A while ago there was some good discussion on post attribution, as in apps or systems used to create posts. Did that ever go anywhere?
@nickposted at 2:20 PM with [poster-boi-icon] PosterBoiI love fall!
A while ago there was some good discussion on post attribution, as in apps or systems used to create posts. Did that ever go anywhere?
@nickposted at 2:20 PM with [poster-boi-icon] PosterBoiI love fall!
In my opinion this is stuck in a state of “is this actually a good thing to support”. If folks want to publish this metadata it is pretty easy to do so, eg with a sidecar. The question is which apps would adopt it, both for display and for annotating the records they write.
The history of Twitter having done this, and then removed the feature, looms over the whole thing. There are clear wins for client developers (visibility of their software). There are more subtle privacy and social dynamics (who is using what software, which posts were made by which individual/device) which are higher-stakes with global broadcast data. I know there are some folks who think this is obviously a net positive, and some folks who think it is obviously an anti-pattern, and a lot of folks in the middle who don’t feel strongly.
So I don’t think it is a “somebody just needs to do the work” sort of thing. Though maybe somebody doing a proof-of-concept would result in a clearer ecosystem-wide assessment and consensus around the design pattern?
Tbh I haven’t heard anyone speak against this as an idea… just that it would be good to have it, but non-Bluesky client apps even coordinating between themselves on a shared field would be kinda pointless until Bluesky displays it. But with Blacksky, Nothsky, Witchsky, Mu etc. getting more use these days, maybe this would be doable.
Checks out. I’m coming at it from a transparency angle for applications and tooling, specifically in the context of AI / ML / CS applications and agents. I think it would be a useful and informative disclosure for authorship annotation and attribution.
If you saw the following disclosure, how would it change the way you view the content?
@nickposted at 2:20 PMI love fall!
Posted using
@PosterBoiBot
I soft-agree that a sidecar makes a lot of sense for this use case.
I’ve been talking with a few other AT Protocol people about the idea of attribution, but in the context of attributing a work to a creator or subject of a work. Like say a photo is uploaded as a blob, using a lexicon to attribute the photography to one did and the people in the photo to other dids. Which leads to also having to have a system of accepting the attribution through some sort of labeler. Like if someone posts an embarrassing photo of you, you’d want to be able to label it in a way that removes your attribution from app views maybe.
In this context you’re talking about, I think the word “source” is better than “attribution”.
We are actually implementing this for the hypercerts and certified lexicons. And we are following your attestation logic @ngerakines.me
A record has a signature field, e.g. https://github.com/hypercerts-org/hypercerts-lexicon/blob/main/lexicons/org/hypercerts/claim/activity.json
And the signature can have a type #createdVia: https://github.com/hypercerts-org/hypercerts-lexicon/blob/main/lexicons/app/certified/signature/defs.json
This way we know which apps created or last edited a record.