i am trying to swap "god i hate software" from my vocab for "someday, together, we will make good software"

there is no leverage in competing with scabware while we cannot distinguish ourselves from scabs: no victory will endure that cannot support the labor that makes it possible.

to honor the craft, to do our works justice, we must separate them from the industry. we must find ways to make good things without exploiting each other for our free time.

@ekaitz_zarraga alright here we go: https://github.com/garbados/p2p-social-bookmarking

@ekaitz_zarraga cool. i've got a demo. lemme see about uploading what i've got so far.

the floor is lavascript

my favorite things about mastodon when i first arrived (uhhh more than a year ago):

- i could say ACAB without getting dogpiled by fascist sockpuppets
- i could nerd about p2p things and folks chimed in with thoughtful takes based on considerable experience, and even material assistance!
- when we couldn't afford food, folks chipped in and we made it through the month ?
- with the exception of the rare sockpuppet, even the people who disagree with you are thoughtful human beings.

arguing about standards like

$ ls
README DONOTREADME
$ ?

Over the weekend I made a PouchDB plugin that transparently encrypts and decrypts your data so that only encrypted data is sent during replication, while encrypted data that you receive is automatically decrypted. This way, a server can't (easily) know anything about your data, but you can still query your data as usual.

It's called ComDB: https://github.com/garbados/comdb

@aral

> the actual long-form DAT address that it maps to is a cryptographic public key that is unique to the content of this site

the url is not unique to content, only to the public key. that's what allows you to update the content of the archive: so long as you have the private key, you can write to the archive at that public key. IPFS urls, on the other hand, do uniquely map to content (which is why they change when the content changes)

docs: https://docs.datproject.org/concepts#dat-links

@ekaitz_zarraga yeah, that should be fine. the only issue with VDUs is scaling, since it adds some overhead to writes (since the update function is custom code that couchdb has to interpret every time) but if you're monitoring the server then you can catch that kind of usage growth before it becomes a problem, y'know?