swamp!
Page 2 31/12/99
A Conflict of AdviceThere are plenty of people with ideasto improve quality, but experiencedpractitioners have their doubts!
Page 3 31/12/99
Following the Paper Trail
The Paper Trail
The Infantry Approach
Lack of Support for Design, Build & Test
Optimism About Developer
Motivation
Expense
Page 4 31/12/99
Developers' ProblemsDevelopers are short of time, budget or both
Need to know : "What do I actually do?"
Page 5 31/12/99
Eurotunnel's Problems
Old systems, non-standard architectures
Geographical and linguistic diversity
Supplier quality approaches
The mythical "Package"
Unstable requirements
New technology
Missing users
Page 6 31/12/99
The Way Forward
- Need for a framework- Eurotunnel's adoption
of CASE*Method- Defined documents
and programming standards
- Help with procurementdocumentation & testing
Page 7 31/12/99
Standards and TemplatesSTANDARDS MUST BE TECHNICALLY
SOPHISTICATED, BUT CLEAR
TEMPLATES EASE THE
ADOPTION OF STANDARDS
LIFE-CYCLE: RESISTANCE,
GRUDGING ACCEPTANCE, ENTHUSIASM
IGNORANCE,
Page 8 31/12/99
Procurement Services
Eurotunnel tries to use "Package" solutions
Procurement guidelines, draft ITTs and contracts
Administration in hands of QA
group
Needs policing as
well as support
Page 9 31/12/99
Communications
Page 10 31/12/99
Inspections and AuditsInterventionist stylePeer & QA reviews for documents
insp ions
Projects and suppliers are all audited
Page 11 31/12/99
Technical Expertise
� Adoption of new techniques
� Early development of tools & standards
� Central repository of knowledge
Page 12 31/12/99
Testing
ET's own standards & courses use structural techniques & inspections
Page 13 31/12/99
Change Control
Both documents and software change frequently
Version control is relatively easy - configuration control is more difficult
Life-cycle: Chaotic Ordered �
Controlled
Page 14 31/12/99
The QA Role
Page 15 31/12/99
The First Steps Forward
Don't try to do it all in one goKnow your customers
Find a framework
Highest value firstDare to be different!
Value = Benefit/(Cost*Risk)
Page 16 31/12/99
There is NO Magic Wand
�Improving Qualityis Slow & Painful�All Parties Mustbe Pragmatic�Make GradualChanges, Matchedto the CorporateCulture�Highest Benefit,Simplest First