My top 10 rants of 2007


I haven't "ranted" in a while (note blog title), so I thought I'd catch up in one post with the top 10 things that got under my skin in 2007. I have ranted about some of these things before. Others will likely find their way into posts for 2008. There are a few that I will probably never mention again. So, in no particular order, here are the top 10 situations and scenarios that bugged me the most in 2007 (at least that I can remember).

  1. "Schools" of software testing. I have a lot of respect for Bret Pettichord, and found his paper on schools of testing interesting. However, every tester I run into is only in one of two "schools" of software testing; the school that labels themselves context-driven, and those who don't care. I've never met anyone who said they were from another "school" of testing. The reason is that all good testing is context-driven. Crud, just about anything I do that I want to excel at is context-driven.
  2. Exploratory vs. "scripted" testing - to be clear, I love exploratory testing. The problem I have is the never-ending arguments pitting exploratory testing against "scripted" testing. Like context-driven testing, I believe all good testing is exploratory, and that exploration is an important part of creating any sort of automated testing. This leads to...
  3. The notion that test automation == dumb scripted testing. While it is possible to write simple test automation that does nothing more than "walk" the application through a scenario, good test automation, can check numerous items throught the test - including both things that a user can see, and things they cannot. Automation doesn't replace human testing (although I'm not exactly sure how a human tester tests something like an API or a filesystem without extensive automation and exploration). At any rate, the argument is ridiculous and gets far too much attention. This leads to...
  4. People who think test automation is only UI automation. When I talk about test automation, nine times out of ten, I am referring to automation of an API test or object model, or some other programmable piece of code. Every time I see an argument against test automation, it inevitably degrades into a discussion about how difficult it is to test UI. Yes, automating UI is tough, but that's not the point. Sheesh
  5. Developers who don't care about quality. Too many developers expect testers to find bugs in their code. Worse yet, some actually get mad when testers find bugs. Devs should be embarrassed when bugs are found in their code, and should have the intention of writing bug free code.
  6. Anyone who thinks quality is the burden of the tester. If <insert name of application here> has bugs, it is not the fault of the test team. It never was, and never will be. Stop thinking this way.
  7. United Airlines. I never said this post was going to be entirely about testing. I put in a lot of air miles this year, and although my only mileage plan is through United, their service and amenities were the worst of the bunch for me in 2007.
  8. Bloggers who only post links. Some bloggers base their entire blog on linking to other bloggers. Sometimes they link the same blogger over and over (and over) again. I guess a better idea would be for me to stop reading those blogs (done - it was time to clean up my blog list anyway).
  9. Ill structured arguments. Similar to my last post on ill structured arguments, is a trend I have seen recently is to generalize any single point of failure as a wide spread failure. E.g. arguments such as "I've seen X fail, so X always fails". Replace X with test automation, code coverage, unit testing, etc. I've seen it all. It's the same as saying "I've seen some people fail driving with a manual transmission, therefore nobody can drive with a maunal transmission" It just doesn't work.
  10. Last but certainly not least (and probably most appropriate is Whining Testers. Beyond my whining above, I'm tired of hearing testers whine about not getting respect, not being listened to, etc. If it sucks so much, do something else, or get another job. Yes, I'm being a little flip, but it's time for this argument to end (I have been sitting on a full length post regarding this "rant" for a while now - I'll pull it out sometime early in 2008).

And that's it. I'm not looking to start any more arguments - I'm just happy to have a place to air all of my gripes. Happy new year.


Comments (6)

  1. Happy new year, Alan!

    And I can’t remember the number of times I have already said – "Whining about your job? Why don’t you get a new one?"

    I am with you on that one.

  2. MSDN Archive says:

    Excellent post, and I agree with almost all of your rants. In particular, "context-based" testing is an absolutely ridiculous concept (well, it’s not even  concept at all).

  3. Peg C. says:

    This was almost better than chocolate AND it made me laugh.

  4. thadb says:

    I agreed with a lot of what you had to say, except for the Schools of Testing comment. While it may be true that only people in the Context-Driven School identify themselves as such, I work with a lot of people who’s mindsets are glued to one of the other schools. Getting them to think in a contextual manner is like trying to pull a lion’s teeth…It may benefit the lion, but you’re gonna lose an arm in the process. Understanding what kinds of core values they have has enabled me to be able to show them how we can update our processes and techniques while still giving them what they perceive that they need. Using the core values of the Context-Driven school has actually allowed me to end the argument of "scripted/automation/regression vs. Exploratory" and move them towards a more cooperative test effort. I think one of the major issues I have is that I’ve seen people use the argument "All good testing is contextual" to excuse lazy, sloppy testing that reverts to the idea of a universal Best Practice rather than think about what they’re testing actually needs to accomplish.

    Other than that, it was a great article.

  5. Alan Page says:

    thadb – Thanks for the comment.

    My point (and perhaps this is a circular argument) is that the testing being done by those with different mindsets is not "good testing". Just about any endeavor done from a narrow minded focus will be less effective than an approach that considers context.

    I took a drivers education class in high school. I got to be a much better driver through practice, and by reacting to context when I drive (e.g. when it’s raining, I drive slower around corners, and I watch for standing water). I don’t consider myself a context-driven driver – I’m just a good driver.

Skip to main content