Checklist: Certifying a Transformation Project
05.03.2018 by Thorsten Reitz
In the past years, we’ve been asked by many customers to review their hale transformation projects. Sometimes users just want to get an opinion on a specific problem they encountered, but often, the scope was wider. Users are concerned at whether the projects they create will result in valid output. We’ve thus now developed a “certification” process for transformation projects. In that process, we check for a wide range of potential issues:
- Are there any performance issues?
- Was the source data mapped as far as possible?
- Were all required types in the target model created?
- Are there obvious semantic inconsistencies or mismatches?
- Are there any validation warnings?
- Does the execution generate any errors or warnings?
- How well are edge cases handled, such as invalid data or data variants?
- Is there sufficient documentation on the alignment and on the used data sources?
- Do all filters/contexts work as intended?
- When using Groovy Script functions or Custom Functions, are these correct? Do they pose a security risk?
- Are all references between features correct?
- Do Geometries use the correct model (e.g. Simple Features)?
- Do INSPIRE IDs and GML IDs use a consistent, best practice format?
- Are codespace values provided and correct, e.g. for
- Do they use valid codelist values?
- Does the output validate using the INSPIRE validator?
- Are Unit of Measurement attributes set correctly?
During the certification process, we document all findings on the alignment, typically using the Notes feature of hale studio, or the comments and tasks feature of hale connect. When all criteria are met, the customer receives a short summary documentation of the checks that were made and the changes made to address identified issues.
If you’d also like to have a guideline how to review your hale transformation projects, feel free to adopt the list above! Let us know if you have any other checks that should be incorporated!