Wednesday, November 26, 2014

Handing over the project to maintenance

Finally the project is nearing the end and handovers are being planned, kick-out dinner is booked and the next project is being staffed up. But there are still some open defects, how can this be?
Shouldn't we deliver a complete software, there should not be any defects, or ?
As stated over and over testing can continue until the fingers bleed, and yet there will be new issues found or some issues might not have been solved. So how do you handle this ?
A good handover is the key, and as usual good communication. The receiving organization must be aware of the current status and if they can accept a delivery with open issues. Also there should be a plan for handling these open issues.
Usually some tool is used for tracking issues during development, QC, Jira or similar. But when it's handed over to maintenance, they might not use the same tool, or the issue handling process is completely different as the product is now in production.
Make sure the open defects are clearly described with steps to recreate.
Ensure that when handover is done that the receiving organization understands the issue and the possible impacts. Also make sure that they understand the issue, what seems obvious to you might not be for another person.
Good practice is also to do a follow up later to see how these issues have been handled. As it is easy to forget and if not fixed they will come back in one way or the other.

Wednesday, November 19, 2014

Emergency room

After spending an afternoon and evening in the hands of healthcare due to bike crash and resulting fracture in the arm, i started thinking on how this is similar to development and test.

The scenario.
I am biking home from work and colliding with a pedestrian resulting i a short flight and a stiff and throbbing arm.

Defect report was raised to “first line support” and providing scenario and symptoms. The initial investigation showed that it most likely was nothing serious and rest was prescribed.
So a minor defect and nothing that needed to be re-written or was a show stopper.

A few days later the arm still was not better so a new call to First-line was made and as the symptoms had changed slightly it was decided that a doctor should look at it.
The minor defect now re-appeared, and was re-opened, but in a slightly different way and a little different outcome.

Finally arrived at the doctors office and was meet by the nurses who made an initial investigation and decided that x-ray was needed. This was done and the images show to the doctor.
Investigating the defect to find the seriousness of it.

Quick meeting with the doctor confirms the fracture and a cast is needed
Product owner checks the defect and decides it needs to be fixed.

Once again meeting the nurses and getting a cast. Information on what to think about and how to handle the cast.
Defect fix information and a workaround provided by the developers. Final fix to be delivered in one to two weeks.

Conclusion
It is important to verify the defect, can it be recreated, and if so will it affect more than one area. A minor defect in one system might affect another part and then become a major.

So remember to test around the defect, as they tend to live in groups

Thursday, November 13, 2014

Stop29119 - A Tempest in a Teapot ?

It has been a few months now since CAST 2014 which seems to have been the starting point for the “great ISO29119 debate”. But what has actually happened since then ?

  • The Stop29119 petition was started and got over 1000 signatures and this was then presented to the ISO-group.
  • Numerous blog posts have been written
    • Mostly posts against the standard, but a few pro-ISO have appeared
  • Debate on Twitter and other media
    • Ranging from personal beef to factual discussions
  • Local discussions appears to have happened
  • Probably quite a few people have bought the standard (me included)
  • No information if any company has actually started to implement this

But what now ?

  • The debate seems to have abated and things have gone a bit silent.
  • No response to the petition, at least to my knowledge
  • We are still waiting for the two final chapters of the standard.
  • Awaiting to see if any organization will pick this up or not

So was the whole debate a storm in a water glass?

Well in a way it was. Probably because only a small number of people did the open discussions.
In another way it was not. It brought the attention to the general community on what is going on and even if most stayed silent, many local discussions seem to have happened. And the information has been spreading.
Some discussions might have gotten a bit out of hand and might have lost focus, but thats normal when you really stand for something. Then again we are testers and discussing is what we do, and question how things are.

Looking forward to read the last chapters and also to see what will happen both within the community as well as within organizations in regards to the standard

Monday, November 10, 2014

Hiking Analogy

After doing a hike up a mountain in near white-out conditions it got me thinking that this is a good analogy for approaching a new system-under-test.

IMG_3929.JPG

You know what it is supposed to look like, what the system should do.
Up ahead is the goal, the end product.
Right now you can only see 100 meters ahead, you see the current function.
You will encounter cliff face - defects or issues

IMG_3931.JPG

Once at the top you’ll see the big picture can say you have a good view of the system
In order to really know the trail/system you need to do the hike several times in different conditions, it might be hard going but once done it is a rewarding feeling


IMG_3932.JPG