1

TL;DR

Do non-crashing bugs always affect Functionality, despite possibly affecting other quality attributes as well? Or is there a more specific quality attribute that non-crashing bugs always affect, like Functional Suitability?


This might seem like an ambiguous or unclear question, but keep in mind that core software attributes are very well defined by standards like ISO/IEC 9126, which describes:

  • Functionality
  • Reliability
  • Usability
  • Efficiency
  • Maintainability
  • Portability

There is a really large list on Wikipedia as well, but many of those approach the fuzzy, flexible land of the blogosphere. I want to keep this question unambiguous and bound to accepted definitions.

I'm looking for a software attribute that any non-crashing application bug affects. For instance, I consider a crashing bug to be a Reliability issue, but I'm not sure how to generally classify a non-crashing bug.

Correct me if you think I'm wrong, but here's why crashing bugs are most generally Reliability issues. If an application crashes (closes unexpectedly, freezes and doesn't recover, etc), it could be a security issue if protected assets are exposed or destroyed somehow. It could also be a safety issue if the application is, for instance, an X-Ray machine control panel. But a crash is always a Reliability issue, regardless of where else it falls.

But a non-crashing bug? Examples:

  • some valid input is treated as invalid

    EX: February 29, 2012 is incorrectly disallowed, but is a leap day.

  • some valid input is represented incorrectly

    EX: 11:59 AM + 1 minute is shown as 11:60 AM rather than 12:00 PM.

  • configuration or assets that are supposed to be persistent do not persist

    EX: User needs to log in every visit, despite selecting "Remember Me"

  • input is ignored in some cases

    EX: File extension is ignored if dragged-and-dropped

These all affect Usability, but that might not always be the case, so Usability is not a completely general classification. I'm leaning toward Functionality, because Functionality pertains to user needs and requirements fulfillment, and I think a non-crashing bug will always deviate from expected behaviour and/or correct behaviour. But something more specific than Functionality is desirable.

ISO/IEC 25010, the 2011 revision gives these definitions for Functional Suitability:

4.2.1
functional suitability
degree to which a product or system provides functions that meet stated and implied needs when used under specified conditions

Note 1 to entry: Functional suitability is only concerned with whether the functions meet stated and implied needs, not the functional specification (see C.6).

  • 4.2.1.1
    functional completeness
    degree to which the set of functions covers all the specified tasks and user objectives

  • 4.2.1.2
    functional correctness
    degree to which a product or system provides the correct results with the needed degree of precision

  • 4.2.1.3
    functional appropriateness
    degree to which the functions facilitate the accomplishment of specified tasks and objectives

    EXAMPLE:A user is only presented with the necessary steps to complete a task, excluding any unnecessary steps.

...which does seem to fit. I can't fathom a non-crashing bug that doesn't affect either Functional Completeness or Functional Correctness or both.

And finally, by bug, I'm referring to all three steps in the dependability chain: the fault in source code, the erroneous program states caused by the fault, and the interface failure caused by the error.

Whew, that was a lot of work for terminology's sake!

  • 5
  • Indeed! That's why I've specifically referenced the relevant ISO standards and favored omission of the relevant wikipedia and blogosphere pages. I'm looking for answers in the language of ISO, specifically. – kdbanman Sep 4 '15 at 18:09
  • 3
    You're missing the point. Nobody except for the most ISO focused pedants actually think of software this way, and categorization of bugs like this is a waste of time at best. All these attributes and names are made up. It doesn't matter what label you give to a bug-free experience; it just is. – Telastyn Sep 4 '15 at 18:40
  • 1
    Hm.. I sense some hostility. I meant no offence. I guess I'm an ISO focused pedant looking for other ISO focused pedants. I use quality attributes for requirements and maintenance prioritization. My team understands many of the ISO words, so they really do help keep us organized, regardless of how made up they are. For example, in the comment you just wrote, every word you used was "made up", but it was still an effective piece of communication. – kdbanman Sep 4 '15 at 19:09
2

I would call non-crashing bugs a "Reliability" issue as well, because you don't just have to be able to rely on a program to stay up; you have to rely on it to do what it's supposed to do.

Imagine a program that never crashed but sometimes thought that 2+2=5 when certain edge cases were hit. I wouldn't consider such a program to be reliable. Would you?

  • Thank you for humoring me. I buy this. To refer to ISO 9126, Reliability is about system functionality while subject to specific conditions, implied or explicit. The "specific conditions" clause is what makes bugs a Reliability issue. A calculator app could guarantee that "the + operator performs addition for all signed 32 bit integral numbers.". The "+ operator" is the functionality that sits under Functional Suitability, and the "all integral numbers" is the condition that places the 2+2=5 bug under Reliability. – kdbanman Sep 4 '15 at 20:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.