[SystemSafety] Collected stopgap measures

[DREW Rae]

Peter,
You are misrepresenting both Boehm's work and the Knight-Leveson n-version experiments.

Boehm was talking about collecting data on past estimates to improve future estimates, not about evaluating the generic effectiveness or efficiency of specific practices.

Knight and Leveson were evaluating whether n-version software was prone to common errors in each version. Their results say nothing about whether controlled trials are good scientific practice.

Derek is exactly right about the problem. Everyone wants to generalise their own experience to make claims about what everyone else should do.

Comparison is the heart of rigorous investigation, and both software and safety communities aren't known for their rigour.

Kind of ironic that the people trying to impose rigour on software engineering are constantly taking shortcuts in collecting evidence to back up their arguments.

Drew

Get Outlook for Android<https://aka.ms/ghei36>

________________________________
From: systemsafety <systemsafety-bounces at lists.techfak.uni-bielefeld.de> on behalf of Peter Bernard Ladkin <ladkin at causalis.com>
Sent: Saturday, November 17, 2018 5:35:25 PM
To: systemsafety at lists.techfak.uni-bielefeld.de
Subject: Re: [SystemSafety] Collected stopgap measures

I find this emphasis on some "proper" way to collect reliable statistics on the effectiveness of
software development to be a little odd. Barry Boehm has written two rather large books, one 37
years ago, and a follow-up a couple of decades later, establishing some basis on which companies can
obtain useful objective quantitative information on their software development processes.

None of them involve performing experiments of the sort Derek Jones is advocating, where you have
lots of teams developing software for the same task in parallel and then comparing results. That
seems to me, if I may say so, a pretty stupid way of getting such information. Besides, it was shown
by Knight and Leveson decades ago to be flawed.

PBL

Prof. Peter Bernard Ladkin, Bielefeld, Germany
MoreInCommon
Je suis Charlie
Tel+msg +49 (0)521 880 7319 www.rvs-bi.de<http://www.rvs-bi.de>





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20181117/144febdd/attachment.html>