30
votes
Suggestion: a way to identify extra-good topics
We have the "Exemplary" label for comments, which identifies comments as particularly good, and even boosts their ranking within threads.
Now that we've had this for a while, I keep finding myself want to do the same for topics. I'll read an article and want to give it an extra boost because it's better than average.
I'm ready for an equivalent to the "Exemplary" label for topics.
This has been discussed a fair amount in the distant past and AFAIK is definitely still planned to be implemented at some point. The "how best to implement it" is the only real question though, since having it be the exact equivalent to the exemplary comment label, where a new one is given to every user every 8hrs and 1 exemplary label being applied gives all the benefits to the post, isn't a particularly good idea when Tildes is this small and has this few topics being submitted.
I have seen talk of a "super upvote", but I remember that was to be restricted to only certain users. I'm talking about something available to all users, just like comment labels.
And, while something like this might be planned for implementation at some point, I'm suggesting that point be sooner rather than later.
Why shouldn't it operate similarly to the "Exemplary" comment label? That label gives the exemplary comment a coloured highlight and a slight boost in ranking. That's all we need for this "super vote" on topics.
I'm not sure why you talk about the quantity and frequency of the "Expemplary" label. Despite the fact there are hundreds of active Tilders, and that we each have 3 potential "Exemplary" labels to apply every day, I don't see the comments sections flooded with exemplary-labelled comments. Between us all, we have enough "Exemplary" labels to cover more than 50% of all comments - but we don't. We exercise restraint and judgement.
That was only one of many suggestions for it and discussed possibilities, as was giving them to everyone but basing their refresh rate and weight based on trust. And it's absolutely not a coincidence that those functions (refresh rate and weight) are already built into the exemplary comment label, since it is basically a rudimentary implementation of that very "super vote" idea.
And if we all had the same amount available for topics, we would have enough to cover 100% of the topics 10x over... because of the "quantity" of topics compared to comments and the "frequency" with which we would be handing out new exemplary labels to apply to the topics if it was the same as for exemplary comments. I mentioned them because that's definitely something to consider first before implementing it.
I also would like to see an exemplary system for topics sooner rather than much later. However as I said, I think the idea needs a bit more consideration before someone puts any effort into implementing it, and I also recognize that there are much more important features, bug fixes and other major quality of life issues to focus on first. Exemplary topic labels (or whatever they get called, which is another thing that needs consideration) would be nice, but isn't critical at all, or a major QoL issue like almost all the other "in progress" issues are.
Exactly. Like all good development, current features are built with an eye to future features, and future features can be built using previous features' code.
Actually, we probably already have enough "exemplary" labels to cover just about all comments, given that lurkers probably out-number active commenters.
Anyway, nowhere in my original suggestion did I say that we should have 3 "super votes" for topics per day. I merely said that this feature should be implemented. I'd be happy with only 1 "super vote" for topics per day. I think this is a very minor detail.
"Exemplary" labels on comments weren't critical or a majority quality-of-life issue, either, but they still got implemented.
Labels on comments absolutely were a major QoL improvement since they have allowed for crowd-sourced moderation. And the reason an exemplary one was included despite not necessarily being the most important QoL improvement is because the system was already in place anyways, since comment tags were on the site since the very beginning. There is no analogous comment labels system for topics yet, which means it would need to be created from scratch. And no, topic tags don't count since that system functions entirely differently than comment labels.
Not to mention the fact that any effects an exemplary label would have on topics need to be considered first, and once decided on would need to be factored in to every topic sorting option, making it way more work to implement. It's not as trivial as you make it sound to implement nor are the issues that need to be sorted out first before it can just "minor details".
On one hand, I like the idea, and I see the appeal. On the other, this essentially creates a second upvote system. How do you discover these
doubleplusextra-good topics? Do you create a new sorting "Most extra-good"? Wouldn't it just be a copy of the usual "Most votes"? Surely, the topics that get the most votes would also get the mostTildes Goldextra-votes. I am honestly against this implementation.I think a more interesting mechanism would be a "Most bookmarked" sorting. Because often, popular (most voted) topics and topics that have a lasting effect (most bookmarked) are not the same. A topic may receive a lot of votes because it's breaking news, or because a lot of people think it's important, or because @Deimos posted an update, or because it's about something that a lot of people find good. But the posts that people bookmark — those are the special kind, the ones people chose to keep with themselves for a longer time. They are probably the most useful, practical, and informative.
Edit: Thanks to @cfabbro, this is now an issue on the tracker.
I think "most bookmarked" could be interesting, but it's definitely aimed at a different purpose than something like an "Exemplary" vote for topics would be.
As a demonstration example, this recent topic by @wanda-seldon is excellent. I'd absolutely give that a "stronger" vote if I could, but I'm not going to bookmark it because I have no intention to come back to it again in the future. I bookmark topics/comments that I might want to reference again later, or that I'll probably want to link to other people eventually, but that's quite different than recognizing something as especially high-quality.
I think bookmarking implies something more like "long-term usefulness", which isn't the same as a measure of quality (even though there will be overlap between the two).
Incidentally, I bookmarked this post. Both because I don't want to lose it and because I plan to return to it later.
It would mirror the "Exemplary" label on comments. These extra-good topics would be given a slightly heavier weight in vote-based sorting, and would therefore sit higher up in lists, getting more visibility.
That's an interesting twist.
However, not everyone uses bookmarks. I certainly don't! Therefore, while some people may think a post is extra-good, we'll never know because they don't bookmark it.
We also have to consider that bookmarking a topic is mostly a way of people saying they want to read it later, not necessarily that they think it's good. When they finally get around to reading a bookmarked item, they may decide it's bad!
Again, how is this different from just having a lot of upvotes? If the topic is unpopular but the mods/admins decide that it needs more attention, then a better mechanism would be an "admin bump", like the one they have on HackerNews. Admins there also usually write why they found the topic interesting, and the OP gets a notification.
This could be rephrased as
The fact that some people don't use some features is not really related to the discussion, is it? Please feel free to correct me here.
I would argue that if the topic was bookmarked in the first place then it is at least a little bit interesting and useful. “Promising”, if you will.
I love this discussion by the way.
LOL, oh neat. I hadn't seen your comment here when I made that issue and thought I was the first to come up it... but it turns out you beat me to it by over an hour! Great minds think alike, I guess. ;)
I added this suggestion to Gitlab BTW @Algernon_Asimov:
https://gitlab.com/tildes/tildes/issues/314
Thank you. Much appreciated.
Because personally I (and as understand it, many others) mostly use activity sort, even the actual upvotes don't really do anything. There's the number, but I don't really notice looking at it much.
What about actually visualizing the amount of votes? Something like adding a sidebar that becomes brighter the more votes topic has? That way good posts would be instantly visible.