[SystemSafety] C for OSs

[Steve Tockey]

Olwen wrote:

"Sorry for frequent small postings ... things are occurring to me in fits and starts ... :"

Not a problem at all, it's actually easier to respond to individual ideas separately.


"Back in the day, requirements elicitation was seen as a key skill of the systems analyst. Emphasis on this as a skill seems to have waned and it is, AFAI can see, often glossed over - if not entirely ignored - in much of the literature on high-integrity development."

It has definitely waned everywhere, not just in high-integrity development. The agile community seems satisfied with "user stories" that are no more and no less than WBS items. FWIW, I have almost 20 years of survey data that indicates that the biggest complaint in the development trenches is "vague, ambiguous, incomplete requirements". It is admittedly the #1 problem and yet nobody (else) seems to want to do anything about it.


"Maybe more emphasis should be given to it in the professional training of software engineers."

IMHO there's no maybe about it. More emphasis MUST be given.


"Another old hobby-horse of mine is the near-universal and abject incompetence of engineers in the practice of writing clear specifications in natural language. Courses on technical writing should be part of every engineer's training. IMO, a great part of the problem of poorly captured requirements is due to the lousy writing skills of the average engineer."

Two comments on this. First, an English prof of mine said, essentially, "An inability to write clearly is a sure sign of an inability to think clearly". I have to agree. People who I consider write good code also write clear English. People who have a hard time writing clear English also write pretty crappy code. It's a 1-to-1 correspondence. I would consider it a fair element in an interviewing situation to have the candidate either submit a two or more page composition or give a 30-60 minute oral presentation, with the intent of my being able to judge how clearly this person can think.

The second comment is that, IMHO, "clear specification" and "natural language" are an oxymoron. Natural languages have built-in ambiguity (which is what makes poetry work, when you think about it) and they are inherently verbose. None of the recognised engineering disciplines depend on natural language specifications. None. They all figured out a long, long time ago that natural language specs don't work. They have all developed discipline-specific specification languages so that they can communicate precisely and concisely.


-- steve



From: systemsafety <systemsafety-bounces at lists.techfak.uni-bielefeld.de<mailto:systemsafety-bounces at lists.techfak.uni-bielefeld.de>> on behalf of Olwen Morgan <olwen at phaedsys.com<mailto:olwen at phaedsys.com>>
Date: Wednesday, October 23, 2019 at 4:57 AM
To: "systemsafety at lists.techfak.uni-bielefeld.de<mailto:systemsafety at lists.techfak.uni-bielefeld.de>" <systemsafety at lists.techfak.uni-bielefeld.de<mailto:systemsafety at lists.techfak.uni-bielefeld.de>>
Subject: Re: [SystemSafety] C for OSs



On 14/10/2019 20:48, Steve Tockey wrote:

Most use of Agile today is as a project management process band-aid trying to solve a requirements problem.


Sorry for frequent small postings ... things are occurring to me in fits and starts ... :

Back in the day, requirements elicitation was seen as a key skill of the systems analyst. Emphasis on this as a skill seems to have waned and it is, AFAI can see, often glossed over - if not entirely ignored - in much of the literature on high-integrity development. Maybe more emphasis should be given to it in the professional training of software engineers.

Another old hobby-horse of mine is the near-universal and abject incompetence of engineers in the practice of writing clear specifications in natural language. Courses on technical writing should be part of every engineer's training. IMO, a great part of the problem of poorly captured requirements is due to the lousy writing skills of the average engineer.


Olwen

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20191024/8897bb4a/attachment.html>