Following up on the SSL outage on the default Gaia hub at https://gaia.blockstack.org, I’d like to take this opportunity to resume the conversation on diversifying the set of available Gaia hubs.
When you onboard with the Blockstack browser, your data gets stored automatically to Blockstack PBC’s default Gaia hub. This is due to the way the sign-in process works in Blockstack – before you can edit your profile or sign into any apps, you need to be able to store data somewhere. This means you need a Gaia hub running somewhere before you can do anything else.
To make it easier for users to get started using apps, Blockstack PBC simply provides a Gaia hub for general use. Unfortunately, but also predictably, this resulted in nearly every user using the default Gaia hub, which caused them all to be affected by the default Gaia hub’s SSL outage. Let me be clear about this–this is not the outcome any of us, especially us at Blockstack, want to live in.
What we’re trying to figure out instead is how to reconcile the need for users to have a good on-boarding experience with the need for users to keep their data in different Gaia hubs. This is already supported at the protocol level, but it isn’t easy to get users to do this since they don’t have any incentive to run their own hubs. I have a few ideas about how to provide better incentives, but I’d also like to hear from the community. What would it take to get you to run your own Gaia hub?
It seems pretty easy according to your old topic, but maybe I don’t have an Amazon S3 thing, and instead want to use the local box that it’s installed on as the storage? Or perhaps my Google Drive, Dropbox, Mega.nz, or Microsoft Azure cloud storage thing?
I think adding more storage drivers (gaia extensions!?) would make a huge difference and make it a lot easier, at least for running it personally or for a small circle of friends. Making migrating easier would also make a huge difference (CLIs can be scary!).
For running a public gaia hub, if it was simple enough to start up, I think it could be recommended for big apps to also host a gaia hub of their own (think postly or stealthy running their own public gaia hubs you could use). Granted this sort of defeats decentralization but users could always have a choice to migrate, of course.
Another idea I’d like to explore at a different time would be the ability to have primary/secondary gaia hubs in case one goes down, or even have the end-user running their own local gaiahub to store/read their own information, whereas their “friends” would need a (different) remote gaiahub to read it, etc., though that has its own problems to deal with in its own thread (where does profile.json go?!).
I actually think there are many power users and maybe even applications that would be willing to run gaia hubs that act as an alternative or replicated version of default storage. There are costs to this, and I don’t see users paying to store files on Blockstack unless it is in a storage system they control. So to offset the costs of running a gaia hub that anyone can connect to, verified hub deploying users/apps could be compensated in Stacks tokens when the network goes live.
There are probably other compensation and incentive models to be fleshed out as well, but I think something that shouldn’t be overlooked is that while, yes, it seems relatively simple to run a hub, I expect the documentation for doing so would need to be even clearer. Tutorials (though they easily get stale) would be necessary.