Tuesday, October 21, 2014

Two things that bother me on #stop29119 discussions

Let me start with my stance on #stop29119 - I have signed and I think everyone in any way impacted by testing in software should sign it - customer/product organizations trying to succeed on software business in particular.

But still, in all the discussions there's two points that bother me and a lot of the critique revolves around.

  • 'Standard' the word and its definition
  • Available options for the contents of the standard

'Standard' the word and its definition

A few days back, I saw Lisa Crispin tweet a reply to Michael Bolton that I'm strongly paraphrasing from my memory. The reply was related to a discussion on agile testing as per Lisa's new book and pointed out that not everyone appreciates time spend on word games as they have a full-time testing job to attend to. It caught my attention, as it was pointing out a problem I feel I'm having with context-driven testing arguments, and a problem that is in the core of discussions for #stop29119.

Many people suggest strongly that there should be no standard. Standard the word strongly implies a connection with the regulations, and through that, compliance and lawsuits of non-compliance. I get that. I agree that standard is a very risky word to use if we mean 'guideline'. But I think I hear the other side on this as well.

I live in Finland and Scandinavia is quite different from USA. Really. The whole discussion about standards isn't that big of a monster in the world I live in. It becomes a monster occasionally with EU regulations and it's always been a monster with things regulated from USA. But in the little fluffy cloud I get to live in, it just doesn't matter that much. I'm sure its not equally bad to everyone in USA either. But in this particular fight for #stop29119 the definition of that word becomes a key issue.

Standard the word has many contexts. After all, we're talking of context-driven testing with testing area not having set meanings on the terminology. Words are communication between people, and for testing terminology we don't believe in set terminology but trying to understand and hear out what the other party is saying. How come the word standard is that different from all the other words we use? How come we can't accept that standard in one context is regulation and standard in another context is guideline.

The fact that this guideline ISO 29119 is totally worthless contents that make things worse is a different story. But the basic idea of attacking the word standard seems like word play I don't even care to win. The risk of compliance requirements is relevant for certain contexts.

Available options for the contents of the standard

The other point that bothers me is how many of the opponents seem to handle the critique of not being constructive as in offering real alternatives. From the friendly, yet disagreeing local discussions I've had, I've caught a strong feeling for the organisations driving these standards, there's a real need of a box they label ISO 29119 but that the contents of that box could be something other as well. It might be that I'm optimistic, but I don't like the answer we tend to give on what to put in the box: nothing. We don't do standards. We don't believe in them. I don't believe in them. But I can respect that the idea that drives my local colleagues towards a standard is a true belief in helping out the ignorant in getting good testing. That box needs contents.

I have a personal suggestion as to what to put in that box. I want to put Rapid Software Testing in that box. But to do that, we'd have to get past the idea of the word 'standard'. 

I have an option for the content of the standard. So, the ball is again on the standards organisation: if context-driven (or rapid) testing is no different from the current contents, how about changing it radically. 

Testing starting from that base of values and ideas would be much more relevant that the paper-piling approach the current contents drive towards. It would be helpful to take things forward for the world of agile and still remain relevant with the more traditional mechanisms. We can't say the same about the current contents.

Win-Win?

To get anywhere from here, we need to start making compromises. Compromises to how strongly we feel about the word standards, compromises on how serious we are about no standard-like documentation about testing can exist. The standardisation organisations need to win too, it's not exactly fair to say we want them to stop doing the business they are in. Or is this really something that can, by no means me resolved?


Disclaimer: I'm tired of being attacked for the choices of words or my opinions. This post is very much my opinion, and I'm not going to stress over my choice of words for other than my good intent. I'm happy to engage in discussion that aims at mutual exploration of what (and why) I think that changes my mind. But please, picking on semantics of words without the intent of communication - let someone else do that.   And reminding me on the fact that I'm not a native speaker is somewhat insulting. I just don't care for the exact words, I care for discussion and understanding of agreement or disagreement - communication.