Rushyo ๐Ÿ‡ช๐Ÿ‡บ is a user on mastodon.social. You can follow them or interact with them if you have an account anywhere in the fediverse.
Rushyo ๐Ÿ‡ช๐Ÿ‡บ @Rushyo

Just had a chat with another "technologist" (haha) with decades more experience than me about the ins and outs of Mastodon's issues. From technical to social to federated to individual to admin to dev to moderator. Many hours of discussion of all the elements.

We agreed that some form of federal moderation committee is likely a must, as is a standards committee (for dev interoperability).

It's just unsustainable that everyone is essentially communicating primary by shouting over fences.

ยท Web ยท 21 ยท 19

@Rushyo Doesn't mastodon then lose its purpose if there is one group making decisions for the federation?

@BinaryBear It depends whether that group is representative of the federation. Right now, decisions aren't being made on the basis of what's best for the federation either, it's a bunch of segmented (largely) selfish decision-making. De-centralised, not federated.

@Rushyo Even if there was a group tomorrow that was made up of people who truly wanted only the best for the federation and keeping it an open, decentralized platform, isnt it possible that this group could be taken over by people who would abuse it? Wouldnt the better option to be to provide the user tools to moderate their own mastadon experience, or are you talking just moderating the technical side of mastadon.

@BinaryBear There's a huge number of issues at all levels, right now. But, having community representatives actually communicate would serve federation as opposed to merely de-centralised silos. Right now we have a bunch of instances/platforms who only communicate to, essentially, hate on each other. Why not communicate on a regular basis? About other problems? Like adults?

@Rushyo im sorry im just making sure that im understanding the example all the way. Are you talking about a cross server moderation team eing able to do something about server A being most filled with content that is hating on/rude to server B?

@BinaryBear No. Literally just having people in a room, communicating regularly. Nothing more than that. No formal political structure other than "Sit".

@Rushyo Ah, Gotcha, I was heading off in a different direction with that.

@BinaryBear @Rushyo No group that's meaningfully usefull is immune to abuse. The two traits go hand in hand.

You try to keep it out of the hands of assholes.

@Rushyo jump the fence and say that to my face ;)

@Rushyo lol there is a standards committee and standards, why aren't they actually implemented?

@superbranch Which single spec for this federation do you feel exists?

@superbranch Why would you want a single spec? Because the pace of change and diversity of user expectations is growing faster than this platform can cope. Mastodon, GS, etc are faster becoming non-interoperable, and the demands and expectations for 'custom features' are rising higher, leading to technical divergence. It's driving wedges in the federated model.

@Rushyo But the spec for communication between instances is largely irrelevant for these cases. People are asking for choices they can make as users of clients and as admins of instances. This is largely protocol agnostic.

@superbranch This is simply un-true if you look at the Mastodon issue list, and the reasons issues are being rejected. I know my client, like most software, can't function in a word of un-defined specifications where anything could break moment to moment. And it has technical a chilling effect.

@Rushyo imho software has users and maintainers should be asking "what should my software work with?" and not "how do I make this automatically work with everything anyone else can throw at it?"

@superbranch This is federated software. If you want non-federated software, that approach is fine, but the entire point of this platform is federation, interoperability, with other software.

@Rushyo I think you can define a sufficiently open standard that encompasses all the standards that are currently out there, but you can't force people to accept it. Federation is about making opportunities to connect, but you still have to put the work in. It's a lot like i18n support.

@superbranch The point is if people are represented in these discussions they won't be 'forced' to accept it, they might actually *choose* to. And then the platforms actually have a base from which to grow, instead of right now where even the basics (should we federate HTML or text?) are subject to dis-agreement, which totally chills the ability to grow and stay connected.

@Rushyo I dunno, it's 2017, we should be able to deal with 2 possible, extremely common text content formats without putting a lot of effort in

@superbranch That's true if it's a technical problem. And yet, this is a social and not a technical issue, so we can't. It's causing all kinds of issues right now, in 2017, in a federation that has existed for a very long time.

@Rushyo I think our difference of opinion may be expressed as: society needs to be more uniform to fit the current tech v. the tech needs to diversify to meet the potential of society

@superbranch The tech is failing to keep up with the change of pace in this society. We need to resolve that. If this is a 'federation', that means federal-level working. If we don't collaborate at a federal level it will cease to be a federation on a technical level, just an archipelago of platform silos, once social progress overtakes it.

@Rushyo I maintain that the standards we have are fine for communication, and what people need is more personal authority in how their clients may be configured

It's unclear in this conversation who knows and who doesn't know what, so just to be clear I'm mentioning that https://www.w3.org/community/ostatus/wiki/images/9/93/OStatus_1.0_Draft_2.pdf exists but was never finalized.

The de facto standard is that plus whatever StatusNet had already done by then, and what GS has done since. It would be nice if this were all documented unambiguously and correctly, so that people could evolve the protocol as necessary. @maiyannah is working on this.

However, the W3C has left it behind and are instead working on, and has presented as a Candidate Recommendation, its replacement #ActivityPub.

@clacke @maiyannah Yup. Well, that's what the GS folks say is the de-facto standard anyway ๐Ÿ˜‰ ๐Ÿ˜Š

@rushyo @maiyannah The protocol has since 2012 slowly been evolved through consensus between various admins and developers. To abandon that would be an enwp.org/Embrace%2C_extend_and_extinguish move.

@clacke @maiyannah I do not feel the dynamics that have worked for the previous years are likely to survive the growth of the network. When the stakeholders involved change dramatically, and the pace of change shifts by orders of magnitude something, as the Beastie Boys would say, has got to give.

@clacke @maiyannah Trying to maintain a slow pace of change in a technical or social environment will not work, it's a question of what replaces it

I do not imagine that there is only one other possible outcome (GNU vs Microsoft is a helluva false dichotomy here). That would imagine Mastodon + other parties were actively belligerent, w/ had corporate goals to achieve. They don't. They can reasoned with & talked to. They're just other devs like you

@rushyo @maiyannah Belligerence and recklessness are difficult to distinguish, and in the end any distinction doesn't matter. The result is a split federation.

@clacke @maiyannah Then give up I guess? *shrugs* if you imagine that other, non-GNU people joining is the end of the road for the principle of openly federating, then you've already "lost" the "battle" by your own reasoning. I think you utterly under-estimate the people involved on the other side of the fence if you can't equate them to anything but a malicious corporation not worth speaking to.

@rushyo @maiyannah People have been trying to talk to people. But they'd rather innovate than federate.

@twitter @clacke @maiyannah who is even 'they' at this point? There's a lot of "us" and "them" in this convo.

@rushyo @maiyannah @twitter You're the one who said "Well, that's what the GS folks say is the de-facto standard anyway ;-) :-)" about a protocol that had a handful implementations and a consensus about what was the de facto standard a year ago. I'm saying that people who talk like that are not interested in cooperation. You say you want a standard, but at the same time are saying that half the network can't be part of the new consensus.
@maiyannah @rushyo @twitter If Mastodon wants to move fast and break things, that means there can't be one standard.

@clacke @twitter @maiyannah I've spoken with the Mastodon dev and their idea of what the spec they are implementing is clearly different to what the GS folks think is the current standard. That's a fact, not me stating an attack on you.

"I'm saying that people who talk like that are not interested in cooperation"

Then, frankly, *you're* being the belligerent fool who can't work with other people who share common goals. Not them

@maiyannah
We're not refusing, it's on the roadmap, which is being published soonโ„ข, with a high priority. (conversation_id)
@twitter @Rushyo @clacke

@maiyannah No we literally went through it right now in the past 2 hours and added it to the "We're working on this now".
@twitter @Rushyo @clacke

@maiyannah But it's happening. Which from a technological standpoint is all that actually matters, if we're talking about the greater aspect of Mastodon / GNU Social.
@twitter @Rushyo @clacke

@maiyannah
I'm telling you what I know from what I see, from what I am working with.

I'm communicating with you about these things because you and Gargron can't communicate with each other.

I communicate with @mmn about GNU Social stuff, and to figure out implementations which are beneficial on both sides.

Sometimes, we need to deescalate. Which sometimes other people have to step in and help with that.

@maiyannah This is yet another discussion. :)
You're obviously harbouring a lot of bad feelings against Gargron, and I feel like it may cloud the way you discuss these matters a lot of the time. It's unfortunate, but I try to look past it, because I know that what you want to say is actually feature related.

@maiyannah It was not my intention to patronize you, but it is from my experience when talking with you about these things.
Which is why I try to steer around directly talking about certain things, which obviously doesn't always work.
I really just want to be able to have meaningful conversations with you, and for us to be able to build towards common goals.

@maloki I read the above discussion and just want to let you know that I'm always ready to know if something on our side should be improved or fixed. I've noticed already that the underscore-in-nick thing causes some confusion and I'm surely going to see if I can take care of the confusing UX around it on !GNUsocial's side.

@mmn yeah I pinged you something on the other accouht @mmn

@maloki quit.im is my image sharing account, it is probably inappropriate/impractical to have any non-imagery discussion with that account :]

@mmn Yeah, but I didn't know and it was the last one I remembered interacting with you from :D

@maloki Right, that thing could probably be fixed by doing a remote-fetch of parent (which I believe #Mastodon does for replies to unknown posts, right?).

I haven't checked whether the feed is fixed yet (@Gargron mentions the Suprrrfeedrrrrr "sanitation" with thr:in-reply-to)

We stitch together instances in !GNUsocial based on the conversation URI though, which is never stored or distributed from #Mastodon (which relies entirely on thr:in-reply-to)

What we have is however a bit of a mess and I don't remember which part is what we actually use and which part just comes along from #StatusNet (where none of this worked and conversations were even more broken).
@maloki The part of the federated Atom feed that tells !GNUsocial which conversation (if the in-reply-to is unknown) it should be stitched into looks like this:

The part of the

<link rel="ostatus:conversation" href="https://social.umeahackerspace.se/conversation/690842"/>
<ostatus:conversation href="https://social.umeahackerspace.se/conversation/690842" local_id="690842" ref="https://community.highlandarrow.com/conversation/429650">https://community.highlandarrow.com/conversation/429650</ostatus:conversation>


Which is an incomprehensible mess that I fully understand if @Gargron hasn't bothered trying to understand. (though I have a vague memory of him asking me about it, or someone else, and I couldn't give an answer. Maybe if I have enough sleep tonight I'll give a go at clearing the mess up tomorrow).
@rushyo @maiyannah @twitter "what the GS folks think is the current standard" is the current standard. Gargron is making his own judgements about what the OStatus spec means. A spec that was written to codify what StatusNet did.

That's "Nuh-uh, we're building the bridge here. That's where the map says the canyon is." There is no sense in a radical interpretation of the spec, it's a best-effort spec.

@clacke @twitter @maiyannah "Gargron is making his own judgements about what the OStatus spec means. A spec that was written to codify what StatusNet did."

It can't both be a common de jure spec AND a de facto non-spec as it suits you. Pick one or the other, stop moving the goalposts. Your position is "if it's not GNUSocial" it's wrong. Well what kind of ecosystem do you imagine that is?!

@rushyo @maiyannah @twitter The goalpost has always been "it's implementation-defined and that's unfortunate, somebody should codify this properly". Starting with doing things differently and deriding people who are describing the terrain is not moving toward defragmentation and standardization.

@clacke @twitter @maiyannah That's all I was suggesting. I said I didn't think OStatus was doing it properly, so it should be re-codified with something that works (or, hell, just rally around ActivityPub?) and the people who are then going to be involved in implementing it should all be stakeholders involved in the discussions of how it is performed.

My position can be summed up as simply: Not talking doesn't seem to be working

@Rushyo @clacke @maiyannah so it's alright if my new mastodon implementation spews garbage into existing instances? it's fully interoperable with mastodon! i'm the head dev and i'm just implementing what i think is the current standard.

@twitter @clacke No. I don't understand who or what you've been arguing against in the middle of all this. You've just added some random inflammatory statements in the middle of an otherwise rational discussion. I understand (at least partially) where clacke is coming from, and I'm trying to make myself clearer too, but you are just throwing perpendicular argument grenades in the middle. It's unhelpful AF.

@clacke @twitter That's why I've been ignoring you, and asked you to stop. I figure we can have a separate debate another time, where maybe the two discussions won't be confused.

@clacke @twitter @maiyannah "You say you want a standard, but at the same time are saying that half the network can't be part of the new consensus."

Drivel. I have no interest in dictating who can or cannot. I just think a standard would stop it from fragmenting, as a basic principle. Simple as that.

I don't even agree with Mastodon's approach, y'know. Your "non-GNU, us VS them" sentiment paints a damn broad brush.

@clacke @maiyannah It's a scaling problem (too many tech changes too fast), not an 'attack' designed to extinguish the network that needs defending against.

If growth of the network were to mean that inevitably the protocols either break compat. or become subject to such an attack, then we can logically derive that federation is a doomed experiment once scaling hits.

I don't think that implied belligerence exists, and I don't think it's doomed.

Ey, Mastofam..
What's your favorite song?
Maybe not 'fav fav', but one you play literally every day

youtube.com/watch?v=pJk0p-98Xz

@Rushyo Thank you for your work. I can't imaging having to deal with a project that took off this crazily. I'm already reeling from the culture shock just as a user. You guys have been far more responsive than I think anyone could fairly expect. Good luck with the growing pains.

@Rushyo It sounds like you're trying to turn Mastodon into twitter. Centralized committees are pretty much the opposite of a decentralized system.

@junglestrike If that's what you've taken from my posts, then may I suggest you've not read them very well.

@Rushyo You are advocating the equivalent of the twitter trust and safety council for a decentralized social network, no? And there are already OStatus protocol standards.

@junglestrike No, nothing even close to that. Basically not even on the same dimension.

And yes, but I already covered that discussion point in depth if you read my posts.

@Rushyo How is a 'centralized moderation committee' different from Twitter's central moderation?

@junglestrike In every aspect but name, which I explained in great depth in my previous posts. Either read the posts or don't bother, but don't ask me to repeat everything I've written because you don't want to read it in a non-personalised form.

@junglestrike @Rushyo problem is that this is different. and existing in a different era. history may have no judgement here

@acw @Rushyo mastodon is not the first decentralized social network, and gnusocial and others somehow managed to survive without any statists forming councils

@junglestrike Gee. If only I'd already written multiple posts explaining exactly why those things are distinct, in great detail.

You clearly have no intention of doing anything but trying to pick an argument for its own sake, and have no intention of reading the discussion that has already been had because then you're worried you wouldn't be able to pick a fight over it.

Bye now.

@junglestrike @acw @Rushyo I wonder if Mastodon's future is predicted by Usenet.

@acw @Rushyo @Ash quite possibly, mastodon's killer app feature is really just a non shit UI compared to gnusocial

@acw @Rushyo @Ash and I just get the feeling the masto guy is more interested in platform building than interop with Ostatus

@maiyannah @acw @junglestrike @rushyo @ash It federating properly is several massive steps behind a lot of much more important things, money, growing his following, much more advertising, setting a good example by going after women who talk back to him / don't agree with him.

@Ash @Rushyo @acw @purplehippo if only he spent as much effort on fixing federation as is being spent on stickers

@junglestrike @Rushyo @Ash its open source. who decides that it goes to platform building and Ostatus

@junglestrike @acw @Rushyo nothing a council hates and will punish more than being unnecessary.

@Rushyo love the imagery, generally agree. if/when you get wind of stuff loop me in?

@bea Will do. I'm not sure if anything will come of it; I'm just beating a drum about the topic to stir discussion. Partly because I'm not positioned to do so (I'm neither a dev nor admin as I'm too busy with e2e) and partly because I suspect there are people with much better experience/knowledge of how everything jives at the moment, and I'd rather their opinions ultimately came to define any solution over mine.

@bea The fact I'm getting a number of scared people throwing hysteria at straw-man arguments, who weren't even involved in the discussion, makes me think it's having at least some of the intended effect ๐Ÿ™‚

@Rushyo yep that's about what i figured! it's appreciated. i haven't modified anything but my own instances so far so i hesitate to call myself a dev but i am VERY interested in this stuff

@Rushyo @bea Been meaning to ask: where is the being added in the stack? Is this at the Mastodon protocol level, above OStatus?

@cs @bea The e2e message format in principle sits atop any social messaging medium (toot, tweet, etc).

The key sharing happens off-channel, deriving trust from de-centralised sources and pinning identities.

The reference prototype focuses on messaging over toots via the Mastodon API to other Mastodon API-using clients.

@Rushyo @bea So this is encrypting (and possibly re-encoding?) text and then sending that as the toot/status content? With this method, doesn't other toot metadata (e.g. timestamps, ids, etc.) remain unencrypted? Why not do it in the Mastodon API/protocol layer?

@cs @bea Whole bunch of reasons. Will explain in more detail in an FAQ, but it largely comes down to two things: 1) It doesn't align with Mastodon goals, so where am I going to integrate it? So far the only issue I've raised (optional plaintext messages) to increase stability has been rejected. Utterly impractical to integrate an entire protocol. and 2) The primary threat agent is a malicious admin, who will necessarily be able to ascertain those details anyway.

@Rushyo @bea Both fair reasons. I was thinking #2 could maybe be solved by making the change at the protocol-level, but I suppose such changes would be so significant, it couldn't even be server-client anymore.

@cs @bea Yeah, the TootCrypt concept works around the assumption of a malleable social network completely outside my control, which frees me to develop something without needing standardisation or agreement. Given what I know now of the state of the OStatus ecosystem, I'm very happy with that call.

A federation-wide moderation committee? Global moderation does not belong in the federation. It belongs back on a centralized network like Twitter.

@lnxw48a1 No, a *federated* moderation committee. As in, a committee of moderators from different instances, who chat to each other from positions of de-centralised authority.

That's completely different from a "federation-wide moderation committee", which would be a central moderation authority.

@Rushyo @lnxw48a1 I'm not sure I see the difference either, tbh.

The instant they start to agree on things, the external effect is the same.

@Elucidating @lnxw48a1 they don't have to agree. nobody can make people agree in such a setup. either they come to compromises that allow instances to work together, or they do what they'd have done anyway. either there's a compromise the parties can work with, or they go back to square one. nobody's hand is forced.

@Rushyo @lnxw48a1 What if they collude though. And wouldn't there be significant reward if they did collude?

@Elucidating @lnxw48a1 They can collude now. Except what they determine can now be done in secret, and nobody has a common forum to bring it to light, so can only scream about it.

@Rushyo @lnxw48a1 If only we had access to a method of disseminating information in a broadcast mode to people interested in Mastodon.

@Elucidating @lnxw48a1 that favours people with social power and reach. it doesn't favour the actual stakeholders. people like me can get discussions like this started because this is what I do. in the meantime, actual instance admins are asking me to be their voice.

That. *Sucks*.

@starhaze @Rushyo @lnxw48a1 Adding more humans will not necessarily make things more clear, which I feel like is a blanket assumption backing "let's form a committee!"

I don't disagree in principle. I just wonder how to put it in practice in a way that's more valuable than just having the admins talk openly over mastodon itself.

@Elucidating @lnxw48a1 @starhaze Actually, I'd argue it's reducing the number of humans to representatives of instances, rather than having the entire community throwing arguments back and forth in a massive squabble.

@Rushyo @starhaze @lnxw48a1 One thing I have learned watching the and running other online discussion groups: that never goes away.

@lnxw48a1 @Elucidating @Rushyo we don't have to eliminate human bullshit, just reduce it

@lnxw48a1 @rushyo there could be a council of soviet deputies, but I doubt that it would be possible to get all of the factions to agree on any universal moderation policy
@bob @clacke @stitchxd @rushyo Such a council of mods could speak only for their own instances. Their decisions MUST NOT and could not be binding on any other instances.
@lnxw48a1 @bob @stitchxd @clacke @rushyo Didn't FIDONET go through this?ย  It didn't work then and went into schism.

I think there's definitely some lessons to take from that, but they're not very well documented. It's likely we'll just have to re-learn most of them with the aid of the communication tools and knowledge of internet communities we have that people didn't back then. If anyone does know a book I can read, I'd definitely like to hear about it.

(BTW. Anyone wants off this thread train just ask, please!)

@bob @santa @clacke @lnxw48a1 See, that committee is adorable. I don't see the problem.

SHIP IT.

@lnxw48a1 @Rushyo @stitchxd @clacke @bob The fantastic part about a federated network, is that it's impossible to lord over all of them. You can create agreements with willing participants, but not force others to comply with your choices.

@bob @clacke @stitchxd @Rushyo @lnxw48a1 If the fediverse ever hits that billion-user level type popularity, there will likely be all sorts of branches and forks and communities that do and do not get along. That isn't inherently a bad thing, people will decide who they want to federate with.

@ocdtrekkie @bob @clacke @santa @lnxw48a1 Agreed, to an extent. I think there's a lot of solvable problems we have right now that are only not getting solved because the drivers are really new and different and scary, not because there's inherent differences that make it impossible. To break up the network for that goal seems counter-productive (counter-social), but is always an option.

@ocdtrekkie @lnxw48a1 @clacke @stitchxd @rushyo and the network effect should hold the majority of it together. Most users will want to communicate without hindrances between servers, and will complain if things diverge too much.

@bob @Rushyo @stitchxd @clacke @lnxw48a1 The network effect also grants the largest software developers and instance admins the most power to define the standard. But I doubt massive growth will happen without at least a few major forks and factions in the community.

seems like this is a recipe for burnout.
@lnxw48a1 first order of business: wth did rainbow dash network do wrong?!