Continued development of existing (and validated) apps - General approach
We have a bunch of apps, keeping them as few as possible helps maintenance, monitoring and overview.
But, we are continuously met by user requests where they want to add new metrics to existing apps, data that could exist in other fact-tables at different granularity and maybe sharing some dimensions.
Is there like a defined standard on how to progress with such a development? Altering the existing (and validated, tried and tested) load-script feels odd... Like if we have one big concatenated qlik-fact from several different DW-facts, adding to that fact could jeopardize the structure of the entire app. Even more so if we have ONE fact and need to add one more.
Makes me feel like duplicating the original app, remove the old sheets, and then make the needed changes for the specific scenario is the way to go. But that could generate a lot of apps... Am I over thinking this? Any thoughts and general pointers are appreciated.
Use your existing data model as a centralized version that is tried and true. Whenever you have request for new applications you could binary load this dm into a new app and apply changes there. Once you have that created you can run the impact analysis api to track the number of apps that relying on the binary DM.
Use your existing data model as a centralized version that is tried and true. Whenever you have request for new applications you could binary load this dm into a new app and apply changes there. Once you have that created you can run the impact analysis api to track the number of apps that relying on the binary DM.
If you're looking for a structured way to manage ongoing Qlik application development, it's worth checking out SenseOps Code Management. It combines version control, app comparisons, deployment workflows, and regression testing to help you continue developing new features while keeping production versions stable and reliable.
