#microformats 2022-11-03
2022-11-03 UTC
jeremycherfas, jacky, ur5us, [manton]2, [manton] and [tonz] joined the channel
# 
[tonz] Question about rel-me. Say I have an organisation website, and several people in that org. If I, on that organisation’s domain add rel-me’s for all team members, And those team members add a link back in their profiles on Mastodon, will all those rel-me’s validate? Or does rel-me look for a specific or first rel-me mention in a page source. The scenario I’m thinking of is having rel-me’s for the entire team on the company’s
ur5us and [tonz]1 joined the channel
# 
[KevinMarks]1 That's not a great idea, as you're implicitly saying that you and your colleagues are the same person.
# [KevinMarks]1 You might want rel="group" and rel="member" links instead, though I don't know if there is a consumer for those https://microformats.org/wiki/group-brainstorming
# [KevinMarks]1 Or you could link to each other with rel="colleague"
[tonz], barnaby, [MiaCrow], [tonz]2 and jacky joined the channel
# [tonz]2 [KevinMarks] mmm, the point here is getting the company site ‘green’ on a Mastodon profile. Which only works with a rel-me.
[schmarty]1 and [tantek]1 joined the channel
# [KevinMarks]1 Could you make each person their own page on the company site and use rel-me to that?
# [tonz]2 we could yes, then you wouldn’t be showing the domain but a specific page in a mastodon profile.
# [tantek]1 [tonz] why would you think it should work? you are not your organization, nor is your organization you so it doesn't even make sense semantically
# [tantek]1 [KevinMarks] did we ever document the Chris & Tara & Citizen Agency problem? Where they all rel-me'd each other and they couldn't be distinguished?
# [tantek]1 verification only "works" with rel=me because it depends on people correctly using rel=me to actually mean hey this other thing over here is another representation of me in particular
# [KevinMarks]1 It was more obvious when we were running the social graph API at Google, because that did the full transitive closure, whereas most current implementations are single hop.
jacky and [chrisaldrich] joined the channel
# 
@MuseAppHQ Metamuse 67 // Dynamic documents
https://museapp.com/podcast/67-dynamic-documents/
With @mschoening and @geoffreylitt
• Data detectors, live searches, and language models
• The beauty of personal microformats
• The @inkandswitch research process
• Why Stable Diffusion is like a slot machine (twitter.com/_/status/1588213414765248512)
# [tantek]1 personal microformats?
[jgarber] joined the channel
# [tonz]2 [tantek] [KevinMarks] what is semantically correct in this case isn’t the starting point though, it is to get a desired effect in a particular case. Users, me here, be users, regardless of what designers imagined. I doubt one can say every rel-me can only mean the one person pointing to their personal page. In this case it is a _business_ account pointing to the business, where such an account on the company’s instance is represen
# [tonz]2 just the shard of me that is me at work. I can see how we might point to individual team member’s author pages on our company site, as [KevinMarks] suggested. But it e.g. would make no sense at all to point to other personal pages representing other shards of their identity to verify their _business_ account.
# [tonz]2 We even are actually all of us the business. In the process of making it worker owned.
# [tantek]1 abusing a technology to get a desired effect will nearly always end badly
# [tantek]1 rel-me for businesses that pretend to be individuals do work in the same rough way as courts (in the US at least) have decided that businesses can act like individuals, which is to say sort of, and only as long as you don't start collapsing business identities with specific individuals.
# [tantek]1 e.g. a business domain and the business's twitter or other social media profile seems reasonable
# [tantek]1 representing "shards" is not a support feature of rel-me. it's all or nothing, no finer granularity
# [tantek]1 probably indicates a different problem to be solved beyond current technologies
# [tantek]1 and sure, "individual team member’s author pages on our company site" looks similar enough to silo profile that individuals could rel-me to those if they wanted to
# [tantek]1 the question of "what is verification or what does verification mean" is a much larger question than any particular technology like rel-me
# [tantek]1 even if it is worker owned, you are not "all of us the business", any more than you are your car or bicycle
strugee and gRegor joined the channel
jacky joined the channel
# [tonz]2 off topic but “even if it is worker owned, you are not “all of us the business”, any more than you are your car or bicycle” is a odd comparison, companies aren’t objects and indeed they’re not individuals either, they’re groups of people. Of course we are _also_ what we are part of. Family, teams, community, are all like that. The group identities are the amalgam of the individuals, and the members determine the outcome.
# [tonz]2 company that’s the same. I don’t claim that’s true for all companies, or that it will remain true for ours should it get much bigger, but right now the identity of the company is fully determined by its team members. In that sense it’s a mutual representation of values. ‘abusing’a tech ending badly’ sounds a bit over the top, we’re talking the equivalent of using html tables for lay-out here. Not ideal, not intended,
# [tonz]2 functional. Anyway, enough said.
[jgarber] joined the channel
# [snarfed] the employee profile page idea (instead of the company's home page) seems like a reasonable compromise. lets them at least assert/prove that they are employees or otherwise affiliated, if not literally the company in some sense. this is common at universities, eg https://web.cs.wpi.edu/~claypool/
# [tantek]1 a business (even worker owned) is closer to a bicycle than a family. you buy into a business, and similar you can sell your share and leave. similar to buying or selling your ownership of a bicycle. none of that applies to family, nor even "community" IMO
# [tantek]1 pretending that a business is a family is a horrible capitalist trope that's been quite well debunked as a mechanism of manipulating workers into investing more emotionally into a virtual entity which is not conscious and cannot and does not reciprocate, no matter how much it may be personified
# [tonz]2 [tantek] I don’t compare the company to a family, merely stating both are groups of people with a number of communalities. And that each group creates its own cutlure / identity. That people can join or leave whatever group does not negate that.
ur5us joined the channel
# [tantek]1 It's precisely that people can join or leave that makes it different. This is as different as the difference between "has a" and "is a" relationships.
bterry1, bterry2 and GWG joined the channel