Catchy Im Back Phrase

"Out of sight, out of heart." A problem that is not visible is not treated .

On the advice of Régis Medina and after attending his lecture at the Agile Tour in Valencia, we are looking for a few iterations to make our issues visible not to let the trainer. For this, we have implemented 2 practices.

1. SEE THE PROCESS OF DEFECTS

The first, already mentioned in connection with a ticket on the resolution systematic problems is displayed on a table dedicated problems encountered by the team (see photo ).

2. SEE PRODUCT DEFECTS

The second is to identify, measure and display daily and automatically non-quality product. It is the sum of all that is correct to the level of quality we are looking for our product (and we are VERY demanding) .
Thus, we urge

• the number of warnings compiling code and tests ,
  
• the n shadow of operations that exceed a threshold of complexity ,
  
• the number of classes not covered 100% automated tests ,
  
• the number of classes that do not meet coding standards ...

history of this metric enriched a curve which is shown daily at the place the open space where there is more traffic, right next to the burndown charts.

Since we decided to make visible these problems, we put so disciplined and continue to reduce this non- quality, as shown by the curve shown.

Once the curve stabilizes at an acceptable level of non-quality (corresponding to a minimum of work in progress) we enrich this indicator of a new class of defects. You will see a leap in hand commented on the displayed curve. Thus, we are raising our level of demand in a controlled manner.

We're not interested in the value of the metric but rather the slope of the curve . Indeed, it reveals very clearly if the team is improving or break.

It is very interesting to view this curve of non-quality alongside burndown charts. Indeed, one often finds that the improving productivity (visible on a burndown chart) comes at the expense of quality (visible on the curve of non-quality). The day view and side by side these two curves can see that we do not play in communicating vessels. Better still, we come to correlate even better than work permits to work faster !

Finally, we also valued this measure of non-product quality. For each category we have identified defect associated value: This is the theoretical time required to correct a defect in this category. Thus, we measure the estimated time needed to obtain a product of the desired quality.

With daily measurement and the combined remaining to and non- product quality, we have tools objective, relevant and complementary to steer our development .