A few days ago I watched Dan Abramov's new talk, "Web Without Walls". It's a great talk, as always! But there are a couple of things I'm uneasy about.

Let me explain:

First of all, it's ironic that my comment on YouTube seems to have been swallowed up by the algorithm 😅. If you open the video, you won't find it unless you sort by "newest comments". I wonder why it doesn't show up in the default "top comments" view, not even at the bottom :/.

This is what I wrote:

25:20 If you're interested in a "galaxy brain" solution, check out the Solid Protocol. It is very similar to what Dan explains in this presentation, but it has universal formats, and it was created by the person who created the actual web (Tim Berners-Lee).

Besides the universal data formats, I also prefer Solid because of its focus on private data. The AT protocol seems more focused on public data and social interactions. If anything, I see it as an alternative to ActivityPub, not Solid.

But there are also some things I like from Bluesky. I really like what they did with encryption, and the way users control their identities seems a lot more straightforward.

Solid has been around for a while, long before Bluesky, but it hasn't managed to pick up any traction. I wonder how Bluesky managed to attract talented developers like Dan and others. It can't just be "the money"... right?

And in case you're wondering, yes, the Bluesky team did know about Solid. Before announcing the AT protocol, they did an "ecosystem review" in which they studied all the existing solutions they would proceed to ignore: gitlab.com/bluesky-community1/

Not only that, apparently Bluesky is not even living up to their own ideals (TLDR: Bluesky is not really decentralized... "yet"): beige.party/@possibledog/11336

Having said that, Dan's talk has made me reconsider my opinion on Bluesky. I trust him, and I don't believe he'd be lying to our faces. So I may dig into the details to learn more.

@noeldemartin i’ve done a lot of playing and research in this space, volunteering at dweb camp for three years now

this is still me from this thread

forum.solidproject.org/t/idea-

point blank comparing apples to oranges of bluesky and solid

to attract me, solid has offered react and typescript for SDKs for my convenience

bluesky offers me http

solid is suffering for reasons i won’t speculate

the docs for using bluesky and mapping routes to resources is more webby to me than the state of solid.

@tychi I know that Solid is lacking in many aspects, I can understand why someone would find Bluesky more appealing.

The point I'm trying to make is that at some point, Bluesky was as experimental (or more) than Solid. But they managed to get a lot of interesting people on board to improve it. Whilst Solid seems to be stuck in time :(. I still think Solid has the potential to improve a lot, but it's been like that for years, "potential".

In any case, this was more of a rhetorical question xD.

@noeldemartin i know it was rhetorical, i want solid to be better too—

i wouldn’t say i find bluesky more appealing per se— i think they’re two different, complementary technologies

the silly thing is, the main problem i think bluesky solves is RSS

i love solid, but i think even before looking at any tech, the main problem solid solves is secure file transfer.

the main competition i see at the solid layer of the stack is sshfs and sshfs is really good. rss is a more personable problem
domain

@tychi Rather than talking about it as Solid's competition on a technical level, it seems like Bluesky is the only decentralization project that is attracting the mainstream's attention, and is getting A players on board. Other projects seem very fringe.

But if you watch Dan's presentation without knowing much about the technical details, it rings very similar to what Solid is trying to achieve.

Hence my comment about it being just the money, or if there is something else there...

@noeldemartin that mainstream attention is partially what i’m getting at. i don’t mean RSS as a technical thing, but as a shorthand— i onboarded my dad to bluesky because the “user api” is 1:1 to twitter, so he didn’t need to learn how to use it.

you and i are both savvy technologists.

my research of watching people use technology— sometimes the answer is as simple as, maybe if solid had exactly the same interface as bluesky and twitter, my dad might understand it.

@noeldemartin to anecdotally use my dad as an example of the type of person that could benefit from solid—

he’s a professional writer. mostly in dealing with sports projections for baseball and football

his target customers buy information from him

he currently uses substack because they handle billing and distribution for him

he primarily needs a place to type text and attach files

he doesn’t need to use substack, but they came up with a more appealing set of tools than w3c for him.

@noeldemartin i guess the reason i’m trying to case study my dad for you—

in my opinion, you’ve got the best grasp on the user experience of building against solid and i admire your work.

this is a complicated space, but i think answers as to adoption are simple.

like, if solid were to want to adopt any given user demographic, i believe the answer is a 1:1 skin of the tool they’ve already been trained on and anything else is not appealing to them.

@tychi Thanks for your comments. Yes, I agree with mostly everything you're saying. But there is nothing inherent in Solid that prevents this from happening as well (making an app that is user friendly, etc.). I'm sure if people like Dan decided to spend some time making Solid Apps, they would turn out as good, if not better, than Bluesky.

So the question remains. What made them choose Bluesky rather than Solid or any other decentralization project...

@noeldemartin i need to watch the video, i’ll follow back up again once i do

tangentially related, this problem space we’re defining together, i’m starting to visualize as “the batteries”

on the launcher app thread, i’ll follow up with my latest research in using gun.eco as a plugin to back the data structure of the custom html tags i use to build my apps

i see a path to use that technique so solid can back those data structures, that might get us competitive batteries.

@noeldemartin long story short, i was using react and redux, written by dan, for a long time in that ecosystem

as react moved away from redux-y paradigms, i decided to make what i liked about that compatible with vanilla js, with less boilerplate, and upgrade-able from custom elements to full web components if needed.

that’s a bit of history on the rationale and personal journey of that thread of mine on the solid forum.

solid was just the first use case i ran at with it.

Sign in to participate in the conversation
Noel's Mastodon

This is an instance-of-one managed by Noel De Martin.