[SystemSafety] professionalism (was: Re: Logic)

[Martyn Thomas]

Steve

I have a high regard for IEEE-CS and I have serrved on policy committees
of the BCS and IEE/IET in the UK. Let me prespond to just one of the
points you make.

On 19/02/2014 01:09, Steve Tockey wrote:
>
>> 5. The governing body must take disciplinary action including, if
>> necessary, expulsion from membership should the rules and standards
>> it lays down not be observed or should a member be guilty of bad
>> professional work.
>
>
> 5. This is already in place in IEEE-CS
>
>

I have spent many years as a consultant and expert witness reviewing UK
and overseas projects that have run into difficulty. I routinely found
commercial (not safety-critical) software development projects that had
the following deficiencies (among many others):

a) The requirements statement was incomplete and highly ambiguous,
leading to major disputes late in the project about whether a
clarification was a chargeable change.

b) The initial plans were incomplete, superficial and absurdly
optimistic. The project was priced to win the contract, not on the basis
of realistic costings, with the intention of recovering profit margins
later.

c) The project risk register, if it existed at all, was superficial and
the risk mitigation activities were trivial (for example: RISK: "the
requirements may not have been captured fully". MITIGATION "Change
Management Board").

d) The plans contained no allowance for any impact from the risks
materialising.

e) Milestone definitions were vague, so milestones (especially payment
milestones!) were claimed as "passed" when work and risk had actually
been moved to later in the project.

f) Key project documents lacked version control.

g) The main assurance activity was testing, at a late stage in the
project, with no planned time or budget for correcting any errors found.

h) Technical decisions were taken for unprofessional reasons (such as
"we wanted to gain experience of using this architecture or programming
language").

i) The lack of specifications and documentation is excused by the claim
that "we are using agile methods".

I could go on (and on, and on ...)

I can say with some authority that there is no appetite in the BCS or
the IET for taking disciplinary action over such unprofessional
behaviour by members (even though both institutions have codes of
practice similar to those of IEEE-CS). I'm not aware of any occasion in
which a UK member of IEEEE-CS has been expelled for unprofessional
behaviour either (though it may have happened without publicity).

Are things so much better in the USA? If so, how was it achieved and how
does the IEEE enforce its excellent ethical standards?


Martyn


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20140219/9315db93/attachment-0001.html>