I’d like to invite you all to share your thoughts and ideas about Lemmy. This feedback thread is a great place to do that, as it allows for easier discussions than Github thanks to the tree-like comment structure. This is also where the community is at.
Here’s how you can participate:
- Post one top-level comment per complaint or suggestion about Lemmy.
- Reply to comments with your own ideas or links to Github issues related to the complaints.
- Be specific and constructive. Avoid vague wishes and focus on specific issues that can be fixed.
- This thread is a chance for us to not only identify the biggest pain points but also work together to find the best solutions.
By creating this periodic post, we can:
- Track progress on issues raised in previous threads.
- See how many issues have been resolved over time.
- Gauge whether the developers are responsive to user feedback.
Your input may be valuable in helping prioritize development efforts and ensuring that Lemmy continues to meet the needs of its community. Let’s work together to make Lemmy even better!
It totally would. In ActivityPub, all objects (like users and posts) have an identifier that includes the domain name. For instance, your ID is
https://midwest.social/u/m_f
. That’s what identifies your user. There is no way to change an ID - the point of an ID is after all that it stays the same and still refers to the same entity. This is a pretty serious limitation of ActivityPub right now unfortunately.I think there’s a FEP that could (or fixes) this. To my understanding ID can be any URI, so there are better ways. I guess it’s hard because it would brake a lot of stuff or how mastodon is build.
Any FEP trying to fix this will be incompatible with existing instances, so I don’t really see how it’s gonna work.
Yeah, it sucks. But I think that at a certain point it will need to happen if we want to make ActivityPub better with better portability.
I wonder who was the idiot who made a persistent ID for identity reliant on a third party factor that can be trivially taken away.
Any plans for solving it that are known?
Not as far as I am aware - I don’t think you can really fix it within the protocol, i.e. without a breaking change. Then you may as well make a new protocol.