Sven Slootweg (soft-deprecated) @joepie91@pixie.town

@riley@toot.site I imagine that that might be why :p

What it used to be:
- many early adopters were anarchists/communists/etc.
- a huge public commons of highly reusable, high-quality, collaboratively developed libraries on npm, built for the benefit of the public, not for profit
- high degree of interoperability between different people's work, no need for pointless busywork to redo the same work over and over again
- successful(!) 'community specs' designed through community consensus (Promises/A+, CommonJS, etc.), gaining near-universal adoption
- fundamentally different structures from other language ecosystems, both technical and social, to make this work

What went wrong:
- large influx of users from other ecosystems due to hype in startup circles, unfamiliar with the established practices and reasons why
- early adopters failed to effectively convey and explain the ideological basis
- corporate adoption and subsequent capture; increasing "business value", leading to corporate steering of many essential pieces of the ecosystem (language spec, Node.js, etc.)
- npm became npm inc., a for-profit corporation, eventually being acquired by Github due to its large userbase, placing control over the public commons and its namespace in private hands
- ideological basis was forgotten, early adopters eventually left for greener pastures, now an almost purely parasitic environment of people leeching off the commons without guarding its integrity or health
- community-consensus specs started being replaced by "official", by-decree-from-up-high language specs (CommonJS -> ESM, Promises/A+ -> ES Promises)
- widespread adoption of these "official" specs, even though they were in many ways worse, due to their "official" label and many people assuming that what a central authority says must be correct or better
- rapid increase in shiny, well-marketed new tooling that is not interoperable with the existing ecosystem at all, and frequently works less well
- more and more commercial/proprietary 'sidecar' services (eg. Snyk) that you are expected to use, sometimes replacing open initiatives
- now an ecosystem and public commons that is rotting in every aspect with no real hope for recovery

... we should probably learn from this?

@bram_dingelstad Probably worth trying that first, would probably be simpler to implement :) Though I wonder whether it'd survive any transcoding process

@bram_dingelstad I don't recall any of the specific algorithms unfortunately, but a quick search turned up this one which seems promising at a glance: https://www.iasj.net/iasj/download/cac16130d9ed9c32

@bram_dingelstad AFAIK there are watermarking systems which survive encoding and compression! Precisely to prevent stripping out the watermark. And it'd be kind of poetic to straight-up use a sharing-prevention mechanism for sharing stuff :)

@bram_dingelstad How about encoding it into the video data itself, like how 'invisible watermarks' work?

@claasgefroi@mastodon.social "[...] und es fehlen noch die prominenten Namen, denen man folgen kann, zum Beispiel Elon Musk."

Tell me you don't understand the fediverse without telling me...

(I promise this is not a shitpost)

another poll

@devurandom@cybre.space Sounds good :) Probably a good idea to list Delta Chat after Matrix though, as it seems to not have any forward secrecy (just like PGP which it's apparently built upon, via Autocrypt)

@devurandom@cybre.space Also, +1 for pointing out that DMs on Twitter are also unencrypted :)

@devurandom@cybre.space A couple of suggestions:
- "Jabber" -> XMPP, which is both the official and the common name nowadays, and distinguishes from the jabber.org service
- "Element" -> "Matrix using eg. Element", to distinguish between the protocol and a specific client
- Removing "Telegram" or at least adding a warning - its homegrown E2EE crypto is highly questionable and generally not taken seriously by cryptographers