I think this misses the point in a lot of ways. There are showstoppingly huge practical issues with schematising data, including standardisation, version control, schema access, schema localisation, schema politics - who owns the definitions? - and many more.
But there is a valid point about data ownership. A lot of current privacy issues and political problems are created by the way that applications own your data. If I want to access a Photoshop file ten years from now, chances are good I'm going to be paying Adobe a fee to do it - likely in the form of a subscription.
This is crippling, because it means Adobe (MS, Apple, Amazon, etc, etc) have a choke point over my own personal access to my work.
Of course I can export documents to some other format, but in principle what I can and can't do with my work is controlled by big corporations.
Open source alternatives don't fix this, because outside of developer tool space they're usually poor competitors and never have the leverage - nor the quality - to become an industry standard.
At some point this data siloing became a "personal" computing principle, which is unfortunate because it undermines the idea of a computer as a personal tool.
OS schematising wouldn't really change this. But forcing open access to the internal file structures used by large corporate applications would be a game changer, because it would allow shell-like automation and composition of data that is currently trapped behind corporate paywalls with either no low-level access at all, or high-friction save/load only import/export features.
But there is a valid point about data ownership. A lot of current privacy issues and political problems are created by the way that applications own your data. If I want to access a Photoshop file ten years from now, chances are good I'm going to be paying Adobe a fee to do it - likely in the form of a subscription.
This is crippling, because it means Adobe (MS, Apple, Amazon, etc, etc) have a choke point over my own personal access to my work.
Of course I can export documents to some other format, but in principle what I can and can't do with my work is controlled by big corporations.
Open source alternatives don't fix this, because outside of developer tool space they're usually poor competitors and never have the leverage - nor the quality - to become an industry standard.
At some point this data siloing became a "personal" computing principle, which is unfortunate because it undermines the idea of a computer as a personal tool.
OS schematising wouldn't really change this. But forcing open access to the internal file structures used by large corporate applications would be a game changer, because it would allow shell-like automation and composition of data that is currently trapped behind corporate paywalls with either no low-level access at all, or high-friction save/load only import/export features.