The following describes the software development cycle that is carried out internally at Athento for the development of the product.
- A functionality request is received from Canny, consultancy or internal.
- The requirement analysis is performed and documented in the task.
- According to priority, the task enters the Scrum planning (normally every 4 weeks) and is analysed by the product team to resolve doubts regarding the specification, propose a first design and estimate its effort.
- At the time of coding, a branch is created for this functionality in the GIT repository
- When the functionality is implemented, a Pull Request is created, each of them has to pass the following checks.
- Have more than 3 approvals.
- Have more than 4 successful builds
- No failed builds
- No builds in progress
- At least 2 approvals by default validators, who are people who have at least 4 years of experience in product development.
- No change requests
- Have evidence available
- Have documentation
- Each build contains:
- When the PR passes all checks, it is merged with the main branch (fast-track) and starts to be deployed to the instances. Normally, all QA/UAT instances are automatically updated every night, and some production instances are also updated. These updates allow to detect any bugs that have not been identified during the development and testing process.
- The rest of the instances are updated approximately once a month.
Considering Athento's development process, can there be bugs in an instance after upgrading a version?
Athento's development process is always looking to increase quality and to improve when it is detected that some functionality has not been of the necessary quality. Athento works continuously to ensure backwards compatibility whenever possible and to increase the amount of testing.
However, since Athento is a highly customisable and configurable tool, testing all possible configurations becomes an exponential task.
For this reason, Athento cannot ensure that there are no bugs after an update and recommends:
- Keep instances as up to date as possible, as the longer an instance is out of date the bigger the change in the next update may be.
- Report as soon as possible any incident you detect following the instructions in the article: How to report a support case?