Monday, June 23, 2008

What do you do when you cannot get any code out of your developers

A few years back, while working for an R&D company, I was part of a group of commercial developers and testers brought on to bring a research project into production development. After quite a few assessments on the prototype, our architecture team decided to keep the underlying engines, however, completely re-engineer the UI and business-logic layers.

So, the technology for the UI and business-layer was eventually decided upon.

At this stage, I was the sole test team - so I worked closely with my team of 10 developers to spec out the entire application.

However, one of our lead developers got caught up in the technology. He had decided on a framework, gave the devs an overview of it - and set them on their merry way to code "the new world"........... until....... a few days into development, he decided there was a better technology to do what he wanted to do.

So, after re-scoping the framework, and re-educating the entire development team in the new technology he had found on the net the night before (and had spent all night working on it)...... he set the development team on their merry way again, to re-code the code they had already written, in the newer, better technology he had found....... until....... a few weeks into development, he decided there was a better technology / way to do what he wanted to do....

So, at this point, there were developers who were getting quite upset re-working the same area every time this guy changed his mind.... and then there was me - going out of my mind waiting for code to come through.

Yes, I worked with each of the developers in the development environment - as I couldn't get a release out of them while the framework was constantly changing - however, frustration was an understatement.

When I did get a release out of him - I heard those words that I had come to dread "hey guys, I saw this new advancement in last night, I think we should do it this way".... yes, here we were, back at square one....

So, with all business specs, SRS's, Functional Flow Block Diagrams, test strategies, test plans, test scenarios, test cases, etc written - I was here stuck with nothing to test. I couldn't start automating, as the foundations of the application were constantly changing - even the developers had given up on writing unit tests, for the fear of those dreaded words again......

So, I put to everyone out there - what do you do in situations like these?

Personally - I assisted the developers as much as possible, by doing pair testing with them (ie: sitting next to them in their development environment, and testing their code changes as they made them locally), but this was becoming time consuming for them, as they had to quickly make up for lost development time every time they had to re-work their code....

So, I started self-educating myself with everything and anything testing related I could find on the net - but this proved to become quite boring after 3 months of not getting anything to physically test.

So, out of complete frustration (boredom is a greater stressor to me than impossible deadlines) - I found another job and moved on.......

So, did I make the right decision? Did I do everything I could to make this job work for me?

I think so - if it wasn't for that move, I wouldn't be where I am today - which I believe, is a very good place to be ;-)

No comments: