6.4.2 Co-existence measures

What this heading covers

The population is a set of environments, each with named products and versions. A co-existence result without that list is not reproducible. Conflicts are often resource conflicts: a port, a file lock, a shared library, a device, a scheduled job. The measure should separate a conflict that stops operation from one that degrades it, because the two need different responses. The family is small but it is where a product fails in a customer environment after passing every test in isolation.

Record position: family 7 of 40 in clause 6, under 6.4 Compatibility, which carries 2 families. It is one of the 40 families a programme has to map its measures to, and its elements are listed below. The heading is 6.4.2 in the public contents; the text behind it is not public.

How a measure for this family is built

Count the environments in which the product operated without a conflict as the numerator, over the environments tested as the denominator. For a finer measure, count conflicts per environment and weight them by severity. Record the other products and versions in each environment, because the denominator is a list, not a number.

What to put in place before publication

1. Write the environment matrix before testing, including the versions.
2. Log resource conflicts separately from functional failures.

Evidence to keep

Two questions to ask

• Which other products and versions were in the environment?
• Were resource conflicts graded by effect, or counted as one number?

Where teams usually fall short

• A product tested on a clean machine and declared compatible with everything.

What the number cannot tell you

It cannot tell you about an environment you did not test. The result is a statement about a list, and the list is part of the measure.

What happens next

Interoperability is the other half of compatibility and needs a different environment record.