Advertisement

rumpsteken

image

Biography has not been added

rumpsteken

's contributions
Articles
Comments
    • It is funny how you did not include the very next sentence after the part you quote: "... 25 years of use. Although hard real-time systems and scientific subroutine libraries are not the same beast, ..." You should also know that defects is not the same as software quality. Obscure code can be free from defect, but still be considered bad. Consider this quote from Donald Knuth; "Programs are meant to be read by humans and only incidentally for computers to execute." Also Hopkins and Hatton (2008) does not mention global variables. However, that paper will be extended to include them.

    • I do not think it was misleading. It was clear that he made very broad estimations when talking about dangerous failures. To consider the worst case and say that all arbitrary failures are (potential) dangerous is okay from my perspective. What percentage of arbitrary failures do you think is fair to consider dangerous? If anything is misleading it is to do unfair comparisons. A 100 line program with one global variable is easier to understand then one on 10 000 lines with 100 globals, even if the ratio globals/lines of code are the same. Also, the MISRA standard is intended for embedded/automotive systems and should not be expected to be used in other contexts. I guess you disagree? I think that it is very interesting that you have a different view that seems more 'legal'/law focused. This is an interesting topic.

    • I found a document on safteyresearch.net very interesting. It presents some really horrific stories about unintended acceleration. [1] With this in mind I think what we can read on page 3 in David’s case study [2] also interesting: “… And as we become increasingly reliant on embedded software in our daily lives, for example, with the advent of self-driving cars, trucks, and automotive “autopilot” features, our legal system is likely to face this issue with increasing frequency.” I have the opposite opinion that these kinds of incidents and trials should decrease. This due to development of good development processes, increased knowledge, use of best practice etc. Feels very wrong that obscure, closed source software, could be used as legal defense to avoid responsibility. [1] Embedded Software under the Courtroom Microscope – A Case Study of the Toyota Unintended Acceleration Trial - http://www.kellytechnologygroup.com/main/Cummings-IEEE-Tech-and-Society-Article-Dec-2016.pdf [2] Appendix A: Toyota SUA Incidents Resulting in Fatalities Included in SRS Analysis - http://www.safetyresearch.net/sites/default/files/toyotasuaanupdate_appendices_20101025.pdf