Perhaps this should be not included if it’s broken/private?
EDIT:
Formatting issue:
Also:
Perhaps this should be not included if it’s broken/private?
EDIT:
Formatting issue:
Also:
Hello I’m using the ChatGPT (
chatgpt-4o-latest
) API for daily/weekly summaries since a few weeks and that’s a beast, the best I’ve ever tried.
We use o3 high on the Open AI forum (only visible to tl3) and it is really stellar, if people can afford a reasoning model, it is a very interesting choice. Gemini 2.5 pro has also provided very interesting results.
This is slightly a nitpick but these blurbs feel a little silly because of how artificial they sound.. I’m not sure if that’s intended from the prompt but I feel like they should be gone
Somehow silent=true
didn’t work today. At least I was notified that my post was linked.
discourse-categories-suppressed deosn’t supress notifications/mentions though?
I’m totally for the bold word to be changed from “Muted” to “Suppressed”
Orr the Bot could create a new Topic for each Summary
The ?silent=true
at the end of links causes no notification to be sent when the post is linked. AFAIK Meta doesn’t use that plugin. The summaries are made using AI (perhaps the URL param needs to be emphasised in the prompt?).
yes query
is not talking with path
?
perahps the catgories suppressed plugin meta uses is superceeding this logic?
Why do you think Meta uses a category suppressed plugin?
Being able to watch, track and mute categories is part of Discourse core. It’s for example explained in the guide for new users
But muting a category doesn’t help. Half of this topic is about the problem that @mentions and links still trigger notifications. That’s why the mentions are formatted without @
and ?silent=true
was developed.
Yes, as Moin said, the mentions and links are done as such to prevent notifying users.
For example, adding ?silent=true
or using <a class="mention"><a>
gives it the same look, without notifying the target user.
this wouldn’t be so valid then, it could say “Unlisted ” rather than “Muted”