Ancestors

Toot

Written by Dave Alvarado on 2024-11-26 at 14:40

Followers here, I'm genuinely curious about your take on this.

https://bsky.app/profile/alvarado.social/post/3lbubb4uf3s23

=> More informations about this toot | More toots from dave@alvarado.social

Descendants

Written by Ian Smith on 2024-11-26 at 16:58

@dave This isn't tcp and ip, it is ethernet or tokenring. The protos are the output of orgs that aren't commensurable things.

There are lots of issues in both protocols that deserve to be fixed, but nothing indicates there are the conditions for an 'iron sharpens iron' competition, and the AT side of the equation has a clear bias towards a 'zero-sum worldview' competition brought in by who is paying for it, how it's governed, and the structure & composition of its corporation.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Dave Alvarado on 2024-11-26 at 17:04

@katachora right, sorry, my question is "when extending AT, is it better to compete with bsky or cooperate with bsky?"

Like, I'm not sure that capitalistic zero-sum competition is the right model, but also the current capitalistic zero-sum competition that bsky is doing with Twitter is leading to a much better outcome than masto's disdainful ignoring of Twitter.

=> More informations about this toot | More toots from dave@alvarado.social

Written by Ian Smith on 2024-11-26 at 19:48

@dave oh, yeah, duh. So yeah, obviously, 'compete with', for many of the same reasons that you can't get 'iron sharpens iron'.

The Web is a big place, so it isn't literally true, but there is a cognitive and interoperability constraint that puts these kinds of things (AP or AT; bskyAT or openAT; etc.) in my "there can be only one" bucket and if you aren't directly benefiting by helping a proprietary or closed or clique-controlled protocol or platform or program, you probably shouldn't.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 20:05

@dave I think that the important question is, "who is social media for?". Based on the way that question gets answered, then we can ask how to build social media. And THEN we can ask about which protocols to use to share and spread content.

Twitter and threads are, I think, for the same thing -- money & power.

Bsky is, I think, trying to be for both money & power AND people

Fediverse is, I think, is trying to be for people but is also secretly addicted to power.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Dave Alvarado on 2024-11-26 at 20:10

@katachora I agree with how you're seeing things there, and I'm undecided which direction is the best way to go. I personally think it should be for people, so AT and AP are both on the table as options and I'm not sure which is going the better direction. Right now they're both going in bad-ish directions, but I think both could be turned to a better direction whether their creators want it or not.

=> More informations about this toot | More toots from dave@alvarado.social

Written by Ian Smith on 2024-11-26 at 20:53

@dave I've noodles on this a little, and been influenced a lot by folks who are from groups who get more than their fair share of abuse online, and really think that the wire serialization and messaging semantics have a floor, but above that floor, are fungible. What is important is the identity model, some sort of reputation framework, and frameworks for factuality and credibility that can be woven together just to get to that floor.

It's basically everything social that social media isn't.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:09

@dave "for people" might look like the most obvious thing that any "refugees from Twitter" landing spot should have - and that none of them do, afaik, is to be able to take your twitter subscriptions in, and ask you if you'd like to publish your old twitter handle, then match-make as people show up on the platform by saying something like, "Ian is on social.foo, would you like to resume following?"

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:11

@dave kind of like how Signal does this when a phone number you have in your contacts list joins the platform.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:15

@dave

changing platforms or instances ought not mean abandoning the relationships you had on a different platform or instance and, more importantly, the implied trust those relationships embody.

but it also ought to be a moment where you get to decide, do I actually trust this relationship?

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:15

@dave

In the end, all security problems are just policy problems, and all policy problems are just identity problems, and all identity problems are just trust problems and all trust problems are fraud.

social must be anti-fraud.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:53

@dave mechanically, i think the way that the red hat ecosystem used to work is legit.... Something like:

Fully open rolling release reference implementation feeds into a commercial offering feeds into a fully open point release reference implementation.

That has to mean that the commercial entity takes the community as a dependency not the other way around, because if the community is forced to take the commercial entity as a dependency, the community is captured and ends up being consumers.

=> More informations about this toot | More toots from katachora@hachyderm.io

Written by Ian Smith on 2024-11-26 at 21:55

@dave open has historically worked long-term only when the decisions being made about purpose, design, mechanics, and how it impacts people are uncoupled from quarterly earnings and investment returns

=> More informations about this toot | More toots from katachora@hachyderm.io