[SystemSafety] Safety Culture redux

Mario Gleirscher mario.gleirscher at tum.de
Sat Feb 24 15:39:16 CET 2018


Dear Peter,
dear all,

Thanks for indirectly bringing up one of the many reasons why we started
our survey on finding out the use of techniques known as formal methods
(and, hence, many don't like them because of just the name which I think
is sooooo sad) in various realms of software engineering practice. Last
time I unintentionally triggered quite a discussion and would like to
apologize for my insistence.

Because Steve pointed at a classic (https://doi.org/10.1109/TDSC.2004.2)
terminology issue "bug vs. error vs. defect", I'd like to remind those
who haven't participated of our survey and kindly ask all of you to
forward the link below to software practitioners who might be eligible
to participate:

https://docs.google.com/forms/d/e/1FAIpQLScMbKk3SDY62fzBujtJA-i8XKREZP40MjXjIZvluZSVO9-QXA/viewform

Accompanying a mindset change through good terminology, we also need
empirical studies on finding out causes of the actual social phenomena
we are discussing. We have gained almost 100 responses, however, this is
by far not enough.

Systems are eating the world and there is no time at all to stick around
and rely on anecdotal evidence on "best practices."

@Steve: As far as I can see, the issue with static analysis is that
there is a wide bandwidth of approaches available under this name. Some
are good, some are outdated, some are inappropriate. The good ones that
we desire to use, I believe, rarely come at low cost. On the contrary,
these, even if sold as tools, require high skills and a lot of
experience to be effective.

Thank you and have a nice weekend,
Mario

On 23.02.2018 12:00, Peter Bernard Ladkin wrote:
> It is not just schools. I had a recent experience.
> 
> Recent graduate in math with minor in informatics, now working at a company in which programs and
> sells WWW-based services. They are using Scala, which has some functional programming facility. He
> thinks of using type theory (that is, mathematical type theory, not data typing) to improve the
> "correctness" (his word) of the programs. Are we interested?
> 
> Here is a simplified dialog:
> 
> Me: "Is there a requirements specification of the program you are interested in?"
> Him: "What's that?"
> Me: "It is a description of what you (or someone) wants the program to do"
> Him: "Oh, I don't think so"
> Me: "You used the term <correctness>. If there is no description of what you want the program to do,
> what did you mean by <correctness>?"
> Him: "It means the program is working the way we want it to work"
> Me: "But didn't you just tell me there was no description of that?"
> Him: "Huh?"
> 
> And this from someone who is, as far as I know, quite smart and quite capable. In any case, I did
> have the impression he was less interested in this aspect of things than in the prospect of playing
> around a bit with type theory. That's fine. (But it's not what I do any more.)
> 
> I have had such experiences regularly over thirty years of teaching.
> 
> PBL
> 
> Prof. Peter Bernard Ladkin, Bielefeld, Germany
> MoreInCommon
> Je suis Charlie
> Tel+msg +49 (0)521 880 7319  www.rvs-bi.de
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5053 bytes
Desc: S/MIME Cryptographic Signature
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20180224/60b18f5a/attachment.bin>


More information about the systemsafety mailing list