Solid is a noble effort but I am skeptical that it would solve the problem, for two reasons:
1. Application functionality is often deeply coupled to its data format. When building on a standardized data format, how does one add new, novel features? If you extend the standard, then you now have a non-standard data format, and the features you added won't be recognized by other apps that use the format. No one wants to wait for the standardization process to complete before they can ship a feature, so inevitably apps will ship various incompatible extensions. Then, when you try to move your data between apps, you find that a bunch of stuff breaks. There's no clear solution to this, other than for everyone to stop innovating, which obviously isn't what we want.
2. How does having standardized data formats suddenly prompt decentralization? Yes, it means you can more easily try out competing apps, but those competitors still face the same high barriers to entry. VCs don't like to fund the second player in a market, much less the tenth. Open source and hobby projects aren't suddenly able to compete when they still lack resources to run a service.
This blog post actually started out being specifically a response to Solid, arguing that decentralizing storage alone (without compute) is not enough, but I ended up not feeling great about targeting them so I cut that part out...
I, as a programmer, am almost always able to migrate my data from an app to another, by using APIs, the command line and tiny bugged scripts that fetch data from one service, reformat it, repackage it and save it to another service, which uses totally different formats.
Maybe a better solution would be to create easier ways to do that, so everyone will be able to move their data.
And in practice, the small, new entrants to a market will write migration tools to migrate away from the entrenched players as a way to get customers.
If this isn't happening, then there's apparently no will from either player to write migration tools, which suggests to me that there would similarly be no will to use standardized data formats, unfortunately.
1. Application functionality is often deeply coupled to its data format. When building on a standardized data format, how does one add new, novel features? If you extend the standard, then you now have a non-standard data format, and the features you added won't be recognized by other apps that use the format. No one wants to wait for the standardization process to complete before they can ship a feature, so inevitably apps will ship various incompatible extensions. Then, when you try to move your data between apps, you find that a bunch of stuff breaks. There's no clear solution to this, other than for everyone to stop innovating, which obviously isn't what we want.
2. How does having standardized data formats suddenly prompt decentralization? Yes, it means you can more easily try out competing apps, but those competitors still face the same high barriers to entry. VCs don't like to fund the second player in a market, much less the tenth. Open source and hobby projects aren't suddenly able to compete when they still lack resources to run a service.
This blog post actually started out being specifically a response to Solid, arguing that decentralizing storage alone (without compute) is not enough, but I ended up not feeling great about targeting them so I cut that part out...