Testing is often the ugly step sister of IT. So it is often marginalized and sometimes forgotten.
Lately I have seen the opposite effect and that scares me as much as the other. One of the indicators is in communicating what testing is required. For me this is the dreaded word "Regression Test".
As systems evolve and the evolution of object oriented programming and service oriented architecture (SOA) evolve into the enterprise we need a new way to communicate the requirements for testing. Let's take an example...
I build a simple service to find out average shopping basket size a customer buys at Starbucks for their loyalty card to try and help the associate/barista to up sell me a bit - suggest an item I might be predisposed to add - for me it is a biscotti.
Starbucks goes and builds a service within their customer master file / customer relationship management system or just right from their enterprise data warehouse. Project X goes and builds this service for $250k. Now it is time for integration and user acceptance testing. What do I need test and where do I stop. This is where someone saying we will need to regression test this drives me nuts.
So where does that start or stop. The purpose of our component architecture is to decouple our systems to make them more independent.
- I should have to do a request response test with my source system.
- I should have to do a request response test with my user interface.
- Load test on end to end of above
- What else... some or all of the rest can be tested in production
That's the problem with regression testing as a word for me, it allows someone to say I need to test downstream systems that also touch the data to see if they are effected, or other source systems and so on.
If I need to make this item to shiny and perfect it will never get there and my project cost of $250k will all of a sudden be $1.5M. There goes my business case, agility or chance at innovation.
So don't get me wrong, we do not want a train wreck, we want to be able to control and master our risks and evolve our thinking. I know a guy other than me who thinks this way and he saves his corporation millions of dollars every quarter and also accelerates his innovation and delivery and so far knock on wood has not reduced his quality unlike some people think will be the result.
Now I will step off the soap-box. It's all about Rapid Results...