Notices by Canageek (canageek@cybre.space), page 24

@Canageek if it is written down, there is a record of it and you should assume an unintended party will read it.
With mastodon, we pass those messages to lots of servers, so there's LOTS of copies.

@Canageek the even cooler thing about this concept living in 1995 is that it's meant for people to be offline too.

one could run scuttlebutt over a sneakernet or occasional dial-up.

I could go on a sea expedition for 6 months and document some scientific findings and all my work would be time and series verified. Then I could sync up whenever I hit a port with a connection.

@Canageek the concepts aren't advanced beyond 1995.

everyone has a local copy of messages that they sign with a private key (been around since the 70s), then you send your data to other people who can verify it.

they've even written a totally HTML only (no frontend JS) SSB / patchwork client which would run fine in netscape navigator :P (https://github.com/ssbc/patchfoo)

@Canageek

Yeah, I tend to think along the lines of the files system being the database and each file is a small data structure consisting of filename, timestamp,(title,content,links) where the stuff in () is the text and links might be filenames to parent files and comment files.

The other idea I had once, instead of a federated model, was no servers.

Bittorrent based social network. Messages, pictures, whatever, all go into the swarm. If I want to follow you, my client asks around until it find a copy. The more people follow someone the easier it is to get copies when they are offline.

This would also have a sort of weird non-chronological arrival order. Messages that more people see arrive first, as do messages from people online.

I'm wondering if you could make a really simple social network without using a database. (Well, you'd probably need one for login information, but other then that...)

Each users messages are saved in a text file. The people they follow are also saved in a text file. When they want to see messages they request the messages from each user one at a time, and reassemble them on the other end.

The client requests the messages a small number at a time until it reaches messages it has already seen.

"Only hashtags are searchable" is a lie though. Only hashtags are searchable BY SEARCH ENGINES THAT FOLLOW THE RULES.

We *JUST* had a client that was archiving every toot and letting people full-text search them. It isn't hard. People are going to toot things thinking it is safe, and it will come back to bite them.

Mozilla giving web users what they REALLY want: "The new Firefox lets you stop websites from asking to send you notifications"

https://arstechnica.com/gadgets/2018/03/the-new-firefox-lets-you-stop-websites-from-asking-to-send-you-notifications/

Surprisingly enough, the original HTML standard was responsive. Your browser would authomatically trim your texts in your browser viewport. No need for CSS/JS and bloatstrap.

But of course people only wants eye candy stuff.

To be honest, the internet would be much more efficient if it was text mode only. But that doesnt make "moneys" for the "corporates".