Largely agreed, although I'm conscious of the fact that folks seem to survive without most of these. For example, search is problematic in Mastodon architecturally, even beside the folks who simply don't want it to exist. I think these are all desirable, but I'm trying not to firmly assume them.
(a11y is a little more important than the rest, IMO, but I think is more about the clients than the protocols.)
I feel like Discord has a decent basic identity approach: you have an account, but that account can use different nicknames across your different group forums. You can even have multiple accounts to isolate things further. But true anonymity is hard.
I'm very conscious of this topic, given how much effort I put into Querki's three-level identity-management system. (Which is pretty much the state of the art under the hood, even though it's never been fully realized in the UI.)
That said, part of me would prefer for identity to be entirely outside the scope of Nusenet per se -- that it should be a consumer of a well-designed identity architecture, not a provider of one.
But the nickname point is a good one (and similar to Querki's concept of "Person"): it's sometimes helpful for the same public Identity to be customized per-community. (Separately from the ability of a single private User to have multiple public Identities.)
As for anonymity, I think folks have done a good job of hashing through most of the points above. Personally, I don't think anonymous posting should be possible -- the history of the Internet suggests that that's a recipe for disaster, although we might want to carve out a niche for opting in to anonymous posting on the small scale (specific posts for a limited time window) for specialized requirements. Anonymous reading is a very different question, and not at all obvious to me: there are arguments both ways, and it could be a per-topic setting.
It's a complex and important topic, and I'm not sure which bits should be part of Nusenet per se, and which belong more properly in a different part of the system.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Date: 2023-06-20 12:33 pm (UTC)Largely agreed, although I'm conscious of the fact that folks seem to survive without most of these. For example, search is problematic in Mastodon architecturally, even beside the folks who simply don't want it to exist. I think these are all desirable, but I'm trying not to firmly assume them.
(a11y is a little more important than the rest, IMO, but I think is more about the clients than the protocols.)
I'm very conscious of this topic, given how much effort I put into Querki's three-level identity-management system. (Which is pretty much the state of the art under the hood, even though it's never been fully realized in the UI.)
That said, part of me would prefer for identity to be entirely outside the scope of Nusenet per se -- that it should be a consumer of a well-designed identity architecture, not a provider of one.
But the nickname point is a good one (and similar to Querki's concept of "Person"): it's sometimes helpful for the same public Identity to be customized per-community. (Separately from the ability of a single private User to have multiple public Identities.)
As for anonymity, I think folks have done a good job of hashing through most of the points above. Personally, I don't think anonymous posting should be possible -- the history of the Internet suggests that that's a recipe for disaster, although we might want to carve out a niche for opting in to anonymous posting on the small scale (specific posts for a limited time window) for specialized requirements. Anonymous reading is a very different question, and not at all obvious to me: there are arguments both ways, and it could be a per-topic setting.
It's a complex and important topic, and I'm not sure which bits should be part of Nusenet per se, and which belong more properly in a different part of the system.