The reason for me writing this particular post was one of curiosity. Often – not always – there are problems when using words that have a different meaning to the intended one. I found it curious that our way of using the word ‘counting’ is actually different to what it really means. So I wrote this post.

I agree with you that with the word ‘counting’ in most situations it doesn’t matter. What I find important is to that we’re aware of the difference and can take a conscious decision IF it matters in a given situation or not. In very many ways it’s a training exercise to look at words that are different from our intended meaning and find out where they come from – Epistemology in action.

LikeLike

]]>Hi, thank you for the reply.

Re-reading my comment, it came out more blunt than I have intended. I do apologize for that.

Anyway, the point I was aiming for was this –

Sure, the definition you gave is a very formal one, and I am not trying to dispute the correctness of it.

What I am wondering about is the value that we gain from applying such a formal definition to our situation. There’s a joke about a guy who gets carried off by the wind in a hot-air balloon, when suddenly he sees someone on the ground and asks him “can you tell me please where I am?” to which that person responds “you are inside a hot-air balloon, about 30 feet above ground level” (one version of this joke can be found here – http://www.poeticexpressions.co.uk/poems/Engineer%20v%20Manager.htm ). Applying the mathematical definition of counting to the situation of bug count seems to me a lot like this situation – it’s technically valid, but practically useless. All those philosophical definitions are inconsistent with the usual usage of “counting” (For instance, given enough time, I can count to infinity. it won’t do me good, but anyone can count any infinite countable set – that’s why they are called “countable”. https://en.wikipedia.org/wiki/Countable_set).

So is your claim about not being able to count bugs – the theory part is a nice idea to play with during your spare time, but for any practical need, bugs are countable the same way PI equals 3.1415926 (if you need more precision – http://www.geom.uiuc.edu/~huberty/math5337/groupe/digits.html ) neither of these is actually PI, but the difference it makes to any practical problem is negligible. As long as I can point and say “this is a bug”, it doesn’t matter to me that there is no strict definition of bugs, because I can define that “for the purpose of measuring, a bug is anything that the product manager agrees is a bug” or “a bug is any item in the bug tracking system that isn’t being rejected”.

So you *can* measure bugs. Whether you *should*, or what meaning can you get out of these numbers is another thing altogether. I don’t see a need to hide behind rigid definitions that say we can’t count it.

LikeLike

]]>thanks for commenting.

To adress your first question “What’s the value of playing with arbitrary definitions and then claiming that they are not enough?”

The definition of counting is not an arbitrary definition but a mathematical definition. I’d like to ask you to follow the link and read up on it, if you haven’t done so. My claim is that the way we are using counting for bugs does not satisfy that mathematical definition, which not everyone will be aware of, that’s why I wrote this blog post. I like to make sure that I fully understand what I’m saying and in this instance I found that I didn’t at first so I shared it with the community.

You also wrote “…every team has a firm understanding of what is considered a bug in their product(s).” If there would be a firm understanding there wouldn’t be a need for discussion. I may have a different outlook of what firm understanding is though, that’s where definitions are useful so that we can agree if or if we are not talking about the same thing.

“The fact that something can’t be strictly defined does not mean you can count it.”

That’s where I respectfully disagree. What you can count are bug reports as I stated further down in the article, no problem with that. You can call anything you wish counting but, and that is the important part, you can’t call it counting in a mathematical sense.

LikeLike

]]>What’s the value of playing with arbitrary definitions and then claiming that they are not enough?

I can count bugs. I know them when I see them, as does everyone else. In some cases, What I consider to be a bug isn’t considered as such by others,and then we can have a discussion. A bug is some behavior of the product that is deemed a bug by the people in charge of that specific product. In some companies it is useful to separate bugs from change requests, in others it is not, but every team has a firm understanding of what is considered a bug in their product(s).

I can count bugs the same way I can count the number of chairs in a room full with chairs, stools and sofas. There is no way to define a rule that will state “this is what a chair is” without including some stools or a sofa, or without excluding some proper chairs. The fact that something can’t be strictly defined does not mean you can count it.

I can also count really easily – one bug, two bugs, three bugs etc. Given enough time, I can also count to infinity (I currently am counting up to infinity – one increment every birthday, and I’m making a steady progress).

So, are we saying that counting bugs is meaningless? I can agree with that, but playing with formal definitions seems a bit off for me. People can, and do, measure the amount of bugs in the software, or, if you want, you can call it “estimating”. The result is similar.

LikeLike

]]>LikeLike

]]>LikeLike

]]>I would like to add “assumption”. Anyway, it’s a good list. Thank you!

I’ll go and copy it 🙂

-Thanh

LikeLike

]]>LikeLike

]]>LikeLike

]]>LikeLike

]]>