Dynamically added content warnings for muted keywords, long posts, and animated imagery

this thread is for a feature set i have had i believe four different users request, that has a lot of power in it, and it is tied into the idea of combining content warnings and hidden images for better user effect: dynamically added content warnings for muted keywords, long posts, and animated imagery

Continuing the discussion from Reworking CW/NSFW image systems for better userflow and design:

1 Like

Wouldn’t having a mute but visible message enough and easier to understand?

“This message has been muted by your filters, click to show the content of the toot.”

this is essentially what the content warning does?

hell, the content warning could, in fact, be, the full text of your warning, that’s the idea.

by using the same UI as the existing content warning feature, and implementing it as such with the same backend code, it is functional with any client that supports the content warning feature, while having recognizable UI the user has seen before to make it more easy to parse. i do not understand how it is less easy to understand, i only see it as being MORE easy to understand.

1 Like

Why do you need automated content warnings for animated imagery when you can disable autoplaying gifs and you can see that it’s a gif?

I think it should be different UI because you want to avoid the confusion between filtering of your timeline and content warnings.

If I post a CW with the message used for filtered message, how would you know that it’s a simple content warning?

this is simply an explanation of how to roll this into the changes that i suggested in the other thread, and how to help add a universal symbol for “hidden content”, the read more button. i understand that we have a working function for this already. i am simply proposing all sensitive content be wrapped under a single system so as to allow the systems to be more easily understood by users.

EDIT: as far as i can understand your post it is “we already have a feature for this, why improve or iterate on it” and the answer is, simply, to improve and iterate on it to make it better, to unify hidden content with a single, global UI to always be recognized as 'this is sensitive, hidden content"

why would we ever want different UI? if anything we could have a symbol that is added to the content warning field that when hovered over, says “content warning dynamically added by filters, not part of original post”, or just that text. i do not see why there should be different UI, and in fact disagree and think there should be one unified symbol/feature for hiding content.

1 Like

“Auto-CW this keyword” is a pretty simple design decision and it’s how every other software with keyword-filtering works; although being able to say “just don’t show me the post at all” should be an option. If I’m getting harassed with a slur, I don’t want that slur appearing 100 times in my notifs as CWs…

“Truncate posts higher than…/longer than…” has been a request for a while and this is a good way to do it.

1 Like

agree one hundred percent, i just know this topic has not gotten any love or even a single response on github so i wanted to port it here. keyword muting entirely, as well as an option to allow dich muted keywords in your mentions is a thing I’ve wanted since the start as well. good to have that clarification, I’m sorry i forgot it.


It’s one of the most powerful anti-harassment features out there and Twitter has had it for a bit, albeit hidden away for some reason. It’s something we could’ve and should’ve had a while back especially since we already have RegEx.

1 Like

i requested it approximately forever ago, gonna make a seperate issue thing for it

linking this

1 Like

Would a counter work here? Tiny badge to show how many things have been muted rather than a bunch of collapsed toots. I’m thinking about this in the case of accidently adding a custom tag that resolves ambiguously/ends up grabbing more than intended, maybe coupled with an option to selectively remove the taged toots from the timeline? Kinda like a junk mailbox notification.

i… really dont know what you think a counter would accomplish. what is the intent? to not let people click the show more on content warnings? that’s the whole point of having a cw feature is to allow it to be less harsh than a full mute.

please do explain, i just don’t get it, your suggestion only worsens any content-warning-mute feature.

keyword muting in general could use a counter, for things that are fully hidden, but i am not sure how much value such a feature would have at all?

@shel wanted an option to hide the collapsed toots?
I’m just thinking of the UX issues we had with regex filtering not indicating it was being applied and the subsequent lost toot syndrome.

ah! this is a misunderstanding, we’re talking past each other.

so, wjat sje; was saying was she wanted this feature i already outlined: Keyword muting for anti-harassment and content filtering

namely, there should be two options for keyword filtering: the option to put it behind a content warning, and the option to fully remove it from the timeline, chosen when setting up the keyword filtering. this is all outlined in the other topic, a counter is entirely irrelevant to the feature shel was saying needed to exist alongside content warning mutes

Having a counter tell me how many people are calling me a tranny faggot in my mentions is kinda counterproductive to giving me peace of mind and not exposing me to harassment. The ideal situation is if I mute those words in my notifications I wouldn’t even know if someone is or isn’t calling me slurs

1 Like

Shel’s right about this – mute for a harassment context should absolutely not have a count. It’d defeat the purpose in some ways, really.

Having said that – the option of having a count may be useful to some folks for muting unrelated to direct harassment (say, “I’m not into Eurovision but I still want to know when people aren’t talking about Eurovision at that level any more”, or “I normally follow the news but everything RN is about LHC protesters, I’m interested but not THAT interested” or even “I need to take a day off X issue but I need to be aware if it happens to blow up so I know what to catch up on tomorrow”).

Maybe a tickbox that is off by default, but when ticked, enables a counter for filtered items? The default behavior would handle harassment in one click, but allow someone who’s curating to have finer control, IMO.

1 Like

Would it not also block your friend’s story about similar harassment and their request for support?

That’s why I think a counter is a good idea. Same reason we don’t give our spam filters total indemnity. Now, limiting that counter to people who we’ve let follow us/we follow, maybe a good consideration.

Or allowing the counter to only count people we follow.

1 Like

Yeah, sorry I was thinking about private accounts for some reason.