💾 Archived View for station.martinrue.com › skyjake › 6a12de47d979434aac73b7b69343615f captured on 2024-08-25 at 04:04:21. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2024-05-12)
-=-=-=-=-=-=-
Small Cosmos tweak: publications like smolZINE are always sorted to the top of their day and can only be the root of a thread. @kelbot
As zines are not regular log posts they may link to many things and if one or more of the links happen to be feed posts, it’s not appropriate for the zine to become a thread reply to one of them.
gemini://gemini.cyberbot.space/smolzine/smolzine-issue-18.gmi
2 years ago · 👍 krixano, kelbot, resetreboot, lykso
gemini://gemini.cyberbot.space/smolzine/smolzine-issue-18.gmi
GeminiLUACH
I'd not heard of this before, but I really like this concept. Maybe I'll incorporate this (or some variant, if I find it ill-suited in practice) into my gemsite. 🤔 · 2 years ago
I can see it must be hard to try to infer when posts are linked but not "in conversation". Still its nice to see your experiments in this area in Cosmos, good luck! · 2 years ago
@marmaladefoo
GeminiLUACH
Yeah, having additional semantic information would help a ton with figuring out how to detect and present discussion threads.
Unfortunately, unless this is really strongly supported by whatever tool an author is using to write posts, I suspect only a small minority of authors will use it... · 2 years ago
@marmaladefoo Given the tight constraints in analyzing post contents and presenting the results, I think that linear threads manage to give some context while also showing the rough semantic relationship and, importantly, chronological order.
The general problem here is what to do with posts that have multiple meaningful links to other posts that are not in the same "conversation". I don't want to show items multiple times (trying to stay compact), so they shouldn't become replies to any one post. I've yet to implement this as a general rule. · 2 years ago
... Or ... you could have SmolZINE as both a response to other posts, and as its own top level if you think it is privileged in some way · 2 years ago
Everything is intertwingled. We can't really say on a network which node is at the top of the graph. unless we know the semantics of the link, which is not generally available. I wrote a short suggestion/proposal a while back about how we can markup our links with a bit more context. Something like this can help, even if this is not the answer https://github.com/LukeEmmet/GeminiLUACH · 2 years ago