63

Continuing in my series of questions to you: Last week, I heard some great origin stories. This week, I’d like to refocus a bit on what could be better (in your view). I’m very curious about whether we’ll find near universal agreement or if we’ll discover that different communities have very different needs or ideas of how things “should” work. So my question for you this week is:

  • If you had a magic wand that you could wave that would let you fix ONE thing that would help your community (not the network, not the company, but the specific site community with which you most closely identify) overnight, what would you change (and why?), and what would the result look like - how would we know it was different?

This can be something technical (a site “pet peeve” - props to Mithical for suggesting as a topic), or could be something in policy, or could be the size of your mod team, or permissions of your users, or something about the environment in which you operate, or none of the above!

I hope it goes without saying that I don’t have the ability to magically “produce” technical change. While I might be able to advocate for technical solutions, I’m not committing here and now that we will build anything suggested here. I’m just looking for ideas - and I would ask that you lean toward social or community based ideas rather than technical fixes, if possible.

  • For bonus magic internet points, feel free to hazard a guess about whether your beliefs would be widely shared or whether you might be a bit of an island on this one.

Community "ONE thing"
Code Golf and Coding Challenges Stack Exchange Making all users use the Sandbox before posting a challenge on Code Golf and Coding Challenges
All communities Changing age requirement needed to be a moderator
All communities Stop people from enforcing the rules without questioning and understanding their purpose.
Stack Overflow
All communities Making vote counts visible regardless of reputation
Code Golf and Coding Challenges Stack Exchange Built-in leaderboards on Code Golf
All communities Fix a 10 year old security vulnerability
All communities Time Travel! (Or at least a cautionary tale of community dynamics)
Stack Overflow Discourage overly situational debugging questions, encourage more generally useful questions.
All communities Convert Answer to Comment
Puzzling Better spoilers
Allow more syntax highlighting
Interpersonal Skills
Software Engineering Encourage more community members on Software Engineering in curating minor issues in questions in favor of closing them
All communities Could we have Chat (and meta) on the sidebar?
Tag Hierarchies/Dependencies
Decrease the limit of simultaneous hot questions from the same site
Science Fiction & Fantasy Custom question-flow for asking story ID requests
English Language Learners Easy way to find credible definitions for words and idioms without leaving the site
Bring chat into this century
All communities A community advocate.
TeX - LaTeX Restore previous "community promotion" conventions
Improve or officially support sandboxing
Meta Stack Exchange Fewer off-topic questions from new users.
All communities More capability to customise the Help Center
Project Management It looks like you're asking a question about Scrum. Have you considered asking it on Project Management instead?
Provide more details about spoiler content without revealing
All communities Site-specific new user onboarding
Shared Drive for valuable Google spreadsheets
Fix a major past mistake which has lasting effect to this day
Force new users to take a quiz
Stack Overflow Duplicate closure should be a more rewarding experience
Role-playing Games Make it possible for RPG system tags to always appear first, and display them in a special color
Multimedia Help Center
Ask Ubuntu Archive feature
Graphic Design Modify the copyright release clause a bit for this site so that a stronger community could be built, a mechanism for sharing SVG files
English Language Learners Change the way questions are tagged
Tag Reform!
Better accessibility.
Allow non-bumping minor edits, but review them on /review
Invite low rep user to a chat
Retain existing users
Biology Ability to search Species ID posts using pictures
Ask Ubuntu Better integration with other forms of Ubuntu documentation
Code Golf and Coding Challenges Stack Exchange Automatic accepting
Ask Ubuntu Obsolete Answer Flag
15
  • 6
    Since this is seeking input from users from all sites, it should probably be [featured].
    – 41686d6564
    Jul 28 at 19:28
  • 9
    Would be really nice if people could (and would) read (and follow) the rules of the sites they posted on. Some sites have closure rates of 20% and more And I know that for Code Review one close reason takes up 60% of these closures.... Jul 28 at 19:29
  • 1
    @Vogel612'sShadow- I should have said this, but it would be great if you could put that in an answer so that people can comment on it.
    – Philippe StaffMod
    Jul 28 at 19:29
  • 13
    I may post an answer, I'm still looking over my wishlist :). But, I just wanted to say thank you for asking this specific question. There's been a lot of "estrangement" from the communities on the network over the past few years, and this is a great first step towards fixing that Jul 28 at 19:30
  • 7
    @cairdcoinheringaahing- thanks, that's really valuable feedback to have. :-)
    – Philippe StaffMod
    Jul 28 at 19:31
  • 1
    @Philippe it's not really intended as an answer, it's more of a venting thing. While the point is one that could be made in a constructive manner, I don't have the self-control necessary for that and wouldn't want to get the discussion (or more accurately this instance of the discussion) off to a terrible start. Jul 28 at 19:42
  • 4
    @cairdcoinheringaahing if all else fails, post it all 😁
    – Journeyman Geek Mod
    Jul 29 at 5:20
  • 4
    For me to reconsider participating on these sites ever again, this would have to be fixed - because it is an unaddressed injustice and a stain on the company's conscience, for one, but also because it getting fixed would indicate that some dysfunctional patterns within the company (which could become detrimental to volunteers again) might actually have changed
    – Pekka
    Jul 31 at 15:03
  • 3
    @Pekka you may be interested in this answer here
    – gnat
    Aug 1 at 9:10
  • 2
    Could we get a list of the answer headers in the question so one does not need to scroll so much?
    – joojaa
    Aug 3 at 16:27
  • @joojaa was it done before? Won't do it without OP telling it is OK. Aug 4 at 9:10
  • 4
    @ShadowWizardWearingMaskV2 no objection.
    – Philippe StaffMod
    Aug 4 at 11:58
  • Does this answer your question? Let's Plan the Second Iteration of the Stack Exchange Quality Project! Aug 9 at 7:13
  • Post seeking input from the Ask Ubuntu community: meta.askubuntu.com/q/19683 Aug 22 at 18:53
  • @joojaa Done! Please do any necessary edits :) Sep 12 at 12:02

46 Answers 46

56

Making vote counts visible regardless of reputation

According to this, the main reason viewing vote count is restricted by rep is because it's costly on the server side. It'd probably take a bit of work, but I'm sure it'd be possible to optimize things a bit so that anyone who has an account can view vote counts.

This isn't quite as important on main sites, but on meta and per-site metas it's often important to be able to see the vote breakdown. On CGCC, for example, a lot of our rules are based on voting on meta questions and answers, so not being able to see vote breakdowns makes it difficult for a new user to determine if something meets the requirements to be a valid answer or not.

Even if an answer has +10 score, without being able to see the vote breakdown you have no way to know if 10 people found an answer to be incredibly useful, or 1000 found it to be helpful and 990 did not.

This is definitely more a smaller issue, but it'd be nice if newer users (or established users on other sites) wouldn't have to resort to workarounds like userscripts to be able to see vote breakdowns.

8
  • 1
    I asked for this in 2016 and was told performance was still an issue then.
    – xnor
    Jul 28 at 21:47
  • 3
    @xnor I'm not asking specifically for the counts to be shown by default, just for the rep requirement to be lowered or removed, which I think would be much better performance-wise. Jul 28 at 21:51
  • 20
    If vote counts should be restricted, I'd be at least happy if there was an association bonus-like way to unlock them. It's quite annoying being very used to a feature on one stack and not being allowed to use it on another. Yes, rep does mean there is some measure of trust to a user, but I don't see why you should be trusted to see the vote counts separately on each individual stack.
    – VLAZ
    Jul 29 at 6:07
  • 2
    I agree; the part that makes this even sillier is that the information is publicly accessible through SEDE and the API anyway. There's a userscript that gives you the functionality on any Stack site, it just feels weird for it to require a seemingly arbitrary 1,000 rep.
    – zcoop98
    Jul 29 at 15:01
  • SEDE and the API are all very well for technical users -- what about the non-technical users? If you're trying to fix a problem on meta sites (which I agree exists) why not ask for e.g. a Poll mechanism -- it's long been the fact that shoehorning what is effectively 'polling' into a Q&A format does not work well. I'd make an answer out of this but it isn't the biggest issue for 'my' site. Jul 30 at 15:25
  • I see one issue with this, you need some site specific experience to understand the logic behind vote counts. It can be overwhelming enough to get to 1k, I don't know how far seeing vote counts might be TMI before 1k. An argument can also be made that it motives users between the 500 rep review queue privilege and the 2k edit privilege.
    – bad_coder
    Jul 31 at 0:54
  • 4
    And if performance is still an issue, can we at least allow users to see vote counts on their own posts?
    – David K
    Aug 2 at 15:59
52

Hi, we're special!

This is what the front page of IPS looks like to someone that doesn't have a profile there. This is what it looks like after you've made a profile and log in. And what you end up seeing after finding us on HNQ: This or this depending on whether you're logged in. The 'Questions' page too, where people are likely to end up browsing other questions: it looks like this.

Let me add some more screenshots to compare those with: A forum, one of the Facebook groups I'm in, and a not so-random subreddit.

Spot the differences. Both that Facebook group and that subreddit have a box in their right sidebar with rules and information about the thing you're looking at. All IPS gets is a big black banner with some text. Our meta (does a newbie know what meta is?), chat, our help center... All the things that contain information about what makes our site just a little bit different from others: they're all hidden away behind this button.

As for the forum post I threw in there: "Closed" behind a question doesn't immediately make it clear that this is because the question is unsuitable for the site. On that forum I screenshotted, for example, 'Closed' means as much as 'I'm not looking for reactions right now'. If your post was problematic, it would've been deleted by moderators/admins. But for a new user to SE, there is no place that explains what 'closed' means (until they open the question), they just see a list of titles and assume they can ask about similar things.

Do you think a total newbie would want to know who the newest valued associates are, or what the site statistics are? That they're immediately looking for tags to watch, or to create a free team? Or would it perhaps be better to give each site a part of that right sidebar, where they can explain "Hey, if you're thinking of asking a question here, look at this and this. If you want to answer, here's your relevant info! Also, don't use our comments to argue or write answers in". I want that box, similar to what my Facebooks groups and some not-so-random subreddits have. I want to have the ability to shove it in people's faces: Hi, we're special, and this is what makes us different from other sites you may be used to!.

12
  • 11
    A million times this, yes! Jul 29 at 8:31
  • 3
    Code golf also had this problem, but recently, we got a nice box in the top right corner. Maybe you can get one too.
    – Adám
    Jul 29 at 11:26
  • 3
    @Adám The one on Code Golf is a nice example of what that could look like, but I was really thinking something that can be edited by the moderators on the site itself though, something a bit more flexible than having to depend on SE devs to make/update such a box :)
    – Tinkeringbell Mod
    Jul 29 at 11:36
  • Law, MiYodeya, and Christianity all have one, that I know of. Might be good to check with the mods of those sites how they got it? Probably a status-review on their meta sites?
    – LShaver
    Jul 29 at 16:32
  • Also MSE has an FAQ... wonder if it would be possible to get an FAQ to your meta questions, where the title is a rule, and the link gives context?
    – LShaver
    Jul 29 at 16:33
  • 5
    @LShaver IPS already has a FAQ. That post is buried on our meta, and the only way to get it linked on main is by keeping it featured, which only shows the title and no other context. A box would allow for a lot more context, really think something like that Facebook screenshot. I don't mind if it collapses and someone can press 'show more'. But it needs to be front and center, and something moderators can do without having to status-review anything and wait for SE (the current sites with disclaimers had to ask SE, some before the current status-review process was in place)
    – Tinkeringbell Mod
    Jul 29 at 16:44
  • 4
    @Tinkeringbell I love this. We have 176 unique communities on this platform. Let's celebrate and highlight that more. :)
    – Rosie StaffMod
    Jul 29 at 19:07
  • 2
    Similar to Rosie, I really like this too. Maybe we can get it done - I shall poke around.
    – Cesar M StaffMod
    Jul 30 at 0:03
  • @CesarM I'm not sure what you're planning to do after poking around but... This answer really is just a write-up of a 'shower-thought', not a fully thought-out feature request, I never thought about whether this could have any downsides, for example. As an answer to this question, it's also a bit more hidden than a proper feature request would ever have been, and as such it may very well be that people who absolutely don't want this or see serious downsides to it have never seen it. If there are such people, they aren't going to love SE for just shoving a box like this in their faces, I guess.
    – Tinkeringbell Mod
    Jul 30 at 8:19
  • 5
    @Tinkeringbell oh! Sorry! By Poking around I meant socializing the idea internally and seeing what people think. I like it, but it's not something we have planned or even discussed, so it starts at discussing it with different folks and seeing how it goes. If there's interest/anything actually concrete I'll update and it would go through our normal feedback collection/development process: as in, we would give a chance for people to give thoughts before launching it. For now, just talking!
    – Cesar M StaffMod
    Jul 30 at 16:14
  • @CesarM Pfew! Glad to hear that :)
    – Tinkeringbell Mod
    Jul 30 at 16:15
  • 2
    It is great. It is worth noting that for quite a lot of years all sites had this box.
    – Nicolas Chabanovsky StaffMod
    Aug 4 at 8:25
48

Fix a 10 year old security vulnerability

You can add images from any source in a post (or in chat). This allows things like:

  • Determining the IP addresses of specific chat users or (with less accuracy) highly active users on main
    • This allows tracking which questions they view, which chat transcripts they open, and more
  • Gaming the Announcer-class badges
  • Making it appear to network administrators that people on school or work computers are navigating to porn sites
  • Phishing passwords from people using old browsers
  • Silently taking advantage of certain CORS vulnerabilities in other sites
  • Freezing users' browsers with specially crafted SVGs
  • Using advertising links to make money

Additionally, after these "images" (usually not actually images) are removed from an offending post, one can link to the revision page, which appears trustworthy but still contains the payload.

7
  • 6
    How would this help fix one specific community (like Philippe asks for in the question)? Jul 28 at 20:24
  • 9
    @EkadhSingh It's admittedly not as focused on one specific community, but it's definitely something that many members of the community of the site that I spend most of my time on have agreed is concerning. From what I can tell this falls under the "pet peeve" section of the question. Jul 28 at 20:30
  • 1
    Is this the issue that killed Twitter oneboxing in chat? Jul 29 at 5:50
  • Its a different security issue IIRC :D
    – Journeyman Geek Mod
    Jul 29 at 9:54
  • I would be unhappy if we couldn't use SVG images. Maybe they could be supported via a variant of Stack Snippets.
    – PM 2Ring
    Jul 30 at 20:14
  • @PM2Ring I definitely don't think disabling SVGs is the right solution, but requiring hosting through a trusted source is a good start, and I'd imagine some sort of restriction for ridiculously large SVGs (if things like imgur don't already have any) could probably be implemented on top of that without too many issues. Jul 30 at 20:41
  • Yes, a trusted host would be ideal. AFAIK, Imgur still doesn't support SVG. meta.stackexchange.com/q/92568/334566 I mostly link to Wikimedia SVGs, but I occasionally link to my own creations, which I host as GitHub Gists.
    – PM 2Ring
    Jul 30 at 20:54
45

Duplicate closure should be a more rewarding experience

The main community I use is Stack Overflow. We have more than a decade worth of questions - chances are that most simple queries are answered. Yet we get them again and again. We are drowning in duplicates. Some show up multiple times a day.

There are multiple issues with closing duplicates that make it a chore and not even viewed in good light.

Closure seen as bad

Duplicates are closed and many users seem to think that closure is somehow bad. There are two misconceptions here, in fact:

  • closure is not bad. It is supposed to be a temporary state until the question is improved. It is quite neutral, not any sort of judgment on the user as a person or driven by malice.
  • Duplicates are a type of answer. This discrepancy is jarring because dupe closures are still closures. Yet, we close as dupes not as a temporary state1 but to immediately give access to answers. In some cases, the answers have been curated and improved for years.

These misconceptions put people off.

In the ideal magical world of my fantasy, duplicate closure would not be presented as "closure" but re-framed as a way of answering.

1 Unless the duplicate was wrong. But we do not aim to close against the wrong target.


Question askers do not cooperate

The misconceptions do agitate a lot of them. This leads to some fighting the duplicate closures. Here are some of the common arguments:

  • The title does not match. But we do not close based on title but answers.
  • The top answer does not help. Again, we close because of all answers, not just the top one.
  • The duplicate does not work for their code. Very often this is because their code uses different names - it is not foo it is bar. Or values - not 42 it is 215. And that is it - everything is the same.

We close as duplicates to give access to ready-made answers instead of having to wait. Having to then further spend time convincing people that yes, this is the answer but you have to maybe modify slightly or maybe even read further down is frankly exhausting.

Not all question askers disagree or fight the dupe closures. However, I seriously sometimes avoid closure and just leave it because I do not feel I have the mental fortitude to start explaining yet again why a would answer something.

In my imagination I picture question askers being happy they are linked to a high-quality existing resource. Because they recognize that it does solve their problem.


Question answerers do not cooperate

Another reason to close as duplicates is to consolidate all knowledge in one place. This only works if all the knowledge is in one place.

Very often, however, we get answers to duplicate questions. Sometimes many. A common and simple enough question can rack up 5+ answers by the time it is closed.

Sometimes, duplicates are not even closed. It is extremely frustrating to start searching for a duplicate and find 3-4 different questions that are all fundamentally the same but you cannot really choose any one of them because all are open and all have different answers.

Gold badge closure power helps a bit but not everybody has that.

In my dreams, people are motivated to find or vote for a duplicate instead of answering.


Searching for the correct dupe target is not very pleasant

I cannot entirely blame people for answering duplicates. Having to look for existing answers is a literal thankless task. The vast majority of the time, at least.

Some duplicates are easier to find. Others take more time. Regardless of the type, I personally start to get exhausted after having to look for a dozen duplicates in a row. It feels like I am just a personalized search assistant. I imagine others might have even less of a tolerance to that.

A simple answer is very tempting. It might be easier to write one than find the correct duplicate target. Also, answers get upvotes, comments, and reputation in general. Finding a duplicate gives you either nothing or frustration.

With some imaginary powder, I would make it so somehow proposing duplicates has some sort of reward. Some incentive to do it rather than just because consolidated knowledge is how what we want of Stack Overflow. It will help mitigate some of the dissatisfaction related to this.


Amount of incoming questions is high

This might be a personal thing but what I also find exhausting is just how fast questions are coming in. At more busy hours, there might be 10 questions a minute. At less busy hours maybe 1 question a minute. Normally, it is somewhere in between.

In the time I spend evaluating few candidate questions to see which most closely matches as a duplicate target... there may be another 30-50 new questions added. A lot of them asking things that were already asked years ago.

That is only about the tags I monitor. At least in that corner of Stack Overflow, it seems impossible to keep up. We need to have more people curating given the number of questions we get.

Hopefully, all these magical transformations will make it easier for users to find an answer to their questions. Which will reduce the tide of new duplicates that come in. At least a bit.


With a magic wand, I would like to make the duplicate closure more pleasant for everybody involved. Close voters, answerers, as well as question authors. I just hope there is magic strong enough to do that.

4
  • 6
    This is an excellent answer (+1). If I come across any magic which is strong enough to do any part of what you're requesting, I'll certainly let you know :-) . Jul 31 at 18:27
  • 3
    I really like the "duplicates are not a closure" idea. Perhaps rather than just closing the question and saying "go over there" it could result in the answers appearing under the question that has been marked as a duplicate. Not quite a merge, but not quite alone either. It would mean that the person asking the question gets all the help right away instead of that jarring feeling you just told them to go somewhere else. Of course leaving the question locked to answers is good, and the answer box should be replaced with a "go to the original to post an answer" info box.
    – Mokubai
    Aug 2 at 10:55
  • I would add that for old questions finding a dupe in most cases doesn't even give you an accepted flag, because the queue is drowning in them and they're mostly never seen by other people. I've read multiple stats posts, but the flagging stats for my (and OP's) use case is the opposite.
    – EvgenKo423
    Aug 4 at 19:36
  • @EvgenKo423 yes, trying to CV an old post is pretty hard, I agree. It helps if you get a gold badge hammer :) That finally allows you to do something about old questions. But... also means that you have to deal with the old questions which weren't dupe closed and most likely spawned multiple copies :( Just trying to research all the potential dupes is a chore. Sometimes it's unsolvable because there are two or even more competing questions all with good answers. Nothing you can easily pick a clear winner of.
    – VLAZ
    Aug 4 at 19:41
38

Improve or officially support sandboxing

Code Golf is probably the site with the biggest need for this feature, though other sites like Worldbuilding have sandboxes as well.

Currently, sandboxes are just meta questions. Answers are drafts, and comments and votes are used for feedback. CGCC's sandbox currently has over 3000 answers, and the limitations of this system have become apparent:

  • It's difficult to review drafts, and often unrewarding
  • It's difficult to get people to notice sandbox posts, so they need to be regularly advertised in chat
  • This means that new users' posts can sometimes slip past and receive no feedback, even when they do everything right and use the sandbox as requested
  • It's difficult to get new users to notice or use the sandbox, when they're the ones who it would likely be most helpful for

Despite almost every new user being requested to use the sandbox in the tour, welcome modal, ask question guidance, and most reviewers' welcome comments, we still get plenty of questions which are either off-topic or lacking the necessary clarity or objectivity that the site requires.

I don't have any specific solution in mind, because I don't know how much work these would take or how much work SE would be capable of putting in. But, ranging from the simplest to the most effective, here are some ideas:

  • Add a link next to the "ask question" button, taking you to the sandbox
  • Add a feature where a meta question can be marked as an official sandbox (maybe a mod-only tag), which could add various things:
    • Disabling the "there are already more than 30 answers" pop-up
    • Sorting by active by default, regardless of the user's sort order on the rest of meta
    • Linking to the sandbox (in a very obvious way) in various highly visible places, such as the sidebar and review tab of the ask question page
    • A way to show the most recent few drafts somewhere visible, like the sidebar, to encourage reviewing
    • A way to quickly migrate questions to and from the sandbox
  • Add an official solution for sandboxing, like a sandbox subdomain or something like collectives

Any solution that would have a meaningful impact would probably take a lot of work for SE, but it would be immensely appreciated on sites where sandboxing is encouraged, and it would likely decrease the number of closed or negatively voted questions on sites like CGCC massively.

3
  • 7
    It might be worth thinking about what other sites need sandboxes too - y'all might not be the only one.
    – Journeyman Geek Mod
    Jul 29 at 16:51
  • @JourneymanGeek I definitely think the idea of sandboxing could work with a lot of other sites, and the main problems with its adoption are likely the same ones CGCC struggles with. I think a highly visible sandbox that makes posting drafts (and reviewing them) easy would help a lot with that. Jul 29 at 16:54
  • 1
    This would indeed be very useful compared to most of the other suggestions here! Jul 29 at 23:02
38

Could we have Chat (and meta) on the sidebar?

This is kinda a request that goes across communities, but also is about communities (Very Zathras, I know).

In the old days, chat was a "first class" citizen of the network, sharing a top bar with the main site and meta. Both (per site) meta and Chat are second class citizens compared to Teams. It would be very nice to consider having per site meta and chat represented on the left bar - as our second and third spaces.

People don't really find meta (or end up here by accident) or chat as easily as they did, and that is detrimental to the vibrancy of the respective spaces for each community.

3
  • 11
    I agree. I really think that if we're going to have it, we should expose it. I don't like "hidden" features, because that creates exclusive "in" groups (those in the know).
    – Philippe StaffMod
    Jul 29 at 12:48
  • I only found chat by literally clicking every kind of link I could find on the homepage to figure out where it went. Would love for that process to be easier :)
    – bobble
    Jul 30 at 2:02
  • Yep, I honestly am barely even aware that Chat exists, other than when people refer to it in Meta posts sometimes. (Then again, for me in my timezone, I have never found Chat at all useful in the few times I have tried it.) Aug 4 at 23:15
36

This is something that recently came up over on SFF as it is very similar to one of the questions (#9) asked in the election questionnaire. In my answer I mentioned new user onboarding. However, this is something that looks to be being worked on at the moment and some good steps have been taken there so I won't focus on that here. I'm sure @Catija will also be glad that I'm not going to ask about site theming this time around!

What I do want to focus on is something a bit more specific to SFF though not solely applicable there:

Tag Hierarchies/Dependencies

Many sites in practice employ Tag Hierarchies, meaning that every question with this tag must also have that tag.

  • On SFF, for example, all [avengers-endgame] questions should be tagged with [marvel-cinematic-universe] and in turn all [marvel-cinematic-universe] questions should also have the [marvel] tag. This means if you're asking a question on a specific MCU film, you need 3 tags by default out of the 5 maximum. See here and here for example, and there's a long-standing feature request for tag hierarchies.
  • On Literature, similarly, all [nineteen-eighty-four] questions should be tagged with [george-orwell], and the same for any story with a known author, while all [war-and-peace] questions should be tagged with [leo-tolstoy] and [russian-literature]. This means if you're asking a question on a Russian novel, you need 3 tags by default out of the 5 maximum. The site has a clear, consistent, and rigid tagging system, but the 5-tag limit is often a problem.
  • On Puzzling, again, all [geometry] questions should be tagged with [mathematics], and all [sudoku] questions should be tagged with [grid-deduction] and [logical-deduction]. (Actual tagging isn't fully consistent with these policies, but a few users are working on it slowly.) This has come up a few times on meta, and tag hierarchies have been requested again.
  • On M&TV they have something similar to us on SFF with universe/franchise tags and specific work tags. It is also likely applicable to the more technical sites where library tags and version tags probably want the main language tag all the time: the main meta feature request linked above refers mainly to SO.

In short, we need a way to make it easier to group tags together that rely on each other: something more automatic than just writing "please use this tag together with this other tag" in the wiki excerpt and then a lot of editing when new users inevitably fail to notice or read the wiki.

The change here would essentially mean that we could mark tags as dependant on one another and so if a question is tagged with Tag A it will also always be tagged with Tab B, C, etc.

This might mean a change in the number of tags allowed on a post as well, or that the higher level tags are not shown on the question but are linked to it in some other way, but the posts will still be discoverable through those higher level tags too.

14
  • 3
    FWIW changing the way privileges are awarded to come from an action based system rather than a rep based system is something I've spoken about on and off over the years but that's at a system level rather than community level and might be out of scope for this discussion. It would also be a major overhaul of the way the core system works. Jul 29 at 10:26
  • 1
    I got permission from you to edit this answer, hope it isn't too much :-) I was already drafting an answer along the same lines, so I edited yours to avoid redundancy. Jul 29 at 11:03
  • This would also be used (to a lesser extent) on Code Golf with the [math] tag. We don't have any requirements to use tags with it, but if a challenge is tagged with [math], it likely has others such as [arithmetic], [number-theory], [sequence] etc. Jul 29 at 11:07
  • 6
    Tag taxonomy is something that I find really interesting. I'm always amazed and the level of thought that goes into it - and most people never appreciate it, but it's so important. Thanks for bringing this up.
    – Philippe StaffMod
    Jul 29 at 12:47
  • 3
    +1000 !!! Tag hierarchies would be extremely helpful on a lot of sites! Also, PhysicsOverflow (which is not in the SE network) already has this implemented, so it can't be too hard. @Philippe I love this suggestion! Jul 29 at 22:50
  • 1
    We could use this on Genealogy to clarify our tagging significantly Jul 30 at 10:20
  • Oh yes, Code Review could definitely use this for language tags. We'd also like a group of tags that are used like discussion and bug. A group of which one tag must be selected before posting (again, for language tags). Code is always written in a language, we don't accept pseudo-code questions.
    – Mast
    Jul 31 at 7:49
  • I requested something similar which would be fixed with tag hierarchies on CR meta, but was shot down. We have language tags which must be applied to questions, and language version tags which can be applied to questions. [python-3.x] alone is not a 'valid tagging scheme'. It'd be so nice to be able to mark all our game tags as [game] children and all our algorithms tags as [algorithm] children. We can't manually fix this because a {game} using {algorithm} in {language version} would go over the tag limit!
    – Peilonrayz
    Jul 31 at 12:10
  • @Philippe yes, it's important. Please, keep it flat and make people accept that hierarchies are just a solution looking for a problem. Change "tags" to mean "categories" as the help center calls it and make it more explicit that they should be unambiguous and well defined.
    – Braiam
    Aug 1 at 20:40
  • 1
    "every question with this tag must also have that tag" I feel you are using tags wrong: meta.stackexchange.com/a/18879/376358 Point 1: <Each tag should stand on its own>
    – Federico
    Aug 9 at 12:05
  • 1
    @Federico I feel you've misunderstood the point that's being made there: "if a tag only makes sense when used in a group with other tags, it's wrong" - the tags do make sense on their own. It's just that those questions are also in scope to the broader tag(s) and so need those as well. Aug 9 at 12:08
  • @TheLethalCarrot I see what you mean
    – Federico
    Aug 9 at 12:11
  • Couldn't a system not automatically complete the tag list if the relation between the tags is that strict? If tag A always requires tag B then simply add B automatically as soon as A is added.
    – Trilarion
    Aug 23 at 10:34
  • Mentioning in case anyone wants to see one live implementation: Codidact has had tag hierarchies for some time now. Sep 27 at 18:05
34

Stack Overflow: Discourage overly situational debugging questions, encourage more generally useful questions.

This likely means less focus (especially by high-rep users) on if the asker has put in enough effort to deserve a solution and more focus on whether the question is a specific, clear programming problem.

As Shog9 puts better than I ever could, the only question we should be asking about effort is "has the asker put enough thought into the problem to formulate a clear, specific question?"

This means no more comments that are basically just "What have you tried so far?" (which you can still post as long as you add "so far") or "Stack Overflow is not a free code-writing service" (it is).

Such a strong focus on problem-solving effort emphasizes debugging questions so specific to the asker's situation that they're unlikely to be of use to anyone other than the original asker (Shog9's above-linked answer, again, explains this better than I can). Frequently, such debugging questions technically comply with all the rules, get answered, the answerer gets some rep, and then the question is never seen again by anyone who finds it useful for the rest of eternity.

33

This is a short one:

Bring chat into this century

9
  • 16
    I mean, the site was created in 2008, so chat was definitely created in this century... this decade though, that'd be nice. :D
    – Catija StaffMod
    Jul 29 at 16:56
  • 4
    Luuklag's not totally wrong though: if I build a model T Ford today, it's still of the last century. :). The interface looks old, and it's missing modern-appearing skinning, for instance.
    – Philippe StaffMod
    Jul 30 at 2:33
  • 2
    @Philippe and stickers! ;) Jul 30 at 10:51
  • 2
    A critical factor here is tech debt. Chat going the way of the mobile app is going to have a pretty detrimental impact for many of us
    – Journeyman Geek Mod
    Jul 30 at 13:41
  • 3
    Chat on mobile is still atrocious. If it's redone, make sure mobile isn't forgotten.
    – Mast
    Jul 31 at 7:58
  • 2
    Yes, @JourneymanGeek, theres almost no question that chat will need to be reworked. There are issues with it on .net, and it’s a heavy lift to get it pulled up to modern standards. And mast, one of the linchpins of my pitch when people ask me what I want to accomplish here… is mobile. we won’t be ignoring that.
    – Philippe StaffMod
    Aug 1 at 13:22
  • 5
    If there's a rebuild - an API would be a welcome part of that. There's a ton of community built resources that rely on somewhat jerryrigged bots that would benefit/need work when that happens
    – Journeyman Geek Mod
    Aug 1 at 14:35
  • You mean all visuals are flat and huge (thanks God SE still has at least some volume in its design and reasonable sizes!...), messages with a big padding, on different sides of a window, emoticons with indistinguishable emotions (still nobody have beaten the Skype or ICQ ones...), stickers with even more indistinguishable emotions or meaning in general??... Please, NO!! I see nothing good made in web design in the past 5-10 years. I'd rather bring most of the things into the previous century... TL;DR: Please, be conservative with what you're gonna do!
    – EvgenKo423
    Aug 4 at 20:27
  • 2
    As a possible voice of disagreement: I still find SE chat the most useful of all the chat tools I have to use. That includes Slack, Discord, Teams, Signal, Wire, Messenger and others. SE chat wins hands down. So if modernising it means making it anything like those others, please don't!
    – Rory Alsop
    Aug 19 at 15:05
31

Stop people from enforcing the rules without questioning and understanding their purpose.

We have a lot of rules here, and that is a good thing most of the time. But the rules tend to take on a life of their own, independent of the purpose they used to serve. Most of our rules boil down to ensuring that people are civil and that the content meets a certain quality threshold. The specific rules are simply patterns we saw that indicated problems.

This can get to a point where people simply match patterns and act on anything that matches that pattern, no matter if the content actually presents a problem or not. This is in part due to some of our rules being inherently difficult to judge, but my impression is that in the end we always end up with more of a "letter of the law" enforcement than any true understanding of the purpose of our rules.

Half of this problem are popular, but inherently problematic rules like those against homework questions. They're very easy to overapply to posts that are not a quality problem. But you can misuse most of our other rules as well.

This is not about a single site, but about two that I feel attached to. How much of an issue this is probably depends quite heavily on the dynamics of the community and the topic, some topics are inherently more difficult to deal with.

8
  • 8
    No, WIKIPEDIA has a lot of rules:en.wikipedia.org/wiki/Wikipedia:List_of_policies_and_guidelines. We are but babes compared to that particular mess (but your point is valid, nonetheless)
    – Philippe StaffMod
    Jul 28 at 20:06
  • 2
    Separate SE sites interpret the rules different ways. On Bicycles we're pretty forgiving about grey-area posts, but other SE sites are hard-nosed and delete-happy. Upshot - not every site is the same, and being pulled up for not enacting the rules the same as another site is meh.
    – Criggie
    Jul 29 at 1:19
  • 3
    @Philippe I don't know much about the admin side of Wikipedia, but I think the best analogue of Mad Scientist's excellent point here would be this Wikipedia rule (hat-tip Shog9 for the link), which we don't explicitly have as a principle here on SE. I'd be in favour of codifying something like that in the help centre, which might help to address the issue that MS raises. Jul 29 at 8:43
  • 3
    Ah, @Randal'Thor. You mean WP:IAR (Ignore All Rules) right? I can opine at great length on that one, but my feeling has consistently been that it's analogous to the Queen refusing royal assent. It's a safety valve that you hope that you never need to use. Ive seen IAR invoked maybe twice, legitimately, on Wikipedia (and that's in 10 years). I invoked it once myself, and was wrong (I failed to consider some things - I was younger and stupid(er).) It's a hard one, & in my opinion it's a great statement of philosophy, but it needs to be behind glass with a red "in case of emergency" hammer.
    – Philippe StaffMod
    Jul 29 at 12:42
  • @Philippe And here my ignorance of the Wikipedia admin world emerges :-) I didn't realise that over there it's a rule so strictly defined as to need being "invoked". In the SE context, I was thinking of it more as a philosophical point to include somewhere in the help centre, to remind some over-enthusiastic appliers of letters of rules that sometimes it's better to consider what's best for the site than to blindly apply written policies. Jul 29 at 12:59
  • 3
    As a specific example: SFF meta has long had RIP posts when famous writers or actors die; some people said that's not what meta is for, we had a big discussion, and Shog said essentially "meta is for whatever you want it to be for, is this causing any actual problem?" There's a case where it might've been helpful if mods could've pointed to a clear principle of "don't blindly apply rules if there's no actual problem to be solved". Of course you have more experience of what actually happens when such a principle exists. Would be an interesting thing to discuss at more length sometime. Jul 29 at 12:59
  • In my experience as a casual Wikipedia editor, mostly watching other people from the sidelines, whenever IAR gets invoked, it nearly always results in much argumentation and debate over whether it was appropriate to do so, whether the rule being ignored is even legitimate in the first place, whether the person invoking IAR is being "disruptive," etc. It's always a drama magnet and pretty much never results in everyone going "Oh, well that's OK then."
    – Kevin
    Jul 30 at 21:28
  • @Kevin, that’s exactly correct. That’s why I make a big deal out of iar being used legitimately. I’ve seen it invoked a lot, but very rarely correctly.
    – Philippe StaffMod
    Jul 31 at 19:30
30

Better spoilers

The site I started on, Puzzling, has a lot of spoilers. Many (if not most) answers have a significant amount of content in spoilers, and many questions have a hint (or hints) in spoilers. That's the site culture. It comes with how we've tried to force the SE Q&A model to fit a puzzle-solving site.

Given that, it would be wonderful for spoilers to be improved. There are two big things:

All that content that Puzzling has in spoilers could benefit from better, easier formatting; people are reduced to using funky hacks or just giving up on proper formatting altogether. The times that I've had to help someone figure out how to spoiler code, or explain that tables don't work, or fix up a post with an apologetic "I'm new and can't figure out the spoiler formatting" etc. are numerous. And making spoilers more accessible seems like a no-brainer (given a magic wand - I'm sure it's more complicated behind-the-scenes).

For something that combines both of these bullet points, take my crusade for Markdown table conversion. I have an SEDE query to find posts which have MathJax tables, so that I can edit them into Markdown ones to improve accessibility. But... a lot of the tables I find are inside spoilers, and I can't fix those. Markdown tables simply don't work in spoilers, and even if they did, my target audience of vision-impaired-folks wouldn't be able to get to them anyways (assuming they're keyboard-only).

So given a magic wand, I would give my community better spoilers.

1
  • 8
    This is one that I haven't heard come up before. Thanks!
    – Philippe StaffMod
    Jul 29 at 12:46
30

More capability to customise the Help Center

The elements of the Help Center that moderators can customise on behalf of their community are quite limited and limiting. The fixed elements are all about the working of the SE 'engine' and there isn't really enough 'space' to talk about the workings of the specific community, especially the community-specific expectations/aspirations for good questions and answers.

Yes, we can include useful guidance that fits naturally in the current structure, such as what is on- and off-topic, but there's other stuff we have either to shoehorn under a irrelevant topic -- making the specific information hard to find, and making that topic (What topics can I ask about here?) longer and longer and less likely to be read.

And yes, we can point to Meta questions but I don't believe Meta questions are a good means of communicating with new users -- the format is confusing to them. Why is it a question? Does that mean it isn't a policy? Why isn't this in the Help Center anyway -- does that mean it's less important?

Give us the ability to have sub-pages in the Help Center -- I won't try to design the structure here, but stuff hanging off 'how to ask a good question' and 'how to write a good answer' would be a start. And somewhere we could put our specific privacy guidance, guidance for including images of sources, guidance for asking for software recommendations, guidance on 'lookup requests', ... All the stuff that is current community policy that we expect/prefer askers and answerers to take into account. If nothing else, it should help improve the experience for newcomers and maybe that will make them stick around -- we have too many hit and run questions and we need to convert some of those into community members not drive-by posters.

6
  • 7
    Agreed, why /help/on-topic is editable but /help/dont-ask isn't is beyond me initially. And /help/how-to-answer should also be editable. On a similar note parts of the Tour would benefit from it and that expanding to cover answering more. Jul 30 at 11:33
  • 2
    Agreed; new users shouldn't have to go on a safari to figure out how to format their first post on a given site... Right now, the only way to know what's on topic for sure or how certain post types should be written in some cases is to dive into the site's Meta, which isn't ideal, and is kinda hidden for anyone not familiar with Stack Exchange.
    – zcoop98
    Jul 30 at 14:17
  • @zcoop98 Plus Meta can be intimidating (even more so these days with Mods and Staff announced in big bold orange and blue). When I venture to other sites (which I do) I stay well clear of their Metas, so I can't blame new users for not venturing into 'ours'. Jul 30 at 15:31
  • Candidates for your list: guidance for homework questions, guidance for conceptual questions (big picture), and guidance for big-list questions Jul 31 at 11:21
  • @P.Mort.-forgotClayShirky_q Not things that are a problem for genealogy.se but relevant in the bigger picture, yes. Jul 31 at 11:24
  • 2
    Yeah, the onboarding experience, quite frankly, needs some work. That's really high up on my list.
    – Philippe StaffMod
    Aug 4 at 8:14
29

For the SO community, more people willing to take action rather than just leaving comments.

If a question requires a dozen comments to eventually dig out all of the information required to answer it, it should have been closed in the first place until the op makes the necessary adjustments to avoid people posting guesses as answers. Ideally this would cut down on answered dupes and allow us to more effectively route users to time-tested answers.

A few potential options:

  • Open up close/reopen voting sooner for high-traffic sites to reflect the fact that the majority of users are sub 3k
  • Open up privileges based on actions rather than (or alternatively/in addition to) reputation
  • Some just-in-time info when a comment discussion is detected to suggest improving/editing the question and/or closing
  • Mind Control/Mass Hypnotism
  • Artificial Intelligence

I am aware that users that are sub 3k can flag posts for closure, but the UI for this is confusing. It should instead follow an identical workflow to closing questions rather than being hidden within the flag dialog. My goal would be better informing users of the tools that help keep this place functional earlier rather than it feeling so adversarial until they've been here long enough to realize why things are the way they are by empowering them with the very tools we're using. You're far more likely to pay attention to a privilege you've earned than one you were introduced to as preventing you from getting the help you need.

7
  • 2
    I'd add to this "reducing the stigma surrounding having a question closed or marked as duplicate" too (I might spin this idea into its own answer tbh). I feel that it would be hard to increase the amount of general and duplicate closures if new users continue to deeply dread closures and see them as a kiss of death rather than simply a chance for improvement. We as a community can probably improve our messaging in this regard, but design and onboarding changes would help too.
    – zcoop98
    Jul 28 at 20:51
  • 18
    I'm a big proponent of rethinking how we award privileges - while I know that not everyone understands how to effectively close or edit a post early on, I also don't think it makes sense to force people who have shown that they can do so effectively to earn the rep - and, well... not everyone who earns the rep knows how to use the privileges, either... that said, I'm sorry to say that we're not planning anything in the future along the lines of your fourth bullet.
    – Catija StaffMod
    Jul 28 at 21:58
  • 5
    @Catija Oh good, that means the fifth is still on the table
    – Nick
    Jul 28 at 23:13
  • 13
    Pretty sure she meant that the FIFTH was off the table. I'm actively working on the fourth.
    – Philippe StaffMod
    Jul 28 at 23:51
  • @Philippe Maybe you should focus on the fifth instead. Seems much more practical in the long run :P
    – VLAZ
    Jul 29 at 6:04
  • 2
    Yeah, well, I like to think big. Out of the box, we call it.
    – Philippe StaffMod
    Jul 29 at 12:34
  • Hey Kevin! Can you please make the necessary edit in the row which is related to your answer (Stack Overflow row) in the table in the question? Sep 12 at 12:17
27

RPG.SE: Make it possible for RPG system tags to always appear first, and display them in a special color

There are two features I think RPG.SE would really benefit from (that would likely also benefit other sites on the Stack Exchange network), and they're both summarized well in existing feature requests on our Meta:

Basically, on RPG.SE, most questions (though not all) are primarily asking about a particular role-playing game system, and often a specific edition of that system - whether they're asking about the rules, or a problem they've encountered, or a group-dynamic issue. This is generally reflected by our tag system, so whatever system a user is asking about, there'll usually be a tag on the question to indicate it.

The name of the first tag on a question appears before the question title in the title of your tab/browser (when viewing the question), and in Google search results. In addition, the system/edition tag is often an important signal to users to help them judge whether they have the relevant expertise to answer the question.

Since D&D is a popular game, a D&D edition tag will almost always be the first tag listed on a D&D-related question... But there are thousands of other RPGs out there, and most of them aren't so lucky.

Say there's a question about the Lancer RPG. Because there aren't very many questions about Lancer on the site, then if other tags (e.g. ) are added to the question, they take up the first spot in the question's tags instead - meaning we have to either intentionally avoid adding a useful/relevant tag to the question, or cause the relevant RPG system to not be listed in search results/page titles.

Making it so certain tags can be designated to always appear first would allow us to have the best of both worlds. As stated in my feature request (the first one linked above):

This would work like the four special tags on meta: all special tags are sorted by popularity first, then all normal tags are sorted by popularity.

This would allow us to always have the RPG system/edition (where specified) be clearly listed for a question, while also letting us include other relevant tags (even when they're more popular than the system/edition tag) without causing any problems.

In addition, as stated in the second question linked above:

It doesn't help that a lot of RPGs have names which might just refer to a game element or procedure of play:

Even when a tag like this appears in the tag list, some users (especially new users) don't realize that the tag refers to a particular RPG system; some of them simply assume the question is about D&D, and answer accordingly. This just results in frustration/annoyance for both the querent (who's getting an answer that isn't relevant to them at all) and the other users that help moderate the site (because they then have to flag it as "not an answer" and/or vote to delete it). It also contributes to an overall sentiment (well-documented on our Meta) that RPG.SE often feels more like "D&D.SE", making players of other RPGs feel unwelcome - or at least feel like the site isn't going to be useful to them.

The two suggestions I've linked above - making it so RPG system/edition tags can always appear first in the tag list of a question, and displaying the tag in a special color - would help avoid/solve these recurring problems. This would be a definite quality-of-life improvement for RPG.SE.

(I think there are other sites that could also benefit from the ability to designate certain tags as "special" as well, such as tags for particular video games/franchises on the Arqade, or specific programming languages on Stack Overflow - RPG.SE is just the one I've spent the most time on.)

5
27
+100

Archive feature

On Ask Ubuntu we have a lot of questions, and especially a lot of answers, which were super relevant and useful to a lot of people at the time they were posted, but which no longer have value in helping folks find out how to do x, because how to do x has changed over time. Many such posts are highly upvoted, jump up on search engines, and if they are answers show at the top of the by-default-vote-ordered pile.

We don't delete such posts, and I don't think we should delete them or even necessarily edit them to update them (often this would involve replacing all of the content, perhaps making it the same as other existing answers to the same question), because they have a social history value and perhaps can be of use to people interested in user experience and various other aspects of software development - who knows? But the mass of obsolete material is only increasing over time and gradually making it harder for visitors and reviewers to find answers that actually give accurate and current information. It's not at all the case that everything old is out-of-date, but enough of our posts are out-of-date that it's an issue.

Recently Monica from the Ubuntu community team gave me the idea of an archive feature. The way I think it could work would be different for questions and answers.

For answers, 5 users could vote to archive an answer, and it would be pushed to the bottom of the page in all views under a notice that the below answers are archived and may not work any more.

For questions, 5 users could vote to archive a question, and a notice would appear at the top of it stating that it is archived and may no longer be relevant, and optionally and preferably, linking to newer questions with current answers on the same topic.

10
  • 5
    There currently is work being done to adress the problem of outdated answers: meta.stackexchange.com/questions/361097/…
    – Luuklag
    Aug 2 at 7:42
  • 2
    @Luuklag do you think I should post this there instead (or as well)? (I totally disagree with Glorfindel's answer there about Ask Ubuntu. That silly close reason we have does nothing to address this)
    – Zanna
    Aug 2 at 7:50
  • 4
    If you have something new to bring to the table there (be sure to also check out the linked MSO question) go ahead and post there as well. Do make sure you answer that question specifically, so a blatant copy-paste is not recommended.
    – Luuklag
    Aug 2 at 8:01
  • are you aware of historical-lock feature available for questions?
    – gnat
    Aug 4 at 8:33
  • 3
    @gnat yes but I don't like that feature because it prevents anyone from ever interacting with the post again, and doesn't actually help people find up-to-date information instead. The main problem we have is with answers rather than questions anyway.
    – Zanna
    Aug 4 at 8:48
  • 1
    @Zanna I read the recent outdated answers threads and this archive feature seems to rarely have been mentioned. It's an interesting perspective because Ask Ubuntu and SO are different in nature and a particular site probably allows to narrow the overall problem.
    – bad_coder
    Aug 7 at 0:37
  • 1
    @bad_coder I think Machavity came up with something somewhat like this on MSO, mediated by tag badges. I like the idea of using tag badges as this requires field-specific knowledge more than other types of review, but there aren't enough tag badge holders on our site for that to be workable unfortunately. Still, I'm sure there's some way to get it done :)
    – Zanna
    Aug 7 at 4:14
  • Ask Ubuntu doesn't have said problem. This is something created by the community. This question have been updated since '13 every time a new version of Ubuntu comes out. Do that for every question and this problem disappear. Make questions less about the version and more about the problem.
    – Braiam
    Aug 8 at 14:11
  • 2
    @Braiam Ask Ubuntu is the community. This longstanding problem is not going to go away without better tools. People do write new answers to old questions but the new answers are stuck at the bottom of the page. Some questions have no relevance to new versions (e.g. stuff about unity DE). People close as dupe without checking whether the answers are still current.
    – Zanna
    Aug 9 at 4:54
  • At the very least you can already edit outdated questions and answers and add a notice that the practice is outdated nowadays (maybe with some reasoning why and preferably with a link to a better solution). Not sure if the archive sort at bottom feature is the best solution. A trending sort order (which is on by default) might work better/more smoothly.
    – Trilarion
    Aug 24 at 11:20
25

Site-specific new user onboarding

Onboarding new-users has been a focus of several past and current initiatives. However, from my perspective these have always been too focused on Stack Overflow.

As moderators and curators on other sites, we don't have any ability to change how the asking process looks for our users even when our requirements and the types of "bad" questions we get are completely different from the ones on SO.

New users shouldn't be so surprised that medical advice and self-help questions are off-topic on MedicalSciences.SE, Biology.SE, and Psych&Neuro.SE after they type in their question (and no, the ugly disclaimer on MedicalSciences is not the way to go; it's also not even visible on the mobile site. This is a legal/liability feature, it doesn't do anything for site quality and anyone experienced in dodging ads and other irrelevant material on the interwebs is going to automatically filter out that yellow background text as "extra stuff not relevant for me to read"). New users to Academia.SE shouldn't be so surprised that the site isn't for asking about their organic chemistry homework.

Close votes and comments should be about responding to the questions that still sneak by the information, not the sites' first line of defense.

Do your research, then give us the tools

I don't think its feasible for the CMs or any other staff to review these sorts of things for every community. The time commitment would be huge across the whole range of sites. However, you have some dedicated experts in each community that are going to be eager to help write copy for this. Let them.

There will no doubt be more opportunity to do A/B testing on SO to determine how to best present the messaging to get (at least some) new users to actually heed the advice given. However, I think the steps taken on SO should be designed with the rest of the network in mind. Think at an early stage about what might be customizable across the network and how to facilitate that, rather than just doing what makes sense for SO and maybe later kind of squeezing in a half-solution for everyone else.

5
  • I think this would be a better fit here but I might be wrong. Jul 30 at 16:35
  • @EkadhSingh Perhaps; I wrote it before that post existed but got distracted and just got back to finishing it up now. Jul 30 at 16:43
  • 3
    At least two other posts here target a similar issue: meta.stackexchange.com/a/368034/401068 meta.stackexchange.com/a/368007/401068 but I stubbornly wanted to make my own take on it, and those other proposals are themselves also focused on specific aspects of site customization of help/onboarding content. Jul 30 at 16:49
  • 2
    I think there’s real promise in this concept and it’s one that I intend to ask my team to develop a couple of experiments around.
    – Philippe StaffMod
    Jul 31 at 19:32
  • @Philippe Occasional tips like these would also help with on-boarding in general. But such things could be annoying and there definitely should be a way to disable them.
    – EvgenKo423
    Aug 5 at 9:56
24
+200

I don't have any specific community in mind. My request is more generic and about the need of the community as whole.

My request isn't technical, and it isn't about policy, but I guess it fits under "something about the environment in which you operate", so I'll give this a try.

The environment in which I operate is Stack Exchange as whole, and something there is still off. Something still prevents many people from getting back to full activity.

So, my request is for the company to fix a major past mistake which has lasting effect to this day. Many, both inside and outside the company, keep saying it's impossible, and maybe it is; yet I think there's no harm in asking anyway. All it needs is one single apology, done properly.

Those who need to know what I talk about exactly know it already; I prefer not to go into specific details here as it would lead to undesired side effects and my request will be lost in the smoke. The essence of the request is simple: fix a past mistake. You (the company) already made major steps towards this.

15
  • 7
    Thanks for the response. My position on this situation remains the same as it did last week: the situation happened prior to my arrival, and I wasn't there. I think it would be height of hubris for me to think that I could come through and thread the needle so carefully as to make people on both sides feel okay with things. Instead, I'd rather look forward and do my best to see that a similar situation never comes up again.
    – Philippe StaffMod
    Aug 1 at 10:51
  • 2
    @Philippe totally fair, but still, you're in a position where you can just hand over the needle to someone who can use it, hence I was asking. Aug 1 at 11:07
  • 15
    @Philippe that's a cop-out, this isn't a "both sides" situation where two factions within the community were pitted against each other or something. A couple of people within the company just decided to channel their inner pointy-haired boss, kick a veteran volunteer to the curb and, to add insult to injury (it seems) as a final step had a bunch of lawyers put a gag order on her. As pointed out in my answer linked above, there absolutely are things that the company could do to fix this & I will take refusal to even look into it as a signal that nothing meaningful has changed since.
    – Pekka
    Aug 1 at 19:12
  • 7
    I rather think that you just proved my point for me. When an attempt to say "anything I say will aggravate somebody, so I'm going to avoid saying anything"... aggravates someone... well, yeah. It reconfirms to me the fact that there is significant study and work to be done before I would even think about wading in there. Not to mention, I still have to answer to legal. Ultimately, this is not my decision to make.
    – Philippe StaffMod
    Aug 2 at 2:52
  • 6
    @Philippe well, I read your position so far as not wanting to do that significant study and work in the first place and that is a cop-out. Nobody is expecting you to have an instant opinion on the matter, and obviously you bear no personal responsibility for something that happened while you weren't there, but you inherit the problems created by those who occupied your position before you.
    – Pekka
    Aug 2 at 8:36
  • 4
    Also it really isn't a situation of "anything I say will make someone mad" like it so often is when there's a highly polarized, intra-community struggle. This wasn't one. Try saying "I'll make a serious effort to look into it and take it up with my superiors even if it might be inconvenient" and it's highly unlikely anybody will be seriously aggravated :)
    – Pekka
    Aug 2 at 8:39
  • 10
    I think you're reading things I never said. :-) I'm not afraid of study and work - but I do have to prioritize things based on several different stakeholders, timeliness requirements, and many other factors. But the reality is, even when that is said and done, it's not my call. I can advocate all day but it's not my call. But I will compromise: I will say that I will make a serious effort to look into it and follow up internally (even if inconvenient) at the appropriate time, based on workload and prioritization.
    – Philippe StaffMod
    Aug 2 at 8:42
  • 7
    @Philippe 👍 thank you!
    – Pekka
    Aug 2 at 8:43
  • 7
    @Philippe To be honest, not saying the right thing will also poison people against you here. "He knows and did nothing" is significantly worse then "He tried to address it but failed / it was tone deaf". I appreciate this is not your decision to make, and i'm sorry that you're going to continue harvesting the bad feelings for this, but unless it's resolved this will never stop, nor should it. Anyone coming into a leadership position at SE that does not positively address this issue will be seen as complicit in the ongoing harm that is being done still.
    – magisch
    Aug 3 at 9:37
  • @Pekka - there was totally inter community struggles. Meta was a mess and we closed down the mod chatroom temporarily. It would be inaccurate, even after so much time, or maybe especially so, to believe 'no one would be mad'.
    – Journeyman Geek Mod
    Aug 4 at 0:52
  • 1
    @JourneymanGeek weren't those struggles largely about the new speech rules though, around pronouns and such, rather than the way Monica specifically was treated for being doubtful about some of those rules (while being more of an actual "trans ally" in real life than most of us here), and the insinuations made about her to media outlets (or at least one outlet) afterwards? I'm not suggesting revisiting the former.
    – Pekka
    Aug 4 at 11:28
  • The new speech rules were... a bit of a distraction, and an unwelcome one - they were both the cause of these events and kinda made things messy - there's an entire spectrum of how people felt about events, and different viewpoints. I would point out (and some of this leaked) - we had to freeze the Teacher's Lounge a over auguments related to this issue - and they went on for quite a bit. It is a monumental mess, and every action or inaction is likely to have its advocates and detractors. If someone can bring it to closure, and keep most folks happy,
    – Journeyman Geek Mod
    Aug 4 at 12:05
  • I think it would be an epic work of diplomacy
    – Journeyman Geek Mod
    Aug 4 at 12:05
  • 5
    @JourneymanGeek true, but rather than "epic work of diplomacy" which might as well be actually impossible, what Pekka asks for is one single apology. Sadly, might be as impossible as amending all mistakes done with the hundreds of moderators. :/ Aug 4 at 12:43
  • 2
    @Philippe "...I wasn't there..." True, but the company was there. On behalf of the company, anyone from the company could actually do something, if the present company wanted that. Maybe you could just mention it internally sometimes that there is some kind of public demand for action and then together with others the company could decide to either act or not to act. I wouldn't see it as a personal thing, rather as a company thing.
    – Trilarion
    Aug 24 at 11:04
22

Allow more syntax highlighting

HighlightJS ships a lot of useful languages, of which SO & other programming sites only use a small fraction of.

Syntax highlighting is very useful for clarity and readability. If someone posts a AppleScript question, the code would be a lot easier to read, and therefore be understood/answered better, if SO had the AppleScript language mode.

I know that this is because of performance, but since most questions only use one language, would it be possible to dynamically load whatever languages the question requires?

This probably would require a quite substantial overhaul, but I believe it's possible.

3
  • I would add that even for supported languages highlighting is not as rich and distinctive as in code editors. At least for Python.
    – EvgenKo423
    Aug 4 at 20:38
  • @EvgenKo423 I think that's mainly because SO has to use light mode, which can't be as distinctive as dark mode, which most people use.
    – A username
    Aug 4 at 20:49
  • Well, the code is on the grey background, so I'm not sure...
    – EvgenKo423
    Aug 4 at 21:07
20

Better accessibility.

When users with disabilities struggle to use SE, this also means that they may not have much of a voice here.

In the 2021 developer survey, 4,1% respondents claimed to have a physical difference with the majority being blind or having difficulty seeing. This is the highest result when we compare to past surveys which varied between 2,8% and 1,7% of respondents claiming to have a physical difference. We don't have such a survey for the whole of SE as far as I'm aware so this is the most representative number I can find for our user base.

According to a 2017 article from the Pew Research Center, "[...] there have also been dozens of lawsuits in recent years, claiming some websites are not accessible to those with disabilities."

This doesn't relate to a specific need of one of our communities but back in 2018, our former SO product manager Joe Friend wrote :

I want to see us make improvements for a range of users with differing abilities. My perspective is that improvements to users with specific needs can lead to improvements for everyone.

Better accessibility would not only make all of our communities more diverse and welcoming, it would also mean an improved experience for all users.

There was a lot of discussion around being inclusive on SE in 2019. Back then, a few moderators started compiling a long list of unresolved accessibility issues related to screen readers, color perception, keyboard accessibility, dyslexia, visual impairment, tremor, and more. Unless they've been resolved, some of these issues seemingly affect core community mechanisms.

A long list of unaddressed accessibility issues on MSE can be found by searching with the following queries:

Questions on MSE including the accessibility and feature-request tags that are not status-completed

or

Questions on MSE including the accessibility, bug, and feature-request tags that are not status-completed

5
  • 1
    One important thing to note is that accessibility solutions don’t just benefit the people that need them, they are features that can benefit everyone. I have no trouble hearing and I enable subtitles often. Text to speech sure is nice to have when I’m driving and get a text. I turn off images when I browse sometimes to conserve bandwidth and battery life. Accessibility is not about catering to some percentage of an audience that depends on specific features, it’s about making a site/service/video game easier to use and more flexible for everyone.
    – ColleenV
    Aug 3 at 16:42
  • 2
    and yet... both of the most notable changes have resulted in a site that is harder to read.
    – Kevin B
    Aug 3 at 17:19
  • And better accessibility benefits all users. Perhaps add something about this to your answer? Aug 4 at 11:43
  • @P.Mort.-forgotClayShirky_q That is why I had added the quote by Joe Friend. I calculated the percentages because I'm curious but they seem to be construed like catering to a small subset of users which is not what I intended at all...
    – curious
    Aug 4 at 13:05
  • 2
    Sort of related. There is an episode of the 99% Invisible podcast that is about architecture accommodating deaf people. I found it quite fascinating. It also does point out how designing for the deaf benefits the hearing people, as well. Some of the design decisions are to limit the eye strain, for example. It is a 12 minute episode, and a great example for how helping people with disabilities can help all.
    – VLAZ
    Aug 4 at 17:34
18

As I mentioned in chat when suggesting the topic, I'm afraid my "pet peeve" or thing that I would change would be the relatively new policy of having to be 18 or older to become a diamond moderator.

As I wrote in my answer(s) to the original announcement, there have been numerous moderators around the network who became moderators under the age of 18, including some of the most well-known or influential moderators on the network.
I also became a moderator at the age of 14. I stepped down before turning 18, which meant this policy change didn't affect my personally - I reclaimed my diamond a few days ago, and I just turned 19 - but it still felt like a slap in the face when it was announced. Given the history of young moderators on the network, and the literal years of time and effort poured into improving the sites and moderating, it gives the impression that all that effort was never really valued.

The possibility of underage moderators was discussed all the way back in 2012 (long before I joined the network). There was a pretty heavy consensus that age should be irrelevant for becoming a moderator, since moderators are selected for ability and temperament in any case (leaving aside the practical issue that it's impossible to actual verify age on the internet).

All this, combined with the votes and responses to the announcement, mean that the community has made its opinion fairly clear: The age of a moderator should be irrelevant. Changing this policy back in order to allow any user who meets the age requirements laid out in the ToS (13 most places, 16 in the EU) to become a moderator, even if it means getting parental permission, would show that the community's values are reflected by the company.

7
  • 6
    As I recall - that there were no moderators who were 'underaged' at the time was something that was discussed when this rule instituted. It would have been manifestly unfair to de-mod someone for a rule that wasn't in place when they were elected. I think part of the concerns was that we'd been moving to a more formal/'legal' mod agreement and the potential for NDAs for certain activities of the mod council (and the working groups that for most part have replaced it) . Its not a reflection of those mods abilities, rather "we need this, to open up these other things, cause these things changed"
    – Journeyman Geek Mod
    Jul 29 at 3:56
  • 1
    @JourneymanGeek That's not fully accurate, now I can't be reinstated as a mod on Drones.SE due to the new policy.
    – Xnero
    Jul 29 at 22:50
  • 1
    Well if you were removed for being underaged I feel that reinstatement without preconditions when you hit 18 should be a thing even with the new rules.
    – Journeyman Geek Mod
    Jul 29 at 22:59
  • @JourneymanGeek Considering a site can change a lot in a few years, I could only agree on that if it was a matter of months, a year at most. After that, elections are the way to go. If the site hasn't changed too much and the ex-moderator was any good at their job, they should have no problem getting a landslide victory.
    – Mast
    Jul 31 at 7:55
  • @JourneymanGeek It is as far as I know.
    – Xnero
    Jul 31 at 11:43
  • 4
    @Xnero - that's a bit deceptive - your age doesn't relate to why you're not a moderator currently. While it prevents you from being reinstated, there's other issues.
    – Catija StaffMod
    Jul 31 at 19:52
  • 3
    @Catija The email you sent me in December says otherwise.
    – Xnero
    Jul 31 at 21:33
18

SF & F

I would like to see a custom question-flow for asking story ID requests. They're about half of our questions, most of them are by new / drive-by users, and most need to be coached in the comments to tell us when they read or watched it, where, what language, and so on.

A custom question-flow that would ask a (new) user to address those points would be really helpful.

6
  • 3
    If we're truly waving the magic wand and to generalise further a custom ask editor based on tags that mods could edit and customise. Jul 29 at 12:57
  • 2
    @TheLethalCarrot or even a manual choice: "do you want to identify a story?", which could be "do you need help with a homework assignment?" or "do you have a piece of non-working code?" on SO.
    – SQB
    Jul 29 at 13:00
  • 4
    @TheLethalCarrot that would require tags to come first...
    – Pureferret
    Jul 29 at 13:00
  • 1
    @Pureferret Good point. Tags are now suggested based on the question body, not the other way around.
    – Mast
    Jul 31 at 9:30
  • 1
    Maybe it's time to revive the Ask Question Wizard.
    – SQB
    Jul 31 at 9:56
  • Custom question flows might be good in guiding users on many parts of SE, but the benefit needs to be balanced with the effort to maintain these custom question flows.
    – Trilarion
    Aug 23 at 10:36
18

TeX.StackExchange: restore previous "community promotion" conventions

Within the TeX/LaTeX community, local user groups have used the community promotion ads to good effect to announce their annual meetings or similar events. Such events are both time-specific and time-critical, and after an event has occurred, an ad dedicated to that event should be replaced by something more general.

This year, however, proposing this year's ads wasn't even launched until late spring, and the ads won't be activated until August 1. By then, at least one group's meeting had already taken place, and the annual meeting of the international TeX Users Group (TUG) starts on August 5; not much time for a useful announcement.

Even worse, it will no longer be possible to replace an outdated ad. This simply leads to embarrassment and the perception, especially by newcomers, that the site is not current, or even that it is not to be trusted.

This is a request to restore the previous practice -- Start proposals for a new year's ads at the beginning of the year, or even a month or two earlier, and activate them as soon as possible after January 1. And allow replacement of "expired" announcements as appropriate.

It's also the case that "commercial" ads for SE/SO are exhibited much more frequently than the community ads, and I've rarely seen only a community ad without it being preceded by one of the organization's ads. While we recognize that SE/SO needs to support itself, it's sad that this appears to be given much higher priority than the needs of its users.

2
  • 2
    Could the mods (ab)use the Upcoming Events bulletin for this? (I'd like a more dynamic ad system than the new one, just wondering for this particular case). Jul 29 at 16:30
  • 1
    I will make a note - I'm not sure what drove the timing this year, but I"ll ask.
    – Philippe StaffMod
    Jul 30 at 2:31
17

Built-in leaderboards on Code Golf

I'm somewhat active on Codidact Code Golf. I also use the Graduation Userscript on CGCC (created back before we graduated and had a design). One feature that both of these have that I'd love to be a part of CGCC by default is leaderboards:

leaderboard screenshot

And a little tooltip that tells you the winner:

enter image description here

As this is doable with a userscript, it mainly involves a bit of custom HTML + JS being added to our site, so it's not even too crazy of a change. Currently, this is only for challenges with the tag, so that might be a bit of a challenge to do.


I find this a super useful feature to have (and is really the only reason I still use an outdated and failing userscript). It allows me to not only see if someone has posted an answer in the language I use quickly (something that can be hard when you have 3 or more pages of answers), but it also gives me a target to beat when golfing.

This is also something that uniquely fits with CGCC. Such a feature makes no sense on any other Q&A site, and would show that SE is willing to put in more effort into site-specific requests.

5
  • 10
    I think that this often goes against our stated view on inter language competition. I think that having leader-boards could possibly hurt our site if not implemented very carefully. Jul 28 at 21:14
  • 6
    I would rather have the ability to tag answers by language for easy searching and filtering.
    – xnor
    Jul 28 at 21:45
  • 5
    Answers by language with best score (in ABC order of language names, not the byte count) looks good to me. This kind of "leaderboard" has been implemented using Stack Snippets like on Hello World, but they're hard to track and maintain, and are actually broken now.
    – Bubbler
    Jul 28 at 23:09
  • There's also scope for this to cover SFF questions asking for the 'earliest' if we could work people into sticking to a particular format.
    – Pureferret
    Jul 29 at 10:38
  • 2
    One small potential advantage: This will obviously require some sort of (loose) standardized format for the language name and byte count, which means other automated tools for analyzing things like language golfiness will work better. Jul 29 at 17:08
15

A while ago I suggested a Dictionary look up in the side-bar for English Language Learners. If I had a magic wand, we would have an easy way to find credible definitions for words and idioms without leaving the site so we could start off on the same page. There are many dictionaries online, and they vary widely in quality; we discussed that a bit here: What sorts of dictionaries should we use on ELL?

We often ask questions similar to “Did you look up word in the dictionary and why didn’t that help?” This isn’t an attempt to try to short circuit questions before they’re asked, but to help us get past the language barrier and try to understand why they’re confused.

1
  • 3
    I think some other sites could benefit utilising such a sidebar for image lookup. If memory serves, a while ago Anime.SE even had an extensive (and I think quite painful) cleanup of questions that could be answered by a simple image lookup
    – gnat
    Jul 29 at 14:06
15

Encourage more community members on Software Engineering in curating minor issues in questions in favor of closing them

As an example, I regularly see good or at least acceptable questions asked, but then the asker adds something like

  • "do you know any further references for solving the problem"?

at the end. Guess what happens - the question collects close-votes, because questions for third-party resource requests are not allowed (interestingly, answers giving references to third-party resources is not only fine, but encouraged).

It gives me the impression certain members don't read the question, but are only interested in scanning them for buzzwords, casting a vote and not even try to give the asker a helpful comment. This is especially frustrating when the close-voters do have enough rep and could easily edit those buzzwords out of the question (or at least leave a comment to give the OP a chance to remove the critical part before the question gets closed).

Other cases include questions which are closed because the title does not really match the question (but could be easily fixed when someone takes the time to read and understand what the OP is really asking).

Of course, when I see such cases, I try to fix the issues by myself, but when I count the number of people willing to do so, including the diamond moderators, it seems we are pretty understaffed.

4
  • We've sometimes had similar issues (or perceived issues) on RPG.SE, with users voting to close something as off-topic for asking for RPG/tool/etc. recommendations instead of focusing on the parts of the question that are answerable/allowed on the site. (There have been quite a few conversations on Meta about that sort of potential problem.) Thankfully, the community is generally pretty good about helping users ask "stackable" questions rather than just closing them and walking away.
    – V2Blast StaffMod
    Jul 29 at 9:59
  • Ho ho, same exact thing happens all the time on Music: Practice and Theory. Good one.
    – user45266
    Aug 3 at 0:18
  • Re "certain members don't read the question, but are only interested in scanning them for buzzwords": Yes, that is a well-known problem on most Stack Exchange sites. (That excludes homework/code writing sites like Mathematics and Unix & Linux.) Aug 4 at 11:25
  • That's a general issue. If you can salvage a question (for example by removing parts) you should prefer that over simple close voting. But often that's not possible, because essential information really is missing. Editors should not have to guess.
    – Trilarion
    Aug 23 at 10:32
14

What does your community need?

MSE: Fewer off-topic questions from new users.

If I had the magic wand, I would wave it in order to instill a deep, irresistible desire for people new to a community to go and read the essentials, first thing.

I've said it before: I've seen a hell of a lot of off-topic questions on MSE and SO - across the network, actually - from new users that have disregarded the on-topic page or the help center entirely (detailed answers to any questions you might have!). After being downvoted and deleted, the authors go and post complaints or questions such as Is it necessary to read all the rules on Stack Overflow? (rene's answer is on the money there).

One of the things I would do to help with this: put a popup on the ask page (just for the newbies) that quotes a few lines from the "Don't ask" page, particularly the bit mentioning:

To prevent your question from being flagged and possibly removed, avoid asking ...

Which I personally would change to:

To prevent your question from being flagged, downvoted and possibly closed or removed entirely...

As that's mostly what I see them complaining about: being downvoted. If we warn them a bit more strongly, they might listen.

These are just a few ideas. I might need to formulate a whole new question, but there you are.

3
  • Doesn't the new ask page do this?
    – LShaver
    Jul 29 at 20:45
  • 2
    @LShaver No. It's nice, but the popup already in place lists the basic steps to asking a good question, but not the consequences involved (downvoting, closing) when you don't bother to read the on/off-topic pages. And that's what they're complaining about. Telling them what happens to bad questions before they post them, right up front, might stem the flow.
    – Ollie
    Jul 29 at 22:07
  • 1
    meta.stackexchange.com/questions/306329/… related tho MSE specific
    – Journeyman Geek Mod
    Jul 29 at 23:40
14

Force new users to take a quiz

Yes, this is fairly radical. But many sites get so many off-topic questions that it's taking the most time of all the moderation tasks. Much of that can be prevented, since many people simply didn't know their question would be ill-received.

All users that are new to the site (1 rep, ignore association bonus) or recently had a question closed get a small quiz. An open-book quiz that's so ludicrously easy that the only way to fail it is if you haven't read the rules. Failing the quiz locks you out for a minute, after that you can try again. 3 questions at most. Heck, we can start with 1, that should already do a bit.

We can't force-feed the rules into new users, but even with the new fancy modals not everyone reads what we shove in their faces. When the questions get closed, some users get really aggressive about it.

Code Review has a couple of rules. 6 of them are very easy and explained in the help center:

  • Is code included directly in my question?
  • Am I an author or maintainer of the code?
  • Is it actual code from a project rather than pseudo-code or hypothetical code?
  • To the best of my knowledge, does the code work as intended?
  • Do I want the code to be good code?
  • Do I want feedback about any or all facets of the code?

If everyone that posted would honestly answer all those with 'yes', we'd lose about 90% of all off-topic questions. This would save a lot of fresh users the frustration of getting their questions closed and make the review queues a lot less populated, which greatly helps with the improved-reviewing-experience SE's got going on for a while now.

I'd like a checklist to have people check all 6 of those, but I know that's asking too much. So I'll settle for quizzing a few of those points at random.

3
  • 8
    Or complete the (hypothetical) Stack Exchange Onboarding Game™. Jul 31 at 10:33
  • 1
    @P.Mort.-forgotClayShirky_q not sure if you're saying this in jest but IMO, we can have something like this. This can be tied to the gamification. system - you can have a short trivia-like game with questions like: "When is it appropriate to edit: a) when I want to say thanks b) when I notice a problem with the post". At the end, if the user got a good score that gets them some rep. And maybe badges This can be repeated to, say, 100 rep or so. Similar to how suggested edits give some rep but with a cap.
    – VLAZ
    Aug 2 at 8:13
  • 1
    There's something to this idea; the more I think about it the more it intrigues me. If we could incentivize them somehow to take a quiz, by giving them something they want, we might get away with it.
    – Philippe StaffMod
    Aug 4 at 8:20
13

There is a highly upvoted feature request about this: Allow non-bumping minor edits, but review them on /review.

In some instances (major retaggings, updating dead links, and correcting typos) it might be preferable that the edited posts do not clutter the front page. (To mention an example I saw recently, a community on some site might decide to split some tag into two separate tags - and the only way to deal with the existing questions is to retag them manually.1) Of course, it is important that somebody looks at the edits - but that could be resolved by a review queue for minor edits.

This was brought up on several metas. Among them, MathOverflow users seem to be one community that is rather opposed to bumping - in fact, such feature request was brought up by one of the moderators on that site. (Of course, that is just my personal impression - it is possible that there are some other communities which were quite vocal about this feature and I simply did not notice them.)

Here are some examples of posts from per-site-metas related to this issue:


1 A recent example of such action can be seen on Politics: Better defining the [polling] and [public-opinion] tags. On a site where I am the most active, I remember splitting (algebra) into (algebra-precalculus) and (abstract-algebra). The tag (algebra) was removed from 1500 questions.

2
11

Okay, here's a little pet peeve of mine for the wand sprites to look at:

Tag Reform!

I'm a Music: Practice and Theory native, and recently I've been trying to go through the site's numerous neglected tags and fix them up. I do notice symptoms of a lot of the functionality of tags going completely over the head of most users. In particular:

  • Tag synonyms are over-restricted. Currently, the requirement to be able to vote on proposed tag synonyms is a net score in that tag of 5, and to suggest tag synonyms one must additionally have the 2500 reputation required to unlock that privilege. Maybe on bigger sites this isn't as much of an issue, but on my home site, I've got the necessary 2.5k rep to suggest synonyms. The trouble is, the smaller less-used tags are the ones that often need synonyms, and it's not easy to get 5 total score in rare tags like or , which only have 4 and 3 questions respectively on all of M:P&T! I think by the time a user amasses 2.5k on a site, they should know enough to be able to at least SUGGEST synonyms, especially when they're voted upon anyway. In the meanwhile, our site has had just two synonyms created this year, and I'm pretty sure that's not a sign of our collective tagging genius.
  • Tag synonyms go relatively unused (partially because they're so restricted). At this point, they're almost a hidden feature, only accessed by those in-the-know. When was the last time YOU checked your site's Pending Approval section for tag synonyms? The description for the tag synonym privilege does provide the necessary information to use the system, but I think we'd see more users start to use this feature if the pending suggestions could have their own review queue!
  • Tag wikis are decrepit. Some of the largest, most common tags have wikis so incomplete that they wouldn't even be valid as answers! They also rarely contain any links to posts or other sources of information. I think it's great that it takes a trusted user (20k) to directly edit without suggesting, but I believe that with a more visible way to access the tag wiki (perhaps with a link in the pop-up when hovering over the tag itself, or allowing them to turn up in search results!), SE's tags would be more complete and useful for new users. On the plus side, tag wiki excerpts are working just fine across all SE, and I like how they are so accessible (cough cough).
  • Tag wiki excerpts could use a rename. It can be confusing for new users to learn that tag wikis and tag wiki excerpts are entirely different things - I remember that when I was new, I thought the wiki excerpts WERE the wiki entries (and since it's so hard to find the wiki, how could I have known the difference?). Why not call them "tag excerpts" instead?

I realize that while this is all my own experience on my specific SE site, I suspect these problems exist on other sites as well, particularly for the smaller ones. I don't know how many of these could be realistically implemented, but I at least feel good about getting these off of my chest.

Here are some more questions to look at about how the tag synonym isn't quite working well enough:

Even Meta's What is a Tag Wiki? How do I write one? answer says about the nonutility of tag wikis, "Start with the excerpt. That's the part of the tag wiki that users will see first, and the only part that most users will see at all".

Oh, and this question was where I got the searchable-tag-wiki idea from: Improving the tag wiki

1
10

A community advocate.

All communities, large and small, booming and busting, or otherwise moving around in circles all have their own pulse and their own quirks to how they work and how they run.

They also have their own challenges, trials, tribulations and triumphs.

What the communities I'm a part of seem to acutely lack is someone advocating for them on a daily basis.

And I don't mean the typical, "I'll run this back and let you know" kind of person. You could get any ol' intern to do that much. For the record, I've never found any value in those kinds of roles, either, since it usually means that the community gets to roll on its back and expose its soft, soft belly, and lay prone to the mercy of the company when the person finally gets back to them. Also by the point we hear back from them, it's a "well we talked about it and this is The Direction™ we're going to be moving in"-kind of deal.

I cannot emphasize this enough.

No thanks to that.

What I really want is someone who's dedicated themselves to that specific community for years; someone who knows the community, its ins and outs, its quirks and peculiarities, and knows how to read the room. Someone who the community can rely on and trust to get their voice heard and stand up for them when the company wants to push changes that may impact or disrupt their flow. Someone who can actually speak to why things are the way they are and break it down so that even D-level or C-level members can understand why.

Someone who isn't going to have panic attacks about talking with us. Someone who is willing to listen with an open mind and is committed to being transparent. Someone who actively cares about the problems we have and wants to see to it that our problems are solved. Someone we can put our trust into because they would have already built that rapport with us.

30
  • 4
    In theory that should be the CMs - least on the company's end, unless you mean a specific 'new' role. Even then - that's a lot of communities with different needs, and even on our end, there's very few folks with super wide knowledge and contacts to be able to do that imo.
    – Journeyman Geek Mod
    Jul 29 at 15:15
  • 4
    We already have advocates. It's the people willing to speak up when given the chance to. Reducing that... "role" down to single person would result in one person that only represents a minority rather than the broad range of people who participate.
    – Kevin B
    Jul 29 at 15:20
  • @JourneymanGeek: In theory, yes. In practice, no.
    – Makoto
    Jul 29 at 15:22
  • 4
    I’m not sure I understand why this role isn’t handled by the mod team. Maybe it could be a more formal aspect of what they do, but who better to represent a community’s interests than someone elected by it?
    – ColleenV
    Jul 29 at 15:23
  • @KevinB: The point I'm making is that we - as in the community - can only speak so loud about certain issues. The real problem is that we are often ignored or sidelined in the conversations when Big Decisions are made that have a direct impact to a community, or they're made as proclamations from above. Yes, we can shout about that as much as we like. Does that mean we're being advocated for? Nope!
    – Makoto
    Jul 29 at 15:23
  • 3
    @ColleenV: I really don't think of the diamond mods as advocates of the community to the company. I think they're out there representing us and our best interests, but I don't have the impression that they talk to the company at the detailed level that I want. You see, they can say a lot of things and point to a lot of things, and the person that they're saying this to or pointing this out to isn't in context or has to absorb all of this and then run it back. Remember what I said about "running it back"?
    – Makoto
    Jul 29 at 15:25
  • 3
    Mods aren't elected for their ability to represent us.
    – Kevin B
    Jul 29 at 15:26
  • 1
    @Makoto It might make sense to have the community specifically choose an advocate, because a lot of mods don’t take advantage of the channels mods have to advocate for their community. There is a “team” and a chat room where mods can talk directly to staff and to each other, and the mods who participate there do impact policy in a lot of instances.
    – ColleenV
    Jul 29 at 15:29
  • @ColleenV: I'd really rather it be an organic hiring process. Someone that the company identifies and recognizes as someone who is capable of helping and leading the community, and is willing to work for the company as well. It's happened before; no reason why it couldn't happen again - or at least more often.
    – Makoto
    Jul 29 at 16:41
  • Well, you’ll never know the answer if you don’t ask. I find it doubtful the company would hire advocates for each site on top of the CM team. It seems unlikely there would be enough work to keep them busy or that they would provide enough value to the company as a single site advocate. Personally, I wouldn’t want my advocate being paid by the company. It seems like that would put their interests more in line with the company than the site.
    – ColleenV
    Jul 29 at 16:46
  • 1
    And what exactly would they do all day? Not trying to be snarky, just trying to understand.
    – ColleenV
    Jul 29 at 17:02
  • 3
    It just doesn’t seem like a full time job to me. Some sites go days without a post on their meta site. Some sites have no issues with the company they need advocates for. I agree that we could use a better way of communicating concerns to the company and getting feedback that we were heard and understood. I’m just not sure that a paid advocate is what we need. I’m an American though, so I probably have a cultural bias against that solution. I’m more inclined toward “the voice of the people” type stuff.
    – ColleenV
    Jul 29 at 17:53
  • 1
    I don’t mean to be a pest. Like I said, if you don’t ask, you’ll never know the answer. I am just bad at resisting any opportunity to tell people my opinion ;)
    – ColleenV
    Jul 29 at 19:28
  • 1
    @ColleenV In a sense - it can be. If you went out to all, I donno, 180? Communities, I'm sure that a significant part of them would be happy to have someone listen and having a person you know can reach out to and will pass on the message to the right folks. While that was roughly a aspiration of the mod council, balancing another job and that sorta thing is pretty hard. I just see that as a core part of CMing as it should be.
    – Journeyman Geek Mod
    Jul 30 at 13:40
  • 2
    @ColeValleyGirl: You gravely understate the scope of this request. Mods in my understanding are still go-betweens with the company and the community. This is a radical shift in thinking in that we would actually want someone who's full time job it is to be a part of a community to understand and advocate exactly for what's needed. Diamond mods aren't in the position to advocate to the company on our behalf since they're cut off from a lot of those kinds of internal conversations that could or do have an impact on communities until the company is just about ready to roll a change out.
    – Makoto
    Jul 30 at 16:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .