Tuesday, August 31, 2010

What a developer wants?

After so many years of driving testing and QA processes I have some possibilities to realize and rights to publish here some of ideas on how things are working from the inside. One of such things is what developers ask of a QA manager.

My top 5 are following:

1. Change defect tracking system

2. Stop reporting duplicates and non-defects

3. Do no find defects too late

4. Stop reporting minor/cosmetic defects in such a number

5. Test design is not review-able

Now back to the details =)

Change the defects tracking system I am always happy when people come to me with their ideas which they think will help them doing their work better. However, I don't like when it stems from the "I don't know what is wrong but I feel that something needs to be changed" intention.

In the past, developers came to me complaining that it takes too much time to work with defects. The only solution they saw was to change the tracking system. I asked in respond what exactly does take time when they work with it. As it turned out the reason of their complaints is not in the system per se but in the time they had to spend understanding the defect and trying to reproduce it. No defect tracking system can change that, so they went out in peace with the idea to leave the old system. Others developers came after... period =)

Stop reporting duplicates and non-defects Yes, this is a problem. But let's see how big it is. On my memory we reported not more than 1% of bad defects. Is that so huge to be whining about at each managers meeting? Nope. Then what was it?... I'll tell you. This is an excuse for not doing own job right. Kind of "hey! they also have problems". It makes them look better in they eyes. What a shame! =)

Do no find defects too late Another problem for testers. We have to find serious problems as soon as they appear in the code. There is no excuse... Wait a minute! What issue are speaking about? That one that developers introduced in the same build? How do you think testers could find it earlier?... We hear such acquisitions too many times. Make sure you take the blame for your own mistakes.

Stop reporting minor/cosmetic defects in such a number The answer is "stop reading them". It's easy =)

Test design is not review-able In majority of cases they just don't care to put some efforts in this. This is merely a laziness that cannot be an excuse :) However, providing a summary of your tests design is always helpful. Remember, those who read your design help you making it better, so it's in your interest to make it easier to them.

Certification... Again?!

Today I responded to the question about certification one of IT professionals here in Belarus. She was interested in my opinion about the certification as it is. I once told to one of testers who came to me for the interview that it is important that one cares to have one. On another hand I said many times that I don't care if one has a certificate up his or her sleeves.

I will try to explain it here. Once and for all =)

Firstly, I repeat I don't believe in certification because it is a try to provide a universal approach and solution to everything. Who does believe that it is possible? I personally don't.

What I believe is personal abilities to think, to solve problems and to actively drive to the goals. This is what I am looking in people. Not just a prove they have heard something about the matter.

The only reason I said this is important is because a person who acquired it cared to manage her career, minded to get new knowledge, which are good intentions. However, reading a professional book would do the same, not to say it would be better than most of certification programs.

Hopefully, it helps =)

P.S. From business perspective it also doesn't matter. Testers with and without certification are equally profitable for the company, so the compensations are the same. The fact that the company has engineers with certificates does not help at winning new projects.