I think he probably has blind spots as we all do but I wouldn’t say he’s naive. More that he was reluctant to buy into p2p possibly because of the economics - that was one objection he raised. I explained how Safe Network planned to solve this but he was skeptical, though continued to be supportive of my work to demonstrate the Solid protocol running on Safe.
What we need is to be able to show people like Tim that it works!
Thats on this community as well.
We need more folk participating in test nets, poking the corners, building simple RUST-based apps in advance of the API.
This would be lovely and we know there are folks who’d like to have a go and others who would help them.
Reality though is that this takes time a commitment that few can muster in practice. When we have more to show more people will come to take a look and in that larger crowd there will be more people able to do this.
This much is true - I still struggle to get much beyond Hello World and stuff I could dash off in 30 secs in bash.
But I console myself with the fact that I am a hardware engineer really and that there are lots of you out there that have more of an aptitude for actual code.
If you could ever find the time, I’d love to see a commentary on your new upcoming eagerly awaited rewrite of vdash as a real world example of a useful application in rust. I think this would be very useful for wannabe rustaceans.
Rewrite? It’s already Rust so curious what you mean.
Apologies, rewrite is the wrong word, I mean the tweaks needed to make it work again.
re Mastodon
The versioning they have now for edits is an interesting parallel…
Mastodon now lets you edit posts
You can edit any post you’ve made, and it will send a notification to people who have already interacted with it to prevent any hijinks or shenanigans. It also lets people see the previous versions.
A more optimistic take on the Mastodon scalability debate.
the scale issues that you mentioned are not exponential, but linear. A user only has the stamina to post a given number of messages and words per day; it doesn’t matter who they’re sent to. There are no new connections being formed with each message, it’s just another entry in a huge database, and the large majority of messages are text.
I’d like to see someone do the maths based on 100m, 1B, etc users.
How many instances would that need? Typically max users per instance is a few tens of thousands but there are many much smaller so it’s going to be a lot?!
How well can this work if one instance ends up following users on thousands (or more) other instances?
I’m not saying it can’t, but that the issue is complex and not just about entries in a database.
There’s more to the argument than that - it’s a long old post - and it assumes messages will stay mostly text, which probably doesn’t hold. But yes, it would be good to see someone do some calculations.
