45

Update - April 21, 2025

The results of the March 2025 Community Asks Sprint have been posted here!


Update - April 3, 2025

Last week, we completed our Community Asks Sprint! There are still some outstanding issues that were started during the sprint but not quite finished, so we're taking the time needed to finish those off. We plan to summarize the results of the sprint next week, and that will include everything we were able to get to during this sprint. Will update soon!


Our third Community Asks Sprint starts next week! If you’re not familiar, the Community Asks Sprint is dedicated time for all of our public-facing developer teams to focus on requests the community has made regarding what issues to address and what changes to make on the platform. Learn more about the first and second sprint.

As we announced in December 2024, this sprint will last two weeks rather than just one, from March 17th to March 28th. This gives our developers much-needed runway to focus on your requests compared to what we had during the previous one-week sprints.

Here are some of the issues we’re planning to work on for this sprint. Keep in mind, these plans aren’t static; the teams may choose to focus on different issues from these.

Once this sprint is completed, we’ll share with you the items the team was able to tackle during this time. We’ll also spend some time evaluating how it went. Critically, we’d like to know whether the extra time let us address more issues and/or more complex issues. These answers will impact how we plan future Community Asks Sprints, particularly their length and frequency, which we’ll keep you informed on. Whatever the outcome, we remain committed to a minimum of four weeks per year of Community Asks Sprints.


With all that said, we’d love to hear from you- What issues are on your mind? What would you like to see us working on this sprint or next? While we’ve come prepared with a list of issues you’ve asked us to address already, we want to know what you consider most important and want to see addressed. We’ll take your answers into consideration alongside those we’ve identified already.

13
  • 7
    I wonder what happens between sprints.. nothing? Sorry for being grumpy and.. ah, I just accept it, I am grumpy! So what you were busy with from 2nd March (date of newest status-completed) and what will you do then until Sprint? Oh, it's a wrong site, my bad, still not impressive.
    – Sinatr
    Commented Mar 12 at 16:53
  • 7
    They walk. Sometimes they Jog. Commented Mar 13 at 16:30
  • 4
    Btw, should I downvote all other posts here to have mine prioritized? Is it a war and I must win or are we friends?
    – Sinatr
    Commented Mar 14 at 16:44
  • 1
    @Sinatr regardless of the answer to the second question, how much difference do you expect doing that to make for your purposes? for me, it looks not enough to matter, and you should consider answering the second question yourself. you really could have just asked "how do you pick what to do and prioritize", which is answered at a high-level in the post announcing community asks sprints.
    – starball
    Commented Mar 14 at 17:44
  • @starball, well, my question will be on top, others on bottom. The devs obviously pick up the most upvoted (more important) things to do, as they always do.. right? My issues are more important to me and I am sorry for other people issues, but they should be nice and wait .
    – Sinatr
    Commented Mar 14 at 17:47
  • 8
    "alright everyone, be nice now. line up behind sinatr". | "And users - don’t forget to vote! While we won’t ever rely solely on voting to determine our priorities, votes are a critical signal that lets us evaluate how much you want us to implement a feature on the network." - JNat
    – starball
    Commented Mar 14 at 17:51
  • 1
    @Sinatr well, I'm guessing you too produce code for a living, and as you can see yourself between the 6th March and the 12th March there were maybe 6 new questions/general remarks relevant for the dev team. I don't know how many bugs you solve per day in your job but in any case I'm worried if you think this is a relevant single metric to judge a dev team Commented Mar 14 at 18:38
  • 1
    @globglogabgalab, SE has over 250 employees. I wonder how many are devs and how many of them are still supporting that "old ugly web-site from which company start in 2009". If it's one or two, then metrics are ok. =D
    – Sinatr
    Commented Mar 17 at 10:41
  • @Sinatr probably they have one or two people dedicated to Q&A bug reports, but during this sprint the number likely grows to maybe half a dozen, if we're lucky, and they only work on things chosen from the list here.
    – TylerH
    Commented Mar 17 at 15:25
  • 5
    Of course, it would be nice if, for a few sprints, they just didn't accept new things and instead focused on burning down the back-log here. We're now several sprints in and still seeing numerous requests in this sprint that have been requested every single sprint to date.
    – TylerH
    Commented Mar 17 at 15:26
  • Maybe a late suggestion but meta.stackexchange.com/questions/383520/…
    – VLAZ
    Commented Mar 27 at 13:39
  • 6
    Is sprint over? What has been done?
    – Sinatr
    Commented Mar 31 at 10:12
  • 2
    Followup post: Results of the March 2025 Community Asks Sprint
    – V2Blast
    Commented Apr 21 at 17:10

29 Answers 29

42

I am (once again) asking you for proper notification for flags/flag responses and suggested-edits/rejections.

While the linked posts provide more detailed explanations, here is a TL;DR:

  • flag responses should be in the inbox because they contain information the flag raiser needs to know, like in this case where a direct question had been raised in the flag: mod responding to flag
  • suggested-edits rejections should be a notification. Currently, you only realize if you made an error in editing when doing another unrelated edit, in the form of a small banner above the edit box. That design is bad, as this is what the editor sees: post editor, with the area the user focuses on Notifying a user about a rejected edit (and the reason for it) will NOT discourage him from editing. If that were the case you need to remove downvote notifications as well. Generally notifications drive up engagement (that's why Facebook added the "poke" function after all). While at it please consider custom rejection reasons.
3
  • 5
  • 4
    I fully support this; it's a great suggestion. Get all the low hanging fruit! 🍊🍎🍐🍇🍋 Commented Mar 26 at 23:51
  • 1
    I'm mostly upvoting this for the former (which would be really helpful as a way to communicate with flaggers), as the latter is arguable either way (against: the right time to inform users of an edit rejection is when they're going to do another edit, so they don't make the same mistake; for: the user may wish to try the edit again, bearing in mind the feedback they received), and I think the banner was recently made harder to miss (I can't immediately find the post on the topic). But I'm not opposed to it.
    – Ryan M
    Commented Mar 28 at 16:02
39

At time of writing, the underlying request has now been marked as .

Just like the original highlighter update, it appears that posts were not "re-baked"– this means that broken formatting is frozen for posts that currently show it, and that proper highlighting won't be restored without an edit to force the post to rerender.

But it's fixed!

lang-javascript
console.log('Hello highlighter!');
lang-js
console.log('Hello highlighter!');

Original request below.


Syntax highlighting has been broken for going on five months now, since the change was made to address Stop guessing/auto-detecting a language when you KNOW it will be incorrect.
Can this get thrown on the agenda?

In short, the behavior of the highlighter in rendered posts now differs from its behavior in the legacy (non-Stacks) editor's post preview, which is wildly confusing for everyone involved. The change also broke a large amount of language aliases outright, so things like lang-js work, but lang-javascript do didsee update above not:

The JS code block works, but the "JavaScript" one does not and appears unhighlighted

Despite showing up as working in the post preview:

Both of the above code blocks highlighted correctly according to their JavaScript syntax

1
  • 2
    This is especially important with the wider rollout of the Stacks Editor.
    – Ryan M
    Commented Mar 28 at 16:03
35

In the spirit of chat getting attention again - may I once again ask for chat and meta on the left sidebar?

It feels an essential step that'd help other work on chat meet its full potential, and restores awareness of community tools

12
  • 10
    It's like you read our mind!
    – Piper Staff
    Commented Mar 12 at 14:15
  • 4
    Keep in mind that most people have probably opted out of the left sidebar since historically it has just contained useless clutter (companies, discussions, collectives teams etc etc). Maybe is we move all clutter from the left panel to a top bar icon, we can use the left sidebar for actual useful stuff like meta + announcements and chat.
    – Lundin
    Commented Mar 13 at 15:01
  • I mean, I'm not on SO, so for me, its mostly that vestigal teams link, from where I'm sitting its just underutilised. Commented Mar 13 at 15:08
  • 2
    @JourneymanGeek curious to know: Imagine, chat is added to the left nav today and gets a lot of visibility. What would you be excited about? What would you be concerned about?
    – EmmaBee Staff
    Commented Mar 13 at 21:39
  • 5
    On the short term - moderation, onboarding, and 'stickiness' - We both want to encourage new users to use chat, and where possible retain the culture and character of chat. Lots of chats are quiet or lack active moderation, so a little heads up and trying to activate folks to come in and moderate/mentor would be helpful. That said, my vision here isn't immediate revitalisation, its people trickling in, and hopefully staying to build that critical mass. Commented Mar 13 at 23:17
  • People who opted out of left side bar probably know about chat and meta. @Lundin
    – M--
    Commented Mar 14 at 6:43
  • @M-- It's not about if they know about it but about making a convenient GUI where useful features are easier to access than useless ones.
    – Lundin
    Commented Mar 14 at 12:40
  • @Lundin Right now the top bar is mostly icons - switching it to text would be nice but in the current design, it makes sense. And more people would be using the sidebar if it was useful - and most people keep the default. On its own, the FR is more for less experienced users or ones less familiar with meta and chat IMO. Bringing back 'old' users to chat needs a bit more work ;) Commented Mar 14 at 12:47
  • @JourneymanGeek I rather envisioned things like Collectives and Discussions to be intuitively found underneath a 💩 icon. But a monochrome one of course.
    – Lundin
    Commented Mar 14 at 12:56
  • Discussions seem easy enough to find- even by spammers Commented Mar 14 at 13:20
  • @Piper - I agree with J.G. that Chat and Meta are the key 'missing' links on the left nav, but a while ago I also raised that the left nav could have so much more functionality - so if this is something you're considering, I hope it's not the only two links that will be added :)
    – Robotnik
    Commented Mar 17 at 5:51
  • @JourneymanGeek Exactly, so it is taking up space at the tight panel where we could place something useful like chat instead. And relocate Discussions to some obscure sub-menu in the middle of nowhere.
    – Lundin
    Commented Mar 20 at 8:30
30
  1. It's way past time to reward the duplicate finders.*

    Finding duplicates is an underappreciated task. Let's change that.

  2. Award Marshal badge multiple times.**

    Same as reviews (that a user can get multiple badges for), we need users to constantly edit and flag problematic content. Let's reward them for that (ironically, the linked FR is a duplicate; see request #1).

  3. A while back I asked for Recognized badge to be "removable" in Collectives. Implementing this in a more generalized way can be helpful: letting moderators/staff/recognized-members to decide whether include their badge when answering on the main site. When a moderator is answering a question on the main site, they are not answering as a mod (meta sites are obviously different). Having the ability to drop the badge/diamond is actually helpful.


* I asked for this during previous sprints: https://meta.stackexchange.com/a/402987

** Copy-Editor badge should also be awarded multiple times.

22

Add search to moderator tooling. Some key moderator lists lack any ability to search text. This includes

  • Open moderator flags
  • Moderator messages
  • Community Team escalations
  • User annotations

Somewhat related is that the Users page (/admin/users) loads a ridiculous amount of data. This is the logical place to add search, but it loads

  • Users with Flagged Posts (all time) - Not necessary for every load
  • Users with Flagged Posts (last 30 days) - Slightly more useful, but does not need to load every time
  • Latest User Messages - Filterable for messages you sent and that's it
  • Latest User Annotations - Also self-filterable
  • Latest CM-Team Escalations - Also self-filterable

Self-filtering was a huge boon, but that's only half the story there. What should happen here is that the data is all loaded via AJAX calls. I tinkered with a userscript (needs some refinement) that simply broke the page up so you could choose what data you wanted and it would only load that data (as with many mod tools, it returns raw HTML, so it wasn't hard to modify). Honestly, this seems like low-hanging fruit.

1
  • 1
    I also have a userscript for that users page which does the same basic thing of splitting up that massive block into individual tabs that also have url parameters to be able to link to and find specific pages and information. Commented Mar 11 at 19:26
18

Pending getting a roadmap again - would it be possible to give a rough overview on planned areas of interest for the next (or even the one after) sprint ahead of time, so we can collate and try to align areas of concern for the next sprint?

1
16

Some things I'd like to see to combat spam

Why?

Some spammers don't bother posting spam posts but rather just add spam to their profile. They do get indexed via unlocking the autobiographer badges. That is more effective, as unless you search for those accounts, e.g. like this, you wont see them on the network itself.

user list of buy-XYZ-accounts spam


As always: If anyone wants to help with the spam on SU, there is a chat for reporting spam accounts on Super User

2
  • 1
    are these official ads or just some dude who likes cash app and google voice? Commented Mar 21 at 21:28
  • @TwineeeThePickleWizard its pretty much just scams and stolen/"hacked" accounts being sold
    – A-Tech
    Commented Mar 23 at 18:33
15

In sprint 1 Allow bountied questions to be closed by regular users was addressed. However, it is now possible for somebody to post a bounty on a question, the question gets closed with no answers, and that effectively burns the rep.

I propose figuring out some refund policy for these situations. It may be "no refunds" as long as it is clear.

10
  • 5
    From a mod perspective - that's going to be interesting. So far, we've been no refunds but that policy was based on the partial assumption users might use bounties to tactically keep questions open. On one hand, it feels fair, on the other, we'd rather not have people bounty questions that should be closed (and possibly demand a refund). A mandatory, automatic (partial?) refund feels the best option if we're refunding Commented Mar 11 at 14:22
  • 1
    @JourneymanGeek perhaps a % of the bounty, related to the percent of the duration it was open + bounties for? That way you get what you “paid” for. E.g. if the bounty would have featured it for 48 hrs and it was closed 12 hrs later you get 75% back? Or maybe a more complex depreciation curve Commented Mar 11 at 14:36
  • 1
    @JourneymanGeek I have a weak preference to a refund only if it's closed shortly after adding the bounty. And it can be a partial refund. But I personally don't care much. I just know this came up after the implementation and it was never addressed, however, it probably should have been considered when the closing bountied questions was added.
    – VLAZ
    Commented Mar 11 at 14:38
  • 8
    I feel like this might be a good question on its own ;) Commented Mar 11 at 14:39
  • 2
    re: "as long as it's clear" - isn't it, though? help center says: "All bounties are paid for up front and non-refundable under any circumstances."
    – starball
    Commented Mar 11 at 17:11
  • 1
    @starball that's from when bounties could attract answers. Now there is a chance a bounty is paid up and then it's impossible to attract an answer. Bounties have always been likened to paying for advertisement. But it's now a broken analogy, because you can pay for an ad that never airs.
    – VLAZ
    Commented Mar 11 at 17:14
  • @VLAZ Meh, if someone wants the bounty from a closed question they might as well edit it into shape. The ad very much airs, it just takes a bit more work to buy the thing. Commented Mar 12 at 7:14
  • 1
    I think it’s true that a revised or at least clear policy is needed, but this doesn’t seem appropriate for a sprint. It might be better to discuss this first before implementing something. Commented Mar 12 at 7:17
  • Isn't the bounty refunded if the question is deleted? So, the refund policy can be this: if the question is closed with no answers and is downvoted, then roomba will delete the question and refunds the bounty. Maybe I am missing something:/
    – M--
    Commented Mar 20 at 2:28
  • @M-- the bounty might lapse before the question is deleted. Various scenarios for that to happen, like the question having upvotes, or having a reopen vote which will disqualify it from auto-deletion until the reopen vote lapses, etc. It doesn't sound like a universal solution. It still could be the policy, if the company chooses that. I just don't think it's the silver bullet for this issue.
    – VLAZ
    Commented Mar 20 at 5:24
15

This is less of an FR than a request for clarification - there's multiple suggested solutions for re-add social media links to posts and user profiles. It might be a good idea to let us know (and discuss?) which solution y'all are going towards

5
  • 13
    This is still tentative, but our last thinking was to implement the third solution in that post, offering multiple generic fields.
    – Frog StaffMod
    Commented Mar 11 at 16:33
  • @Frog that's awesome to hear! If you do go down that route, could you still support some icons for the "common" urls? (X, Reddit, FB, Insta, Mastodon, BlueSky, etc).
    – Robotnik
    Commented Mar 17 at 5:58
  • @Robotnik Or, maybe just set up a simple script that finds favicon.ico and displays that? Commented Mar 21 at 21:26
  • @TwineeeThePickleWizard not if you want them to match the style of the site. Icon packs exist for that reason, for example: here's Material UI's
    – Robotnik
    Commented Mar 22 at 2:53
  • @Robotnik for popular things like Facebook and X maybe, but if you don’t know, then that would be the best option. Commented Mar 26 at 1:02
14

What issues are on your mind? What would you like to see us working on this sprint or next?

As highlighted in this question nearly 4 years ago, Area 51 exposes our email addresses on a silver platter. Almost. (Well, sorta)

From 2013 to 2023, Stack Exchange had been salting our email addresses prior to hashing them, before sending them off to Gravatar to be, in most cases, transformed into an Identicon. From 2023 onwards, they've actually stopped using our email addresses altogether.

However, Area 51 still operates on a 2011 fork of the SE engine, which means that both old and newly created Gravatar avatars continue to use the unsalted hash of our email addresses. For instance, my Identicon on Area 51 can be traced back to: https://www.gravatar.com/avatar/a454773f0a95c0855e768b7c8be13e0b?s=128&d=identicon&r=PG. Here, a454773f0a95c0855e768b7c8be13e0b is simply the raw MD5 hash of my email address. This issue affects all users on Area 51. (If you don't believe me, test it yourself!)

As I'm sure many of you know, MD5 has been widely considered "cryptographically broken" for a long time. It isn't a particularly strong hash in this day and age, and quite easy to brute-force. I have no doubt that among the thousands of email addresses associated with Area 51, at least one could be found in a wordlist or potentially be brute-forced. This has probably occurred in the past, where a user's email address was (mostly likely) cracked from its MD5 hash by another user (back when the hashes were included in the data dumps).

Despite raising this concern four years ago, the status remains . Please, oh please, can something be done about this serious privacy risk?

5
  • If anyone with an Area 51 account gives me permission, I can attempt to crack your email address to prove my point, and anyone can feel free to attempt to crack mine, by the way - I'll be impressed :) Commented Mar 12 at 0:39
  • 1
    I'll try later... Commented Mar 12 at 15:50
  • 2
    @security_paranoid, go ahead -- and if you manage it, send me an email, rather than posting here.
    – Mark
    Commented Mar 13 at 23:52
  • @Mark you don't appear to have a '51 account ;) -- the rest of the network is secure. Commented Mar 14 at 0:04
  • 5
    @security_paranoid, I do have an Area 51 account, but I've got it hidden from my network profile. Try area51.stackexchange.com/users/109290/mark
    – Mark
    Commented Mar 14 at 1:28
13

I'm seeing some profile related work - and I'm on a bit of a roll on my wishlist so - I'd like to bring up profile spam

More specifically - could we hide or otherwise de-index profiles from users with no interactions with the site from search engines or unregistered users?

2
  • 3
    I thought this was already the case? i.e. profiles under a rep threshold aren't indexed?
    – Robotnik
    Commented Mar 17 at 5:59
  • some tests indicate they are, but YMMV I guess Commented Mar 17 at 6:10
12

Can site moderators please get a notification when a question is put on the Hot Network Questions list? Ideally as a setting individual moderators could disable. Or maybe that they could enable. I don't really care too much what the default would be, as long as it can happen.

1
10

Please enable support for the <details>/<summary> HTML tags. I had asked for this in a previous sprint). The original, declined feature request at Please add the ability to fold blocks of code in questions and answers which has been re-asked in a more general form at Please, I want some more HTML. The specific posts for <details>/<summary> are quite well received (+27/-1 for the older one and +17/-1 for the newer one at the time of writing).

10

I'm probably late to the party here, but if I can squeeze this in:

Please remind me when I am wielding the dupe hammer

Having some sort of client-side check to see if the user's vote will unilaterally close (or reopen!) the question as a duplicate in this situation feels overdue.

10

Update: The original bug report has been marked by a staff member.


I'd like to (once again) raise the discrepancy in link validation between regular links and image links, introduced around the time of the Imgur migration. Protocol-relative URLs to embedded images are incorrectly being blocked.

If you try to post an image with a protocol relative URL, you get the following validation error preventing you from submitting:

all URLs must start with https

This is a recent development, likely changed as a result of the Imgur migration. For non-image URLs, protocol relative links are still supported:

Regular links ✅ (passes validation)

[This is a text link][some-link]

  [some-link]: //stackexchange.com

Image links ❌ (fails validation)

![This is an image][some-img]

  [some-img]: //i.sstatic.net/f7Uili6t.png

I raise it here because it's probably a minor fix and something that could fit within a sprint.


How to resolve (definition of done)

I'd prefer protocol-relative be allowed again for images, but ultimately it's the discrepancy that needs to be resolved - either we support it for all link types, or we don't support it at all.

In other words, the validation that checks links should be the same for both image and regular links.

9

What issues are on your mind? What would you like to see us working on this sprint or next?

8

Here are some of the issues we’re planning to work on for this sprint. Keep in mind, these plans aren’t static; the teams may choose to focus on different issues from these.

This issue came upon after Inbox notifications not consistently marked as read after clicking title was "fixed" in sprint 1.

I propose you at least go back to the initial rework of the inbox which made it a point to allow almost all types of clicks to mark the notification as read. That is: left click, middle click, Ctrl+click. Only right click -> Open in new tab was not covered. The breaking change "fix" removed these and only allowed clicking the tiny envelope button as marking the message read (without reloading).

I would also be happy if you expand it to work with right click -> Open in new tab.

However, I hope we do not have to talk about having to fix this again in future sprints. It is getting ridiculous. I would like to see the platform moving forward, not failing to address a relatively minor issue for so long. Particularly, we can probably start addressing other issues from the initial inbox rework.

I almost forgot - once the marking as read is done, then we can maybe also actually address Inbox notifications not consistently marked as read after clicking title because it is still not consistent. It is more rare but the solution that was implemented fails to deliver. Although, it will me a much much rarer issue if I do not have to click and load to each notification multiple times.

1
  • 2
    Yeah - I would hope that a fix addresses all situations. I don't even have "middle click" on my Mac but all forms of "open in new tab" should mark the item as read - two finger click (multi-touch devices), press and hold (touchscreens)... etc.
    – Catija
    Commented Mar 12 at 11:49
8

First, I'm really glad you're fixing Inbox is no longer immediately marking item as read when middle-clicking to open in new tab. It's a huge, frequent paper cut for me.

Other frequent paper cuts I'd love to see addressed are:

8

Looking at existing issues marked "[status-review] or [status-deferred]" (sorted by votes) would be most appreciated.

A tiny list of the top issues:

  1. "Add the ability to ignore users", asked Jul 7, 2009 - 292 votes - 34 answers - 8k views [status-deferred]

  2. "Offer an XMPP method for chat", asked Jul 15, 2010 - 210 votes - 5 answers - 5k views [status-deferred]

  3. "Can we have a way to edit bounty custom message? " - asked Nov 15, 2011 - 170 votes - 3 answers - 2k views [status-review]

  4. "People don't read the tag excerpts because we don't swat them into their faces", asked May 7, 2015 - 160 votes - 2 answers - 2k views [status-review]

There's 397 questions on that list, perhaps some not actionable and some already resolved.

It would be kind to work on getting that down to nothing; even if the outcome must be: [status-declined] or [status-wontfix], but preferably [status-completed] or [status-planned].

Work has already gone into writing the question, posting feedback as answers, and staff altering the status to deferred or review, let's not throw out valid requests by looking at new things to do until we've already worked out what to do with popular years-old requests.

3
  • 4
    does anyone even use XMPP any more? Commented Mar 12 at 10:50
  • 2
    @JourneymanGeek apparently so; because it's multi clients/OS/protocol, including SMS (or even email for you) :), it's fairly standardized. XMPP is not my request, if it's declined then the list is that much shorter. My feedback isn't support for everything on the list, it's finish what's on your plate before diving in for seconds. --- Not sure why this didn't score better, because none of those requests were mine and were wanted at the time by others; we can close that question if no one wants anything there.
    – Rob
    Commented Mar 12 at 20:47
  • 3
    XMPP is a convenient protocol to chat without requiring your friend/colleague/end user to use a given app/website edited by Google or Meta, and it's easy to fine-tune it to your needs, although I reckon it's a bit nerdy Commented Mar 14 at 18:49
6

I really hate idea of putting an empty cup with text "put here your requests now!" and completely ignoring a two full containers nearby. Even a lottery where from all feature-requests and bug-reports you pick a few is more fair.

But ok. In jail we eat what we get.

Just today I needed 2 things and they are not here ... yet.

Notification when comment is deleted - I suspect one of my comments under interesting question was removed, I think what I've written was important, but maybe it sounded rude? or what could be the deletion reason? And I can't decide what to do: re-post same comment, re-post more "nice" comment, do nothing or go to meta and ask why my comment was deleted? Maybe I am just old and simply forgot to click "add comment"?

Improve "Related" algorithm - or rather remove that stupid list of totally unrelated (ok, barely related) questions on the right side. Isn't it a place, where AI-vibe could help?

6
  • 3
    I agree with the opening paragraph, but I disagree w/ the idea that we need notifications whenever a comment is deleted. At most, showing us our deleted comments on our user profile's 'comments' section would be sufficient. I do NOT want a notification any time a comment of mine is deleted. This would require a lot of extra work to figure out when to notify users (e.g. deletion due to question or answer deletion?) and would mean a lot more work for mods due to abuse potential (a troll or argumentative user notified their comment was deleted may be inclined to go re-post it, or go on a rampage)
    – TylerH
    Commented Mar 17 at 14:34
  • 1
    (continued) or go complain on Meta about 'moderators censoring my comments', which is entirely unnecessary. Finally, while I agree w/ the desire to improve the 'related' algorithm, I think that's way too much effort for a two week sprint, especially given the level of domain knowledge the current dev team has of the Q&A platform.
    – TylerH
    Commented Mar 17 at 14:35
  • 1
    I second the "view deleted comments from your profile" raised by @TylerH. I would also like to raise showing your deleted comments in-context. So if I view an answer that I've commented on, show me that deleted comment (with a red background), but not deleted comments from other users.
    – Robotnik
    Commented Mar 17 at 22:38
  • I liked the last idea. Answering robots? should be left to the humans. Finding duplicates? with little human moderation, yes, this should work. Commented Mar 21 at 21:25
  • Definitely +1 for the related questions idea. AI should be used where it's logical to use. Current algorithm can find some duplicates, but it fails to do so consistenly. Well-trained AI would do a much better job.
    – Destroy666
    Commented Mar 23 at 19:55
5

What issues are on your mind? What would you like to see us working on this sprint or next?

Address the issues from trying to fix How to find a question by its old tag when that tag is synonymized with another back in sprint 2.

5

It would be nice to fix an issue with the API:

The previous issue was fixed after 4 months:

Hopefully these bugs can be prioritized so that the API is not broken for months.

1
  • While I agree w/ the effort to fix API bugs, I hope they are not prioritized over other things that affect ALL users (e.g. UI/system issues).
    – TylerH
    Commented Mar 17 at 14:36
5

It would be nice if an option could be added to the API to mark inbox items as read, see e.g. this feature request:

Marking inbox items as read in api v2.2

That's the only missing piece to being able to build one's own inbox.

5
  • ooh, or a inbox notifier of some flavour? Commented Mar 14 at 16:42
  • @JourneymanGeek ... would be nice to have, wouldn't it? Commented Mar 14 at 16:45
  • Notifications of some flavour, especially on mobile is on my wishlist Commented Mar 14 at 16:49
  • Also, arn't we at 2.3 now? Commented Mar 14 at 22:52
  • @JourneymanGeek Yes, I hope one can understand the feature request as "please add the feature to the current version, not to the version which was current when the feature request was opened" Commented Mar 14 at 23:01
5

The flag/close dialog has several serious issues. The easiest way to see these is to review the responses to Question Close Updates, most of which have been unacknowledged (publicly) by staff.

Issues include:

  • The flagging dialog should not put all the custom close reasons under the heading of "needs improvement"
  • The "needs more focus" reason is not only about posts that ask multiple questions. It is also about questions that are too broad and which it is unreasonable to expect to be answered in the SE Q&A format.
  • When you've voted to close with a site-specific close reason, you can't see which one you chose at a later time

Please consider implementing Machavity's very clear proposal.

1
  • 1
    I'd also add to that that the 5 migration sites limitation is nonsensical in close reasons. It's 2025, we have scrollbars. I'd increase it to 10 at least, considering how many SE sites there are that are more suitable for various questions. At least let mods of specific SE sites decide if it's worth to include any more sites there.
    – Destroy666
    Commented Mar 23 at 19:42
5

I really hope Ability to change close reason before it gets closed gets covered at some point. Punishment of no close reason at all or wrong close reason when your mouse slips doesn't serve any purpose.

You could in theory say that it's to prevent too hasty close votes, but I don't think it achieves much in that regard. Plus there are way better possible preventions that aren't as limiting - e.g. just lowering the possible frequency of switches with reasonably short timeouts and what not.

1
  • 1
    Also, there are a lot of cases where the question is, say, unclear at first read. So you vote for that. But then you realise it's actually (very roundabout) description of a duplicate. But you can't change the vote. Another alternative is that it initially matches one close reason but after edits, it matches another. Often something like 1. Unclear/Unfocused 2. Duplicate. Or (at least on SO) the clarification might be that it's a typo/not reproducible.
    – VLAZ
    Commented Mar 24 at 16:58
4

What issues are on your mind? What would you like to see us working on this sprint or next?

Allow filtering of flags by sub-type in addition to their status It's rather pointless letting a moderator write me a personalised message in response to a flag if that flag is down on page 8 or further by the time it's handled. Please let me either filter down on specific flags ("custom" the most relevant) or those with a custom moderator reply.

This could go in tandem with the request made by A-Tech to generate some form of notification on a flag response.

1
4

I use saves a lot and would like to see the search operator for specific saves list implemented. I don't think the saved list feature is complete without the ability to search within them. This has also been brought up on this meta SE answer. Both the question and the meta SE answer have been highly upvoted.

0
2

What issues are on your mind? What would you like to see us working on this sprint or next?

Since the number of questions is nosediving, I suggest the following:

11
  • 2
    What is the benefit for the OP or either site to migrate an old question?
    – Catija
    Commented Mar 12 at 15:05
  • @Catija Same benefits as migrating a recent question: avoiding question closure/deletion (which in turns avoid 404 and allow new answers to be posted) and help site be better organized. Commented Mar 12 at 17:08
  • 3
    I just have a difficult time seeing that as a high-value change and, instead, risks sending low-value, off topic questions to a site where the OP is likely not even around any more to care about what happens to or try and fix the question. If it's really that stellar of a question, a flag for a mod to migrate should be fine in the minority of cases where migration is even warranted. Remember - even if the 60 days goes away, it still limits regular users on a site to only having up to five target sites - most sites barely have 1-2 targets, let alone 5.
    – Catija
    Commented Mar 12 at 17:30
  • @Catija OP is just one of hundreds/thousands of readers. I always view question closure/deletion as a big loss in learning opportunity for any reader (unless for crap question, of course). Eg on SO, many questions that could be on SU are instead closed/deleted. Commented Mar 12 at 17:43
  • 3
    I wasn't thinking only about the asker. You've linked to the post where the policy was changed to prevent questions over 60 days old from being migrated... so there's no info about why you think it should be overturned - is that the correct link? Your other two links go to actual requests to change the policy. Why should an unanswered question be "saved" in the way you're saying? Should the migration option only exist for answered questions? The migration system is incredibly broken and it needs help but I sincerely doubt there's a high volume of valuable questions that mods couldn't handle.
    – Catija
    Commented Mar 12 at 18:38
  • @Catija It's the correct link but linking to some FR would be better indeed. "Why should an unanswered question be "saved" in the way you're saying?" Others may have the same question. I agree it's lower priority, relaxing roomba is much higher priority imho. I just thought changing 60 to 365 to higher would take a second. Commented Mar 12 at 18:57
  • Imo, the best action for off topic questions that are old/to be deleted, but where the deletion would be detrimental to the community as a whole, is to historic-lock them.
    – Robotnik
    Commented Mar 17 at 11:12
  • We need more questions deleted, not fewer. We also don't need chaos in closing off-topic questions. There are ways to improve the question closure/review process (one major way to improve the amount of attention here is to have fewer questions, hint hint), but adding a way to 'cancel' pending votes is not it. I would give this a -2 if I could for those two suggestions. I support the idea of expanding the migration window (both in time and in number of sites we can pick from), though.
    – TylerH
    Commented Mar 17 at 15:22
  • @Robotnik That's already the process for off-topic questions that would be detrimental to lose.
    – TylerH
    Commented Mar 17 at 15:23
  • @Catija saying that migration system is incredibly broken is an understatement. Even < 60 days questions often are preferably closed than migrated. People fail to understand how bad leaving off-topic questions is for the network. Not deleting them isn't a solution, but also not migrating them isn't one either. Let's take ffmpeg as an example. Googling ffmpeg problem results in 90+% of cases in SO questions being the top Google results. Do you know what this does? It ingrains in users' memory that this is the website to ask ffmpeg questions, not SU or video SE.
    – Destroy666
    Commented Mar 23 at 19:21
  • Small notification with close reason is not nearly enough good of a prevention, most people that don't use the site will misunderstand/ignore it. Then there's also the problem of new questions getting answers before getting closed and potentially becoming next targets for high upvotes as the voting capabilities aren't removed. It's a complete mess that wasn't managed well in the 1st place and is becoming worse and worse + will become even worse the more additional backwards-incompatible close reasons are added, without properly solving the problem.
    – Destroy666
    Commented Mar 23 at 19:23
-1

Very happy to see another Community Asks Sprint! I'd like to see some form of community comment moderation implemented. There have been some requests in the past to pull inspiration from:

I'd also like to see you guys implement some or all of the many feature-requests which have been made to try to help with massive spam waves on Super User and elsewhere.

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.