Reworking CW/NSFW image systems for better userflow and design


#26

here’s a mockup of how to better visually distinguish and explain things without any tutorial.

when you add a content warning, it would show the show more button in the center, and a line to show where it breaks, along with faint text that dissappears whenever you type anything into either field to better explain it. if the default-images-to-sensitive option is selected, the blue eye would be crossed out, otherwise it just turns blue instead of grey.

EDIT: probably also edit “hidden until show more is clicked to” “hidden text and images till show more is clicked. images default to hidden even after clicked.”

EDIT: also note the pixel values are obviously off i literally did this in MSPaint on a laptop.

the showmore would be centered, the top and bottom would default to the same space vertically to show you can put however much text you want in either, and it would adapt to text to expand either way as necessary


#27

I love this and would even take it one step further with the functionality concept: treat the top part as the main post and have the second part be the optional “behind the cut” section. Do you think that would work?


#28

My 2 cents: As a user, I don’t want to care about the differences of text vs image. If I put up a toot with whatever sensitive content, I activate the warning for this toot. If I have sth. to say which is not “warning-worthy”, I write it as a reply or follow up to my own toot.

Keep it simple. Everything else seems just pretty confusing.


#29

i really don’t understand what you are trying to imply, but, frankly, if we rework the content of either box in terms of explanation as to what they are used for, or otherwise rework the intents of the warning functionality, which is vital, i only see the feature suffering for it

if we say that the “warning” section is the main section, then suddenly it isnt a thing being added and the descriptive text kind of fall apart.

that said, i can’t tell if you mean to change the descriptive text, or which text box holds the text you type BEFORE clicking the CW button. if you do mean the text-carry-over, i would argue it is counter-intuitive to user flow, and here’s why.

A content warning is, just that, a warning, and generally is NOT going to be what you think first. generally would write as flows in your mind, then say “hang on, this is a bit explicit” or another reason to tag it, and thus tag afterward

if you mean that the orientation (top and bottom) should be switched, i would argue this is counterproductive, as it makes the readmore button that so simply explains the concept not make sense anymore, as you would have to read bottom to top.


#30

Content warnings don’t support rich text features* (mentions or hashtags or links) so while they share the maximum character count with the toot, using them instead of the actual toot text is not advised. They are meant as a label on top (like a headline or summary, if you will).

*: Technical limitation, tried, didn’t work, wontfix if you want to keep the functionality of copying CWs when replying

When hiding images, you usually mention in the toot text why you do that. That becomes the colloquial “content warning”. I do not believe that the toot text needs to be hidden as well, behind a second level of click-through, for that use case.

Short of changing “nsfw” to an eye icon to be consistent with the show/hide image toggles on the actual toots, I think the system as is does not need changing. Besides, on API level you have the choice how to interpret this stuff, so if you want to make apps that behave differently, be my guest - there’s the presence of spoiler_text, and the sensitive boolean, how you wanna display them is up to the app.


#31

From my perspective the only change to this functionality needed (and it is NEEDED) is that images should be automatically hidden when you add a CW - you can let ppl unhide them when composing, but i have seen multiple users post sensitive imagery with a CW but with the image not hidden. It’s simply not intuitive to users that doing that would not result in the image being marked as sensitive.

I just don’t think that merging the two functions into one is a great idea.


#32

It does need changing. We’re not asking for you. We’re asking for us.
If someone tags drugs and I can see a bong that completely defeats the point of tagging the text drugs.
If someone tags food and I can see a cheesecake that completely defeats the point of tagging the text food.
If someone tags lewd or kink and I can see a paddle that completely defeats the point of tagging the text kink.

All posts with attached images and a content warning should be considered sensitive.
Users expect it to - I’ve seen users old and new argue with this poor UX decision day after day.
We do not need the granularity of hidden text, shown image.


#33

I’d argue it’s a given that content warnings should always be located in the same space of the toot especially when there’s a a section explicitly labelled for that purpose.

Additionally the tool as is is frequently misused, so many hastily deleted toots followed by ‘oops forgot to hide image’ even from experienced users. New users have an intuitive understanding that pressing the button should hide the contents of the post and that’s clear from the last influx of new users many of whom made that very mistake.

It’s simply poor design to continue to work against how people instinctively assume the functionality works, particularly in the case of content warning and image hiding systems because it only takes one mistake to cause harm. We can’t prevent all mistakes but we can reduce them quite considerably with this change.


#34

it seems @MililaniVFand @MadDog have both explained WHY the function should be changed, but i would like to explain an enhancement this allows, that i have had requested several times.

the ability to hide long posts by default behind a content warning, and the ability to hide animated gifs by default.

both benefit by images getting treated in the way i outline, as it allows a one-button view to said things.

as such, we could create mutes “mute posts longer than X” and “hide animated gifs” in the settings menu, and make long posts get tagged “long” and posts that contain animated gifs “animated” so people know why it is hidden and can choose appropriately whether to view. this is a bonus IN ADDITION to better UI and design, that allows this feature to be improved a large amount.

dynamic mutes could also be added. “mute posts with X in the text” and they would be put behind a content warning of said keyword or otherwise fully hidden, to allow several levels of muting for best experience


Dynamically added content warnings for muted keywords, long posts, and animated imagery
#35

giving it some more thought i think that merging sensitive image/cw would be okay. its a tiny bit of granularity that is unnecessary and can be sacrificed for better UX

I take issue with what you said here: it’s only the way it’s done because the UI/flow is as it is - it’s quicker to write in the body and mark as sensitive than tag as cw, write in the cw, AND mark as sensitive. It’s a self fulfilling prophecy. The idea is to get rid of that last click. + it encourages good form
As for the second clickthru - it seems to me that the idea is that if you click a ‘show more’ it’ll reveal the image and text at once?


#36

@puphime [quote=“puphime, post:35, topic:200”]
As for the second clickthru - it seems to me that the idea is that if you click a ‘show more’ it’ll reveal the image and text at once?
[/quote]

exactly. a second option would be allowed, to add the sensitive image toggle to default to hidden for when you want to put more explanation under the first content warning and not show the image immediately.

the only changes suggested beyond UI tweaks are as follows:

make images act as part of the body of the post, so cw hides it
do not allow image to be set to default to hidden UNLESS there is already a content warning in place, which is just to ensure that users put a tag of some sort on any hidden content, as it is just better practice for that to be the case.


#37

side note: making images act as part of the text of the post should also allow users to remove the image link and still post an otherwise empty text body, which is currently not the case, and would fit in with the logic of the feature both for content warnings and not.

it doesnt make much sense for the UI to proclaim, when a toot is not empty, since it contains an image, “cannot post empty toot” or the like.


#38

I strongly support making image-hiding work like users expect, i.e. be linked to the CW. Given that massive simplification, this additional toggle complicates the user experience significantly while only being helpful for a tiny margin of edge cases. I think fully merging the two hiding toggles would be the most easy to use and intuitive, and the least cluttered, of all the options discussed here.


#39

i am not against a second level of hiding things, there are visible use cases, but i agree the simplest improvement is totally merging them. i could see it go either way, that the default be to hide the image even after cw is expanded, or to not default it and say the content warning with image symbol after is warning enough. i dont know


#40

There certainly are use cases for the second layer, but they’re edge cases, not the usual use of this feature. What I’m saying is that I think the general simplicity and usability of the UI should take priority over such edge cases.


#41

i dont know if i agree with this but since i don’t have the ability to implement the feature changes, i guess whoever implements it makes the final decision now. i only hope my words and suggestions have been heard.


#42

Someone on Mastodon said my rhetoric was too quick to get hostile. But being we don’t talk about Mastodon using Mastodon anymore, let me explain my hostilities.
We already have functionality as we describe.
This is the simplest change I’ve ever asked of Mastodon.
Take the the event that activates flagging a post as NSFW and giving text posts the hidden #NSFW for federating purposes already and the feature that gives you the ability to put CW into one button, the eye graphic that we already have. Open eye unpressed, closed eye pressed.

This isn’t about the fact that NSFW/CW is dumb. This is about the fact that this is the third non-Mastodon software used to talk about Mastodon.
I want one thing. The inability to have a content warning and a shown by default image.

Collapsed a longer section, see moderator notes for details

This is your chance, @Gargron @maloki @noelle and everybody else who is now officially part of Mastodon.
Are you listening to us, using the third non-Mastodon software to talk about Mastodon (previously Discord and Github)
Are you listening to us? Are you listening to us? Or are you ignoring us, like Gargron always has, like the Gargron that told us “We could just leave” when we said Mastodon wasn’t safe for us to be on? Are you ignoring us, like Gargron has in the Discord time and time again? Or are you listening? [strong profanity removed by @chrismartin] Prove it. Prove it to us. Prove to us you want Mastodon to be more than Gargron’s toy he shares and it can be the implementation of OStatus for users, even the least technical of us.
If you’re not, we should “just leave”, as Gargron once told us to do. But we’re still begging you for the ability to delete accounts - something you once told us you wouldn’t do, and then said you will do, and haven’t months after the fact.

We’re waiting to for you to listen to us, when we talk on your terms. The ball is in your court now.

[moderation: Hid a section which is quite inflammatory, as our guidelines would call for it to be removed (“In order to maintain our community, moderators reserve the right to remove any content”). However, only hid it, to show that we do listen, and we are not trying to “censor” the meaning of the post, but for peace of mind and calmer conversations we have hid it. Please read the FAQ. This is also to be considered an official warning and strike 1, for failing to “Always Be Civil”, and “provide reasoned counter-arguments that improve the conversation”]


#43

I actually like the granularity right now.

I’ve definitely posted jokes that were fine with pictures that were ambiguously not safe for work so I tagged the pictures and gave my rational. @Gargron 's points about CWs not being rich text is pretty valid IMO. I’ve been burnt once or twice by CWing text and forgetting to tag the image NSFW but I think this has more to do with maybe improving the workflow assumptions rather than getting rid of the granularity all together.

Maybe the system should treat tagging the post with a CW as defaulting to both CW and NSFW on the attachments while still allowing for the full combination of possible toots as now? E.G. Specifically untagging NSFW unhides pictures from the default.


#44

I see your concerns @MadDog , and I would like to get some facts straight. Most of it will be hidden by a summary, as it does not belong in this post for discussion, but contextually it belongs here, but I would prefer it if we do not continue that discussion here.

Yes, we hear you

I’ve been, as posted here, busy with my thesis the past few days. And have basically just woken up after the 16hrs of sleep I required after I wound down, after submitting the thesis on Wednesday afternoon, Central European Time. Me, personally, as PM, I only work a maximum of 20hrs / week, and I have CFS (Chronic Fatigue Syndrome), which means that I will not be immediately available on here all the time, but I will do my best to check in regularly. And if you check my post history you can probably see that I have for the past few days not been active (due to the thesis already mentioned), but before that I was quite active.

The following will be posted elsewhere as well, but just not right now, because I am still incredibly foggy.

The discourse is here so we (mainly Gargron and I, but also the community moderators (who’s job description may be extended but that’s a different discussion), can allow the community to discuss between each other ideas, and then dip in and see what is requested.
I have explicitly told Gargron to NOT attend the forum to any major extent, as we are going to sum up and show him what is important. There are a number of reasons for this, and one very important one is emotional capital, it costs a lot to read about, sometimes very strong, criticism of your baby. And it is his baby, no one should deny that, nor wish to take it away from him.
The discussion threads here, he does not need to be a part of. They are to provide a good collected space for the community to discuss features properly, before we implement them. Having Gargron reply to every single thing, would actually just hinder the full discussions by the community. The same actually goes for the moderators/admins and I as well. We do not need to be here to course correct the discussion.

As for the CW/NSFW, I have been listening, and I am incredibly happy to see proper discussion on this matter. And the mockups are helping a lot to visualize.

We have in the past had problems with wanting something to be implemented, but not having a good space to discuss it which isn’t github, but where we can still collect the details and make it into a good cohesive github issue which someone can then utilize to implement features in the way we (the greater community) would prefer them to be made.


#46

This is a really good mock-up.

It does help a lot with increasing the UI for better UX