Follow-Up Poll: Expanding "Mute"

avatar

Thank you to all the Steemians who participated in last week’s poll. In that poll we asked whether you thought that when you Mute another user, their comments should be hidden from your blog for other Steem users, not just you.

We received a lot of great feedback from users who both wanted this feature, and those who voiced concerns like @anyx, @blocktrades, @jarvie, and @holoz0r. @anyx’s comment was especially insightful. He said:

Absolutely not.
Plagiarists, identity thieves, or other fraudsters will mute comments from anyone calling them out on their actions.
This means the @steemcleaners and @cheetah efforts would be useless, so you could say goodbye to them.
The appropriate reaction to comments one does not like is downvoting them.
A better solution to "the problem" is user opt-in mute lists that are community maintained. A community should be able to publish a list of accounts they mute. A user can opt-in to using this list, or the set union of multiple lists they subscribe to.
If one has ever used adblocking, it would function like this -- lists one subscribes to enjoy a curated experience.

The Poll


After considering this feedback, we would like to propose an alternative implementation that we believe addresses the valid concerns that were raised. If you believe that we should integrate this alternative implementation (outlined below) into steemit.com (Condenser), please leave a comment that begins with “YES.” If you believe that steemit.com should stay as it is, please leave a comment that begins with “NO.”

Alternate Implementation

  1. When an author mutes a user, the muted user’s comments will not appear on the author’s posts; BUT,
  2. A prominent note will display when any commenters have been muted by the author saying something along the lines of, “The author has hidden replies from @cheetah, @worman, @spammer99. Reveal all?

With this alternative implementation, it would be easy to identify exactly which accounts are affected, and reveal posts with a single click. If a user is particularly malicious, then the visitors to their blog will see an extremely telling list of all the accounts that have been muted by the author, which would be an obvious red flag.

What do you think? Would this constitute an improvement over the existing UI?

If you believe that we should integrate this alternative implementation into steemit.com (Condenser), please leave a comment that begins with “YES.”

If you believe that steemit.com should stay as it is, please leave a comment that begins with “NO.”

Thank You


A special thanks again to everyone who participated in last week’s poll, with a special shout out to @anyx for his very constructive feedback. We are committed to making steemit.com a site that works for its users and we believe that it is through conversations like this that we will be able to do that.

The Steemit Team



0
0
0.000
97 comments
avatar

I say no to this version, though the curated community list is a fine idea. Reason being is I have muted a bunch of Bernie bots because he was leaving photos of literal shit and dicks on my friend's posts - like, dozens and dozens of them.
So if my friend who he doesn't like comments on my post, and his bots leave nasty pictures (honestly, I don't know if he's still spamming with photos or not, this happened months ago), there's going to be a mile long list of like 75 tags getting his attention so he can come flagging. He already flags me if my friend comments or resteems me anyway, but man, I'm just trying to live my life and not see this unnecessary shit. To make it a list of tagging who's muted would invite more bullying.

0
0
0.000
avatar

If the muting also blocked downvotes then it will actually make freedom of downvoting obsolete, I guess I answered my other post

0
0
0.000
avatar

Right? I'm not sure if there's a technical solution that would solve targeted bullying without also defeating the purpose of having flags in the first place. Especially since so many people have approximately ten trillion alt accounts.

0
0
0.000
avatar

There could be a committee of elected users with some delegated Steem power from Steemit, Inc. which could decide (in case someone complains) if flaggs are justified or not, and if "yes" just counter them with upvotes.
In addition, accounts who repeatedly misuse flaggs in an abusive way (instead using them against spam, plagiarism etc.) could be flagged, as well, after a decision of that committee.

0
0
0.000
avatar

What about totally blocking someone who is randomly, blindly downvoting all your every post and comment

0
0
0.000
avatar

That's what MUTE is for, but it affects your own perspective. OP is asking about MUTE affecting perspective of other viewers on the authors blog.

0
0
0.000
avatar

Yes

This is better as it would show if known abuse fighters had been commenting. There are malicious people who mass comment spam and we need defences against them.

0
0
0.000
avatar

Yes. This is better than nothing.

0
0
0.000
avatar

No.
Although this version is better because it allows the user to see the post, it still can give the author the ability to censor what the reader can read. Most readers won't bother clicking on the Reveal all button, so it's mostly the same as your previous proposal.

I don't understand the reason why this is even needed. The observed quality of any post is never changed if spammers comment under it. No one cares. Only if the commenter has a valid point against the contents of the post(by for example revealing plagiarism), he can change the observed quality. And in that case we shouldn't help the author of the post to mute the comment, as he is the one who did something wrong and deserves less payout.

0
0
0.000
avatar
(Edited)

Actually many news sources don't even allow comments anymore due to the absurd trolling that often goes on. And, many other social sites allow the owner to moderate their own content and have control over what their readers see. Overall this just makes steemit.com less of a social media site. Though I think we already knew it wasn't one.

0
0
0.000
avatar

No, these changes will not improve user experience. Spam is already flagged and hidden.

0
0
0.000
avatar

Also if the user spamming under your post has more Steem power than you?

0
0
0.000
avatar

Meh

0
0
0.000
avatar

Right, especially if the author hasn't > 100 k SP like you ...

0
0
0.000
avatar

I don't think this brings anything good to the table. It's not about how much SP I have. I can think of situations where it could help clean up an author's comment section. However I can also see it leading to an author being more able to spread disinformation and removing all critical comments making it seem legitimate. Sometimes you have to accept the good with the bad.

0
0
0.000
avatar
(Edited)

However I can also see it leading to an author being more able to spread disinformation and removing all critical comments making it seem legitimate.

That's one side of the medal. On the other hand I see many flags given just because of different opinions, personal animosities or even for fun.

Maybe a committee of elected users with some delegated Steem power from Steemit, Inc. might help, which could decide (in case someone complains) if flags are justified or not, and if "yes" just counter them with upvotes.
In addition, accounts who repeatedly misuse flags in an abusive way (instead using them against spam, plagiarism etc.) could be flagged, as well, after a decision of that committee.

If abusive flags could be countered like that, I would agree not to take further measures to protect the blog or an author.

0
0
0.000
avatar

Muting a user is not going to stop them from abusively flagging you. Assholes are gonna asshole. We have to accept some bad things are going to happen sometimes. However this proposal has the potential to open up abuse that is far worse. Maintaining an environment where people can't pass off bullshit as truth, where open dialogue can be had... Is extremely important (imo).

0
0
0.000
avatar
(Edited)

Maintaining an environment where people can't pass off bullshit as truth, where open dialogue can be had... Is extremely important (imo).

You have a valid point here, but I also insist on mine: too many people left or even didn't/don't/won't join STEEM because of omnipresent flag abuse (just recently a potential investor from Switzerland told me he saw all these flaggs, even under official Steemitblog posts, and thus won't buy STEEM for sure).
We Steemians are so accustomed to this that sometimes we aren't aware anymore how devastating the impression for people outside of our microcosm is ...

It's OK for me not to implement the new muting feature, but then there should be other solutions, like for example the idea of an 'anti abusive flag committee' which I suggested above.

0
0
0.000
avatar

But flags (downvotes) are a part of content value discovery ... Their use is supposed to be how the system works. The coming downvote pool will make this more obvious I suspect.

0
0
0.000
avatar
(Edited)

But flags (downvotes) are a part of ...

Of course they are. I never wrote anywhere I was against the option to flag.
But many flags are not used in that sense. For example for quite a while every single comment of @valued-customer got flagged automatically by a whale. That had nothing to do with discovering value or preventing spam (actually the whale added a spam comment under every flagged comment). And that's just one example among many.

If there was an 'anti-abusive-flag-committee' its members would recognize such kinds of flags very easily and could counter them.

0
0
0.000
avatar

why not just leave it the way it is?... is this really a priority right now? 🙈🙈🙈

0
0
0.000
avatar

Then you should respond NO

Posted using Partiko Android

0
0
0.000
avatar

don't waste your time policing other people's comments 🙈

0
0
0.000
avatar

If you don't say the word "no" they wont count your vote. I agree with you doitvoluntarily. You need your voice heard. But in the format they are asking for. They will not count what you say unless you use that word.

Cheers man

0
0
0.000
avatar

I just thought, especially since you agree with me, that way your vote would be counted.

0
0
0.000
avatar

Shaking my head on the way you guys set priorities........

0
0
0.000
avatar
(Edited)

This isn't a priority. It's just one feature we thought might improve user experience on steemit.com. Clearly many members of the community agree, however, many others clearly disagree and we can't know that until we solicit their feedback. We've learned a lot by polling the community on this issue.

0
0
0.000
avatar
(Edited)

This isn't a priority. It's just one feature we thought might improve user experience on steemit.com.

What about a 'feature' cleaning once and for all the trending page as God demands for consumers outside. Enhance and revamp the chronological order of articles posted throughout every #tag here on behalf of discovery and code a proper internal search engine to really find valuable & relevant stored content within the platform at our queries to improve not just the user experience on steemit.com, but the experience of everyone worldwide?

0
0
0.000
avatar

No.
I appreciate the feedback you guys are showing lately but I don't think that is needed today.
I will stand for focusing on the most important things instead of this.
However if you have a team willing to improve condenser, I am sure that you will be able to find something that deserves more attention.
I am impatient to see SMTs and communities implemented, which I hope that will be able to structure better the information stored on the blockchain. Maybe prepare the condenser for this will be a good option.

Posted using Partiko Android

0
0
0.000
avatar

Hear you loud and clear! Thanks for the thoughtful response. Based on the mixed response we're getting I suspect you're right when you suggest that there are other potential improvements to condenser that we could focus on. Thanks again.

0
0
0.000
avatar

No, I think up and downvoting is better, alternative suggestion, automatic downvote comments from … (on my own posts, on all posts globally)

0
0
0.000
avatar
(Edited)

NO

I don't think the features should be merged. Hiding from seeing (muting) is not the same as wanting to hide from replies on a blog (moderating).

I think people being able to moderate their own blog (the latter feature) is okay as long as there is a prominent label indicating that poster moderation is active. And it might be debatable whether this should even be allowed on posts which aren't reward-declined, because if they aren't then it isn't really a 'private' personally-owned blog post but one which has chosen to participate in the shared reward pool.

Also consider how this will work with communities, which probably should be able to have their own rules on what sort of moderation is allowed or required for that particular community.

0
0
0.000
avatar

Great points here. It keeps coming back to communities with how best to improve the social experience on here. We can't get those shipped soon enough.

0
0
0.000
avatar

They're getting closer. And the great part is that it's not a hardfork need.

0
0
0.000
avatar

We need them done yesterday.

0
0
0.000
avatar

How about just moving them down (regardless of Sort Order)?
Or semi-hiding (like currently downvoted comments are shown) with info that's [MUTED BY AUTHOR]
Also, if visitor follows someone, who's muted by the author, then it shouldn't be affected at all (explicitly stated preference of visitor).

0
0
0.000
avatar

I really like this type of implementation.

0
0
0.000
avatar
(Edited)

I still don't know if it is systemically good for the author to be able to unilaterally hide comments from (nearly) everyone. In practice most people won't dig through [muted by author] comments and can easily be fooled by scammers, plagiarists, etc.

I prefer the downvoting model where it becomes a consensus, not unilateral, matter whether a comment gets hidden or not. Perhaps the author could get some extra temporary vote weight for this (considered paid for by the act of posting).

I very much question, particularly after HF21 with free downvotes, whether truly abusive comments won't get consensus downvoted by others. Already it mostly happens that way. The hidden comments are generally the truly garbage/abusive/disgusting ones.

Putting author-disapproved comments at the bottom of the sort order facilitates a variant of forum sliding where the author can (via sock puppet/collaborators) post many author-approved comments to push the author-disapproved ones so far down that no one ever sees them.

Some of these could also be used to help hide abusive rewards in the comments, too.

0
0
0.000
avatar

No.

The last thing Steem needs are echo chambers. I'd prefer a different option again to what is proposed. "The author of this article has muted this commenter, which is why you may not see a reply to the comment from the author."

There are genuine cases were individuals mute people for petty things, when they are attempting to constructively critique or promote discourse with that writer.

Progress is only made when people build upon their differences and disagree, and a platform like Steem should be completely transparent, showing all, and every point of view in regards to a piece of content.

Posted using Partiko Android

0
0
0.000
avatar

Constructive critique towards a person that doesn't want your critique and has muted you would technically be called "Unwanted critique" ;)

0
0
0.000
avatar

NO
This solution is simply ugly. I prefer any of the propositions in the previous post.

0
0
0.000
avatar

No,

This is going to be used by plagiarists and can be used in spreading hate or horrible ideas. This would also work really well to push a certain message. Let's say Steemit for an example don't like everyone who is saying they don't like a fork so they mute all of these people so it looks like they have overwhelming support.

Not exactly top priority.

0
0
0.000
avatar

None of these are problems.
If steemit starts to mute all critiques then that will just blow back even more against them. It's a tool more so than a new standard.

0
0
0.000
avatar

No.
If you want something hidden, flag it.

0
0
0.000
avatar

Right, If you have enough Steem power that may work ...

0
0
0.000
avatar

Or enough courage.

0
0
0.000
avatar

Courage alone won't really help to hide it (but it may help to get hidden yourself in future). :)

0
0
0.000
avatar

Good friends wouldn't let you go down alone.
Rare though they may be, they do exist.

0
0
0.000
avatar

Sometimes, actually, good whale friends would be necessary ...

0
0
0.000
avatar

And the spammer won't downvote you if you mute his comments under the post?

0
0
0.000
avatar
(Edited)

At least the author wouldn't have to see the comment of the spammer anymore.

Furthermore, I think a committee of elected users with some delegated Steem power from Steemit, Inc. might help, which could decide (in case someone complains) if flags are justified or not, and if "yes" just counter them with upvotes.
In addition, accounts who repeatedly misuse flags in an abusive way (instead using them against spam, plagiarism etc.) could be flagged, as well, after a decision of that committee.

0
0
0.000
avatar

No matter how many times you try to step into censorship, the answer is a BIG NO!

0
0
0.000
avatar

I think you have many other much more important issues to fix.
Why you're wasting time with this?
Very unproductive.

0
0
0.000
avatar

No one is presently programming anything for it and even if they get the go ahead it's not very challenging and maybe even a community member submits a pull request for it. So it's not quite fair to consider it waisting time. They're practicing communication BEFORE doing things and i think that's very valuable thing that we should be celebrating.

0
0
0.000
avatar

You're right. It wasn't fair. I'll try again. Maybe their approach is not right. It's more debating than polling. You cant decide for changes on the opinion of ten or twenty people. Community is much bigger.
But I still think that there are many other things to do first.

0
0
0.000
avatar

No , unless we can opt-out of this.
I would like to see every comment on every post without having to click Reaveal all everytime !

0
0
0.000
avatar

NO

I think some legit users would block cheetah, whose algorithm often flags posts erroneously, especially contest posts.

This simplification would lead to even more dichotomous "user bad" "user good" perspective, which is the opposite of "proof of brain".

Posted using Partiko Android

0
0
0.000
avatar

Jesus... When will you start advertising to bring in NEW USERS AND PEOPLE TO BUY THE COIN?? Why are you sitting around thinking of useless features as we slowly slide under the top 100 marketcap?? There are pure shitcoins with better marketing tactics... You have NO marketing efforts to show. WTF!!!

0
0
0.000
avatar

Alternate Implementation...

...What do you think? Would this constitute an improvement over the existing UI?

Bleh! I say bleh. You can take this as a NO a Yes or a simple Whatevah. It seems to me you already have your ultimate decision 'made up'.

0
0
0.000
avatar

Abolsutely NOT! NO!

This is why!

You want to hide the truth. You are AFRAID. That is why you want this. You need to control your bullshit narrative.

Go fuck yourself and the horse you road in on!

0
0
0.000
avatar

NO

I still don't like people being able to censor the voices they disagree with

0
0
0.000
avatar

NO

I believe this is the same wrong approach, just realised in more hidden way.

Make downoting usable and create this with downvoted content. So, that if a persons wants to see all spam, it can be unhidden with one click.

0
0
0.000
avatar

YES

I think this feature is crucial for Steem to grow and attract more serious content creators.

It would be nice in addition to the feature already mentioned if users could access a settings option that would allow them to turn author moderations off. Aka shows a "vanilla version" of Steem. Users could also be informed of the existence of this setting when they first click the new "reveal all button".

The point here is that authors should be able to control the default view that an outsider without a Steem account will get when they visit their blog. However, Steemians who are logged in and want to judge content and see what they want should be able to do so without needing to look for a "reveal comments" option all the time.

Finally. With hardfork 21 hopefully making downvotes more popular and accepted, it would be great if steemit also had a "filter comments by controversy" option similar to reddit. This would make it even easier for users to look also for comments that may have been fairly or unfairly downvoted/muted.

0
0
0.000
avatar
(Edited)

Concerning the downvotes I think a committee of elected users with some delegated Steem power from Steemit, Inc. might help, which could decide (in case someone complains) if flags are justified or not, and if "yes" just counter them with upvotes.
In addition, accounts who repeatedly misuse flags in an abusive way (instead using them against spam, plagiarism etc.) could be flagged, as well, after a decision of that committee.

0
0
0.000
avatar

No. Who I choose to mute should be known only to me. That's my business. Nobody else's.

0
0
0.000
avatar

NO

It be nice as the author to be able to pin or feature a comment on my own blog post. Along with some kind of comment flood spam protection. Outside of that it’s quite a slippery slope to be going down.

0
0
0.000
avatar

NO.

For this, we can vote to prevent comments from being visible.

0
0
0.000
avatar

No.

Muting / posting rights and restrictions should be handled by the "communities" development. Each community can then set its own rules for how to handle spam / comment advertising / abuse etc.

0
0
0.000
avatar

Alternate Implementation

Its perfect, so yes

0
0
0.000
avatar

NO, they considered that each of the same way we want to read a post of someone, we would like to know the opinions of other users, and even more if they are spam or plagiarism, without having to be pressing a button to see the hidden comments, in addition to being in a decentralized network we are subject to others to comment what they want, it is supposed that here the freedom of expression is not restricted.

0
0
0.000