Suggestions on labels
Rationale: labels are a valuable way to receive and give feedback, so it would be useful to have more labels-related tools.
This topic deals with labels received by an ordinary user or given by an ordinary user from that user's point of view (as opposed to non-logged-in lurkers, other ordinary users, and users with elevated privileges).
While labels presently only apply to comments, these suggestions would apply to topic labels when they are implemented, and to other labellable content types should any appear.
The “Gilded” page—Issue 423
Suggestion 1. Users can filter their user pages for content labelled Exemplary.
Unlike all other suggestions, this also applies to users viewing other users' pages, and possibly even to lurkers viewing user pages.
I also suggest that users have an option to automatically expand the Exemplary messages when they see their own Exemplary content.
Other labels given TO the user
The common point is that it would help if users observe the feedback given to them by others via labels. In addition, this would prevent label misuse and abuse.
Suggestion 2. Users have an option to observe labels given to their own content along with the label counts.
Suggestion 2a. If comment vote counts remain generally hidden, users should still be able to see the vote counts for their own comments.
Suggestion 3. Users can filter their user pages for content labelled Malice (but, of course, they should not be able to see Malice messages).
Suggestion 4. Users can filter their user pages for content with any label (maybe with further options like All labels vs “Non-major” ones).
Edit: Suggestions 2, 3, and 4 might go with time lags. Namely, labels given to own content are only visible for content older than X minutes (X can be even 1440 or more) and to users with accounts older than Y days.
Labels given BY the user
Suggestion 5. Users have an option to automatically expand the label pane for the content they have already labelled.
Suggestion 6. Users can easily overview the content they labelled Exemplary. (This is basically the “Gilded” page in the other direction.) In addition, users can see the messages they provided when giving Exemplary labels.
Suggestion 7. Users can easily overview the content they labelled Malice. In addition, users can see the messages they provided when giving Malice labels.
P.S. These suggestions deal with the current labels, but they can be extended to future labels, e.g., group-specific ones.
Thanks @MetArtScroll, some interesting thoughts and suggestions in here. I expect that we'll be talking more about labels pretty soon, because I have a change that's going to make them more important that should be getting deployed soon
(hopefully in a few hours after I finish a few last things)(Edit: got too late, I'll do it first thing tomorrow).Overall, I think we need to be very careful with showing people "negative" labels on their own comments. They can very easily create conflict similar to downvotes, and as some other people mentioned, we previously saw that when they were initially visible. I do agree that they have the potential to be useful to give people feedback, but I think it might be better overall to try to keep that a little more vague and less immediate.
For example, something I've thought about is if a new user joins and some of their first comments get labeled as Noise, maybe they should get an automatic notification/message telling them that some of their comments were marked as not contributing to discussions, and pointing them to some guidelines about commenting (which don't exist yet). I think approaches like that could help give people that kind of feedback without it feeling as confrontational (which often gets people upset and makes it worse).
Tease! ;) /noise
OMG, please be /. style voting!!! Please be /. style voting!!!!! Dang. :(
Still "interesting" changes though. And I am glad the default is going back to all-time, since currently it's a shame that once a topic reaches 3 days old it basically completely dies. I still think we might need a "hide topic" feature before all-time is totally viable though, even with this new "interesting" change.
Your suggestions regarding making comment labels visible to the user whose comments they have been applied to would very likely end up causing the exact same sort of issues that lead to the original comment tag system being disabled and eventually revamped into what it is now.
Suggestions 6 and 7 are potentially good ideas though... so I have added them to Gitlab:
https://gitlab.com/tildes/tildes/issues/495
I don't know how I feel about it. While I agree with you, I think either all or none of the suggestions should be added to gitlab. I know this comment pointless since Deimos reads everything in ~tildes anyway, especially since it's linked to the gitlab issue, but I still feel we should not choose which parts of suggestions should be added to gitlab.
People are more than welcome to add their own suggestions themselves to Gitlab... but I at least try to use my best judgement on what I think is actually likely to be well received there (and not immediately closed by Deimos) and only add those, since I don't want to clutter Gitlab up with every idea, no matter how non-viable they are.
If @Deimos chimes in and asks me to add the ones I skipped (or everything from now on), I will though.
Actually, the very reason why I numbered the suggestions is so that they could be considered separately.
I hope it is clear that I do not suggest that a user observes the labels given by other users to other users. This, when everyone could observe everything, indeed caused those issues with the original comment tag system.
(Unless those issues included users complaining for unjustified labels, in which case I would prefer label abuse not to be hidden from the users abused.)
I understand your suggestion, but IMO one of the major issues with the old system was caused by users being able to see tags that had been applied to their own comments. A lot of threads got completely derailed by people getting angry at certain tags being applied to their comments (even when justified), and accusing others who had disagreed with them, or they were arguing with, of being the ones to apply them, which in several cases caused things to spiral completely out of control.
Abuse can still be detected via the effect they have on the sort... and I disagree, since the person who has labels applied to their comments is generally not the most objective judge of that, as even when there is no abuse and the labels were correctly applied, they can often still feel as if they were unfairly applied.
So would making labels visible to the labelled users after certain time (between 10 minutes and one day) solve this problem?
Users will still see the valuable feedback, they will be able to observe and report true label abuse, but they will not be tempted to derail discussions.
Edit:
I definitely agree, but hopefully we are on Tildes and not on certain other social media. The whole collapse of the original comment tag system occurred when Tildes was about one month old.
Considering all it takes to revive a topic is a new comment due to the Activity sort, and even after 1 day the same problem of escalating arguments and accusations still exists, I don't think that delay would really make that much of a difference.
And IMO the only label that can really be abused right now is the noise label, which is easily detectable by others, so if anyone sees a noise collapsed comment that they feel has been unfairly labeled as such, they can (and should) just report it to Deimos. But honestly, I have yet to see more than a few of those, but even they were borderline cases.
edit:
That's true, and I could definitely be wrong here. It's possible that making the current labels visible to the user who had them applied to their comments won't lead to the same problems as before... but I am still incredibly wary of it.
On my side I can be wrong in overestimating the userbase (especially the future userbase).
I am adding the time lag idea (both for the content, and, more importantly, for the user account age) to the topic post.
Yeah, and perhaps the issue with this idea for me stems from the fact I am trying to think long-term, with reddit's downfall as my major frame of reference, so I tend to think of worst case scenarios and have very little faith in social media users. i.e. "The Barbarians are at the gates!" And I still haven't decided if that way of thinking is good or bad for Tildes yet. :P
I can see both sides of it. But speaking personally, I've noticed that when I see downvotes/negative reactions on my content it tends to put me in a worse mood which leeches into more negative comments. It's not a conscious decision but does seem to encourage the spread of negativity. That's part of the reason I like the voting/labeling system of Tildes, and how it's hidden from the user.
I suspect this might have some value. The issue is the triggering of bandwagon effects - but we're not looking for specifics here, are we? General is enough. Rather than seeing the exact labels on your comments in some specified time period, you'd acquire some basic feedback about your behavior's reception over a longer time frame.
Distancing it from the numbers and from immediate feedback (say, 30 days lag or something like that) prevents the flare ups in the individual threads. The user won't be able to point to one thread or one discussion or one interaction that caused their feedback to shift in some way or another, instead they'd just get a time-delayed average of some kind. They have no target for the reactions where they got such negative or positive feedback. Presumably the admins would still have some access to this information (such as which interactions provoked which feedback in which groups) until it's wiped at the 30 day mark. That info could be factored into moderation tools for everyone to use.
This very much depends on the user's attitude. While seeing negative labels on their content, some would reflect on why they received such labels and adjust better, whereas others would instead blame the labellers. I somehow think that the former ones prevail on Tildes, but see the comment thread here between cfabbro and me.
As for new users, they can be classified into the same two types. For those who take the feedback into account, this suggestion (#2 as well as #3 and #4) will be extremely valuable for the reasons you described in your comment. For those who ignore the feedback and blame others, this will not change their bad behaviour and could lead to thread derailment as described in that comment thread. Very unfortunately, bad actors are usually new users as they typically get banned quite quickly.