@bcrypt @tim I have closed down registration on this instance to prevent this

@Gargron @tim @bcrypt
Hopefully symbolism like this will help pre-bake the culture into one that might avoid that trap.

I feel like this is a race to see who can run the largest instance so they can get advertising $$$ somehow :(

If we can get something like the Patreon model that @five talked about it might sidestep the incentive feedback loop further. I wonder if there are any easy to use decentralized server providers that can get fed money from shared cryptocoin wallet.

@ultimape @Gargron @bcrypt @five How about a multi-sig wallet, like on BTC.com? Easy to user, and still usable without the third party.

@ultimape @five @bcrypt @tim @Gargron I think a #platformcoop instance that's free to users with moderate and transparent ads could be a good thing. Particularly if it kicked revenue back to users based on content creation? So people could choose to pay a subscription to a node for ad-free experience, or transparently participate in a free to use space ad-lite that comps users any revenues beyond operating costs.

@mattcropp I personally don't think that would scale well. Hard to get advertising money without strong deals, and I think ti would only encourage users to game the system.

But I might be biased. I'm still not done writing my anti-advertising thing and it's 10k words (including quotes).

@ultimape that's 10 whole toots! @mattcropp

@Gargron @tim i feel that was the right choice. i would like to run an instance but don't want the burden of maintaining it forever; giving users a migration strategy is a deal breaker for me running one

@bcrypt @tim @Gargron for me I would like to see a way to port follower-lists to other instances, I don't know if that's feasible but it would help being able to move and prevent supremacy of any one node

@sarahjeong @Gargron @bcrypt Working on it.

@tim @bcrypt @Gargron very cool to hear!

@tim @sarahjeong @Gargron @bcrypt Wait, working on it how?

@kevin_redacted @sarahjeong @Gargron @bcrypt

Quit my day-job as an IT manager 2 years ago to work on the topic of P2P identity & reputation. To get an idea, see: https://github.com/identifi/identifi

@tim Oh wow, cool. I wish I could afford to quit my day job and work on interesting projects.

Are you actively attempting to integrate this into Mastodon atm, or just talking long-term?

@kevin_redacted Working on a proposal that includes self-issuing identities, managing multiple identities, accurate tracking of accompanied reputation, controlling one’s social network(s), sharing sensitive data and/or personally identifiable information (PII), trolls, sock puppets, and Sybil swarms.

Give me a few more weeks and I'll drop it here first. :)

@kevin_redacted Oh, and there's also gonna be multi-sig, distributed (read: mobile) reputation, some P2P networking, strong crypto (for bothing signing messages and encrypting data), local data storage, and user- and/or machine-controlled sharing of (sensitive) data.

Should address heterogeneity, non-centralisation, micro-intermediation, pubkey namespace, WoT, JSON msgs, APIs, CLI, revocations, Sybil protection, DoS protection, etc.

@tim Sounds like you've thought things out a BIT more in depth than the kind of quick hack sort of solutions I was contemplating, lol

@kevin_redacted Since ~July 2014. It has been crystallizing for a while. I already can't wait to see how I'll get sent back to the drawing board. I've already proposed it to a few folks, and haven't found any major flaws so far. Smartest kid in the room principle though.

@kevin_redacted Right now I'm just super-excited because so many other people start thinking about, and discussing, the subject. The amount of fresh point-of-views on it is amazing. :D

@tim I bet

@kevin_redacted Prepare the reading glasses. Buffer overflow incoming.

@tim @kevin_redacted That sounds really cool. Do you happen to have anything written down about the technical design? I would love to read more about it.

@cdata @kevin_redacted It's heavily inspired by Identifi, and am currently working on a proposal with many more details. I'll dropping it here first. :)

https://github.com/identifi/identifi

@cdata @kevin_redacted *drop. And here's an article by Identifi's creator, which may interest you: https://medium.com/@mmalmi/learning-to-trust-strangers-167b652a654f#.2eyijxngr

@sarahjeong @Gargron @tim @bcrypt

I agree. Some sort of way to keep the userbase mobile is going to need to become very important.

@sarahjeong https://xkcd.com/1425/

@sarahjeong @Gargron @tim @bcrypt @rich_harris mentioned a ticket about this over on the bird site: https://github.com/tootsuite/mastodon/issues/177

@knowtheory @rich_harris @bcrypt @Gargron @sarahjeong

Issues were next on my to-do list, to study. Thanks for the shortcut! :)

@knowtheory @sarahjeong @gargron @bcrypt @tim @rich_harris ah word, the "account redirect" seems like it would solve a lot of problems.

@schlink @rich_harris @tim @bcrypt @Gargron @sarahjeong Heh. There are a lot of distributed computing problems hiding in here. I haven't caught up w/ the issues well enough to have an opinion about how scared i should be yet :D

@knowtheory @sarahjeong @Gargron @bcrypt @rich_harris @schlink https://maly.io/media/c3rrCCl5gIRSL4sDNwE

@tim @schlink @rich_harris @bcrypt @Gargron @sarahjeong Wait, is this saying that the CAP theorem is newspeak? :D

@bcrypt Why is a migration strategy a deal breaker?

@cr1901 see the part of the sentence before the semicolon

@bcrypt I think I'm missing something obvious. Wouldn't giving users a migration strategy mean that you *don't* intend to run an instance forever? Or is a migration strategy a recurring burden as long as you run the instance.

@bcrypt @Gargron @tim This is such a good point, giving users a migration strategy seems key to this thing working… It also seems like the kind of design choice you'd have to make early on, but maybe there is a conceivable strategy already?

@tim @Gargron @bcrypt And +1 on genuinely interested in running an instance, but the burden of LTS is off-putting.

@Gargron @tim @bcrypt Maybe throttle, instead of close down? I suspect it's an important time, and total growth may be as important as federation-balance? I really want a Twitter replacement so I might just be too excited...

@Gargron @tim @bcrypt Reason I first signed into mastodon.social is because I did not realize I could sign up at other sites to begin with. I suppose has to do with the latest advertising of this site, but what if every instance shows or allows you to choose the site you want to join? Then people would have an obvious choice from the start

@bcrypt @tim @Gargron I mean something in the UI like a dropdown listbox. Currently you need to fetch the list of servers from the github doc, and I rather suspect there are more servers than those listed.

@frankofsandiego @bcrypt @tim @Gargron I've been driving newbies to http://toot.today as a starting point because it gives info (if provided by admins) on instance culture in addition to uptime, etc. It seems to be working out well.

@bcrypt @Gargron @tim does mastodon deliver all messages to all instances?

@bcrypt @Gargron Yep. Centralisation is efficient. Still, lots of different power dynamics here that we can study. #crowdpsychology

I expect the #Fediverse to break up into many smaller pieces, as I've elaborated on previously: https://medium.com/@2W/there-will-be-more-intermediaries-not-less-63a591d4c8eb

@bcrypt that would seem to only affect maybe Unlisted posts, no? (i.e. Public is already expected to be global; private and locked limits scope as intended. This assumes proper encryption of course.) @Gargron @tim

@bcrypt @Gargron @tim which is not even to mention identi.ca, which is literally the predecessor to GNUsocial/mastodon!

@bcrypt @Gargron @tim power law distrubutions get you every time