Notices by Strypey (strypey@mastodon.nzoss.nz), page 136

@clacke you mean like a genie in a bottle? Be careful what you wish for ... ;-)

@scolobb who hasn't had some kind of 'Scorcerer's Apprentice' experience where they accidentally spawned a whole mess of processes that spawn more processes every time you try to shut one down? ;-P
@clacke

@aral
> Crawling public content and third-party tracking are separate things.

Sure. The post you were replying to was more directed at @rick_777

@a_breakin_glass the difference is in what's required to implement the protocol, and what's kicked down the road into extensions. @Aerdan 's criticism is that XMPP kicks way too much essentially functionality into extensions, where AP saves them for things like what #ForgeFed are doing (protocol extensions specific to federating web-based code forges). Keep in mind that AP is a young spec, and some of the older apps are still federating with OStatus. XMPP has had years to sort shit out
@jcbrand

@Wolf480pl yes. Successful consensus process finds a synthesis of the best elements of each idea and concern brought to the table.

@jcbrand it's more of a 'prisoner's dilemma' than a 'tragedy of the commons', but otherwise I agree with what you say here. Jabber was set up to clone an already existing set of proprietary IM silos. Even if they had set out to achieve feature parity with Skype, which started about the same time, that still would have been a clone. If they'd baked in features no chat app had at the time, and got out ahead of the pack, the story might have been quite different.
@z428

@jcbrand so it appears to me that there is a missing element in the ecosystem; customers. People who are willing to pay a regular subscription/ membership so that secure jobs can be created for people to maintain ethical services. If enough #PlatformCooperatives were offering great chat UX on top of XMPP, that would help funnel funds and resources into improving all the stuff we've been bemoaning. This is the kind of stuff I want to facilitate through #OpenAppEcosystem group
@z428

In the short-term, I wonder if the answer might be fussy federation? Where each client gives a list of servers and clients with which it's willing to test and guarantee smooth inter-operation, and what features it does and doesn't support. So Wire, for example, might only federate with XMPP servers/ clients that support its #E2EE protocol, and reliable voice/video calling.
@shura @jcbrand @z428

@shura that might be possible, if there were a single #XMPP client that supports all the major OS, and the web, like #Wire does.
@jcbrand @z428

@jcbrand so here's my question for you, and anyone else who has been involved in developing web clients for XMPP, is there a better way to integrate XMPP support into existing fediverse apps? ie Natively, without giving people trying to self-host the app an extra package and integration to wrestle with. Could @jaywink add XMPP support to his Python federation libary for use in #Socialhome, for example?
@ayy @Goffi @js

@jcbrand just out of curiosity, have you looked into the way #Diaspora has used XMPP? My understanding is that they built a plug-in that allows users to interact with a separate XMPP server via the Diaspora web interface. This means that podmins who want XMPP chat to be available have to install both Diaspora itself *and* an XMPP server, and strap it all together. It also means the #UX of the integration is ... well ... clunky and frustrating (last time I tried it anyway).
@ayy

@jcbrand I've been thinking for a while that there needs to be a standards-agnostic body that facilitates user-friendly inter-operation between #FreeCode apps and services. A sort of 'Free Federation Foundation', but with a better name, because that's a terrible tongue-twister ;) That's why I was excited when the Collaborative Technology Alliance got started, and I'd still like to see it take off:
https://collaborative.tech/ #CTA
@z428

@z428 absolutely. I've been bashing on about this for years, even since most of the global #Indymedia network collapsed under its own weight. #DougRushkoff explains the problem with relying on VC-funded startups as our default organizing model really well in his 'Throwing Rocks ...' book. This is why I'm very excited by the #PlatformCooperative movement.
@jcbrand

@Aerdan I don't think so. Integrating GTalk with GMail was a great design decision, and quite intentional. XMPP had a significant and growing user base at the time GTalk adopted it, which meant they weren't starting out significantly behind the proprietary silos dominating chat at the time (ICQ, AOL, Yahoo, MSN etc). Yes, arguably the actual tech of XMPP hasn't aged well, but what other federated chat standards were there at the time? None that I remember.
@jcbrand @bob

@clacke to me, it reads as either a) well-meaning but ridiculously naive, or b) a bunch of lame excuses for why folks should put money into the Openbook crowdfunding campaign instead of supporting one of the *many* projects that already do most of what they are proposing.

@clacke here Joel from Openbook answering that:
https://www.reddit.com/r/opensource/comments/8w2k4y/can_openbook_an_open_source_securityfocussed/e1skt53/

@61 @jcbrand the only way to have a chance of competing with the network effect of Windows or MacOS, as platforms, is with a cooperative platform that serves all the same purposes for users, of which GNU/Linux distros are only one hot-swap-able component. The #EFoundation ("#Eelo") and #KPIM in Indonesia are among the #SoftwareFreedom orgs that seem to understand the need to pitch an OS-as-a-Platform to end users.

@61 fair cop. But what people mean when they ask that, more specifically, is "when is the superset of OS based on *GNU/Linux* going to be adopted by more than a hobbyist minority as a day-to-day OS". Treating each distro as a "product" would only make sense if Windows and MacOS were products, subject to the same market competition as pencils or bread, rather than "platforms" with powerful network effects.
@jcbrand

@pertho @jcbrand have you tried hosting your own XMPP server? Can you link me to a good #HowTo on setting one up? AFAIK this is not a trivial task.

@LWFlouisa #DavidGraeber writes some great stuff (can't remember which book right now) about how consensus decision-making is a process of finding agreement among a group of people with fundamentally different internal worlds. The #Borg are not capable of consensus, because they share a single meta-mind, where each individual Borg is just a neuron.