The Information Architecture of large organizations is a mess. Until we recognize and take action on the core problem, the situation will continue to deteriorate - from here
It won’t come as a surprise to those who know me well that I am passionate about freeing people’s health data from the constraints of applications and proprietary (or even just local) formats.
Well, it isn’t just health that this applies to. Not by a long shot! This manifesto is a multi-industry push to a new way of ‘doing’ IT. The key principles outlined:
Data is a key asset of any organization.
Data is self-describing and does not rely on an application for interpretation and meaning.
Data is expressed in open, non-proprietary formats.
Access to and security of the data is a responsibility of the data layer, and not managed by applications.
Applications are allowed to visit the data, perform their magic and express the results of their process back into the data layer for all to share.
I encourage you to take a (brief) read and consider joining the 1193 folk from around the world who have signed it. I have.
I would have hoped points 2 - 4 did not even need expressing (computer science 101 since the 90s), but apparently they do. I don’t understand point 5 - this point seems to me like a business logic layer (BLL), which should be kept clear of a clean data layer. I agree with the concept of separating UI from the DAL and the BLL however.
What I have seen from Orion Software seems to fundamentally violate point 2. I would love to be corrected/put straight as to if this statement is misguided or out of line. Field names are in no way self describing, normalised relational structures are not simple, defined or obvious and finally numeric data is often stored as strings and regional datetime data as complex multi-field structures - which by definition is not self describing. Each data entry “page” seems to have its own table - even if the pages are only broken up to improve ease of use for the person entering the data. I worry that many DHBs are now so enmeshed with such data formats that extrication is too large, expensive and risky a task to consider. As you say - this is a common problem in many industries.