Alex Schroeder 🐝 is a user on octodon.social. You can follow them or interact with them if you have an account anywhere in the fediverse. If you don't, you can sign up here.
Alex Schroeder 🐝 @kensanata

I just stumbled over project Solid by Tim Berners-Lee – a “decentralised internet to give control of user data back to the people.” Do you know more? There are various general pieces out there, and a blurb here: csail.mit.edu/research/solid-s and some more info here: csail.mit.edu/research/solid-s
I guess I’m looking for a critical discussion. Any reading suggestions?

Gaaaaah! More info here, of course: solid.mit.edu/ – something about copying and pasting URLs doesn’t quite work as it should on this phone of mine.

@kensanata It's over-engineered and too complex for wide adoption imo. Don't have recommendations for good critiques though. I think most people are just excited because it's timbl's project.

@kensanata My first thoughts looking at it?

SOcial LInked Data? I'm quite happy with ActivityPub for that.

@kensanata Also with it being based on HTTP I worry that it would face the same centralizing pressures as the rest of the web.

ActivityPub does pretty well this way with it's local timelines, even if it's still based on HTTP.

Just something to ask about any decentralized technology: what's keeping it decentralized?

@alcinnz It looks more like a generic data store on the web. Any application can use it. It doesn’t seem to do much else. Since every application can use it, you don’t need import and export functionality. Data exists separate from applications. This is a cool idea. Each Mastodon instance uses its own database, and provides import and export. If an instance goes down, no more data export. Solid night be a solution to this (very specific) problem.

@kensanata @alcinnz
The Solid blurb suggests it is aiming to be a personal data store that everyone can have, which is independent of the apps using it. This is laudable, and is a solution to the centralised data stores "owned" by the likes of Facebook.

However, it still leaves room for app providers to concentrate their power. The #Fediverse model provides a means for communities to work together and potentially (in the future) for everyone to have their own instances.

@fitheach @kensanata I guess that could potentially be useful... But it's not something that excites me on it's own.

That problem space definitely sounds like something that should be running locally though. I mean we all have harddrives (don't we?) so why should SECs get involved? I hope that's where the project ends up.

@alcinnz @kensanata
Solid could be a *future* spec adopted by e.g. W3C, which would be good but the apps purposefully left un-defined. AP, however, is getting there *already* and includes interoperability between apps.

TBL would be better throwing his weight behind AP.

I can envisage a future #Fediverse where apps & data are stored in something that looks like a HomePlug, and everyone owns one. "Cheap as chips" doesn't just refer to potato products any more.

@alcinnz @fitheach I think one user story would be: what if you could write an app that used your Facebook data? And the fail case for Mastodon would still be an instance going down and taking all your data down with it; and local backups being useless for another instance (at least as far as your toots are concerned).

@kensanata @alcinnz
I agree that Mastodon data is a current problem, but I disagree that a common data standard is a solution to Facebook. FB works because of network effect and "enticing" services, Solid wouldn't change that.

Should Mastodon, or the Fediverse in general, rely on external data definitions? Solid might be a long time in the coming.

@fitheach @alcinnz I agree. It sounds like over-engineering to me. And I don’t see the business case for a successful platform. Then again, perhaps one day we will be able to buy Solid storage from Amazon, or users will be able to contribute some of their own Solid storage instead of money to their preferred app. Hm.

@kensanata @alcinnz I jumped in on another convo about this already, so please forgive the potential duplication in my posts, but...

I checked it out a couple of years ago, and thought it was a cool idea that still had some major issues to work out (like its reliance on browser certificates for user identity, and lack of good in-production implementations to use as references).
@kensanata @alcinnz IIRC, when I looked into it a few weeks ago, it still had the same issues, and I had actually assumed that the project had been abandoned, until I saw the same article.

@kensanata I'm open to this. I've been wanting GNUnet to take over for a year now, but I'll go to team Solid if they sort it out sooner.

@klaatu @kensanata GNUnet could use another user to make it a network :/