Conversation

Notices

1. Embed this notice
   Alexandre Oliva (lxo@gnusocial.net)'s status on Wednesday, 16-Nov-2022 11:14:44 JST Alexandre Oliva

   • Javier Alvarez
   let's take banking. it's been federated forever. you need an account before you can transact. but you have to first pick a bank to set up an account, a decision point that introduces such huge friction that banking will never succeed. do I get your argument right?
   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Wednesday, 16-Nov-2022 23:43:34 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     how's that different from federated servers? you have to choose a bank upfront, as much as a server. after that, both mediate and simplify your interactions with the network.
     banks have a profit motive indeed, which creates some biases in the choosing processes, but your distinction seems to rely a lot more on preexisting familiarity with the system than on any material differences
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Wednesday, 16-Nov-2022 23:43:35 JST Javier Alvarez

     in reply to

     @lxo :) so heard me out. Banks are fronts that simplify the federated network that manage money. They exists so you don't have to think about the network. And they are motivated to exist because they make a profit for providing you that service. For our use case there are no "centralized" or "simplified" gateways to the network ( or they are not well known yet).

   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Thursday, 17-Nov-2022 04:53:46 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     how is choosing a bank and a branch thereof for your banking address any different from choosing a home instance for your social media address? in order to wire money to someone else, you have to know their address; this is hardly different. I don't see what distinction you're trying to make. it comes across to me as "I'm used to this, so this is easy, while I'm new to that, so that's unfamiliar and therefore it seems different and complicated and scary" even though it's the really same
     take telephony. you also have to pick a phone service provider, and then find out others' addresses (area code and phone number) before you can call them. and, again, the difference between that and fediverse federation is that you've long been so familiar with the telephone federation that you don't even perceive it as such. ditto email. ditto XMPP. ditto so many political structures. the centralized walled garden make things simpler for exploitation (no way to escape), not for users.
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Thursday, 17-Nov-2022 04:53:48 JST Javier Alvarez

     in reply to

     @lxo that is precisely my point :) The banks have already solved the federation problem for you. You then have to pick the one you like the best. #mastodon still has no "banks". We are still interacting with very "raw" services.

     Another way to see it is on how the web got started. The protocols existed and the nodes could talk to each other. But it wasn't until AOL "packaged" it that we started seeing massive adoption of the network.

   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Friday, 18-Nov-2022 13:04:35 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     you seem to have changed your stance from "federation can never succeed" to "lack of awareness hampers adoption". I'm glad we agree now.
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Friday, 18-Nov-2022 13:04:36 JST Javier Alvarez

     in reply to

     @lxo Because the idea of a bank or phone company (for the masses) is no longer scary. It is an understood entity. There is nothing to figure out, and there are "no consequences" to pay if you choose one or the other. You have to remember that even the word "server" is scary. If, for example, there was a "friendly" layer on top of the network that called them "communities," that would reduce the friction of adoption.

   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Friday, 18-Nov-2022 16:23:57 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     that was quite a lot to unpack from the quotes around never in the head (?) of the thread
     https://hachyderm.io/@jmangt/109348887292670081
     
     and still I feel you're making a distinction that doesn't hold. banking, telephone, email are federated structures got wide popular acceptance without hiding the underlying network, and requiring service users to explicitly make a choice of service suppliers to join the network, without centralized or seemingly-centralized service, or even without requiring boutique ways to join.
     retail bank branches are by no means boutiques. phone shops seem designed to repel customers. email had a wide reach before google, microsoft and yahoo turned an efficient federated system into a recentralized web-based abomination.
     
     so that's 3 examples that counter this theory. do you have any single example to support it?
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Friday, 18-Nov-2022 16:23:58 JST Javier Alvarez

     in reply to

     @lxo oh, my stand is not that the federation can't succeed but that the access to the "naked" network won't reach critical mass. Not without a "boutique" or "central" way to connect to the network.

   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Saturday, 19-Nov-2022 01:57:47 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     only because the assertion you posted claimed it "never" could. my mind irresistibly sets out to check the assertion and, because of the structure of the phrase, seeks a single counter example.
     now, you had quotes around never, so I took that as meaning you didn't mean literally not once, but more like virtually never, almost never. so the assertion checking involved looking for situations in which it held and those in which it didn't, and expecting the former to be a supermajority. which is still the opposite of what I see.
     now it looks like you're saying the "never" meant something more like "whether or not", i.e., being federated doesn't affect the outcome by pretty much leading to failure, as a reading of your claim might suggest, but rather that some details of how federation is exposed to users of the service may have some effect on it.
     that feels like moving the goal post, but the new question is a lot more interesting indeed.
     I'd pose that they offered something sufficiently interesting for enough people to overcome the barrier of understanding the federation model as to make it common knowledge.
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Saturday, 19-Nov-2022 01:57:48 JST Javier Alvarez

     in reply to

     @lxo I think you are looking at this from the point of view of whether a federated model can succeed.

     The question you should be asking yourself is: What made those federated networks succeed?

   • Embed this notice
     Alexandre Oliva (lxo@gnusocial.net)'s status on Saturday, 19-Nov-2022 11:23:58 JST Alexandre Oliva

     in reply to

     • Javier Alvarez
     I understood that's what you've been talking about all along. I just disagree that it's such a big deal. yeah, I get that centralized services don't involve that step. I get that making choices before you know what they amount to can be quite hard for some people. but people still learn enough to choose a place to live, a bank branch to set up an account, a phone company, an ISP, and those choices often have a lot more impact than choosing a mastodon server. switching is so easy, especially early on, that I believe the solution for this non-problem is to state early on that, if you don't like the community of the instance you end up at, it's easy to move later, so don't sweat over it.
     but there are other non-problem design choices that seem to make this more complicated, such as the conflation of instance and community, the lack of a "choose for me" option that picks an open instance at random (with documented caveats), or a "give me a short tour" that names a handful of instances for the new user to pick from to get started. wording that makes the choice less of a commitment than it seems to be taken as now is likely to lessen the apparent burden that leads to this misperception of difficulty. but the difficulty is a matter of perception. people have no trouble deciding whether to set up an account on twitter, or facebook, or instagram, or some or all of them, and hypothetical intercommunication possibilities between them wouldn't change that.
   • Embed this notice
     Javier Alvarez (jmangt@hachyderm.io)'s status on Saturday, 19-Nov-2022 11:24:00 JST Javier Alvarez

     in reply to

     @lxo Here is an example of what I am talking about.

     https://twitter.com/owillis/status/1593608286460420101