26

The sixth Community Asks Sprint is coming up soon (end of October or beginning of November – we will update the exact dates soon), though the last one concluded only a few weeks ago. We do these once per quarter, but since the end of this quarter is a holiday season for many people around the world, we are getting this one done on the earlier side.

The Community Asks Sprint is dedicated time for us 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 previous Community Asks Sprint and everything we resolved in it.

The focus of this upcoming one is going to be on site design and usability. I know that is a pretty broad topic, but basically we are looking at asks, new and old, which will improve the ease and efficiency of using Stack Overflow and Stack Exchange sites.

Let us know what UX/UI- or design-related asks would improve your experience on the sites you use. You can link to existing requests on any meta site, or mention something new. We have a few in mind, but we want to consider taking on what the community would find most impactful in the present moment.

Once this sprint is completed, we’ll share everything we were able to tackle, and we will also ask about what themes you would like us to consider focusing on next (so save those comments for the next post, please).

18 Answers 18

32

Currently, the company has several incorrect ideas about how web accessibility is supposed to work. These range from misunderstanding WCAG guidelines to viewing accessibility and usability as separate things. Unfortunately, SE are practically experts relative to the rest of the industry. This stuff is not well-documented. This makes it difficult for me to recommend (what I believe would be) the single most effective intervention to improve UI, design, and UX: "learn the basic principles of web accessibility".

So I'll repeat my suggestion for the Community-a-thon: please participate in Q&A about accessible web development practices. This solves three problems:

  • It lets you ask "stupid questions", to be answered by whoever has relevant expertise. Stack Overflow loves questions like these, provided that they're not too broad. We'll have no way to distinguish ignorance from thoroughness. You can even self-answer them, where you're confident.
  • It gives us insight into the state of the company's understanding of design and accessibility, beyond what precious fragments we can glimpse from public communications. This will make it easier to identify why you're doing particular things (which may not actually be mistakes after all!), improving company-community communications for the future.
  • It means there'll actually be a reliable, vaguely-comprehensive repository of accessibility-related information, something we're sorely lacking at present. Being able to point other people at a Stack Overflow question to say "here is what you should be doing instead of what you are doing" would be so much easier than having to write two paragraphs of explanation cross-referencing three sections of WCAG and a handful of blog posts (yes, these are the experts, no I can't prove that, but they are: that snake-oil vendor is lying to you).

I understand that this is an unorthodox suggestion: after all, sprints are usually about concrete features! But without an understanding of accessible design, your attempts to improve things could easily make things worse.

Take blockquotes. These used to have a coloured background (dark-grey or yellow, depending on theme), but this was changed to a white background to improve text contrast. But this made it harder for users to distinguish quotes from the surrounding text! So to fix this, the text was made light-grey… reducing the text contrast. So now blockquotes are still quite hard to distinguish from the rest of a post, and they're hard to read: strictly-worse than the old black-on-yellow colour scheme!

That kind of thing will keep happening everywhere until you understand accessible design. Nobody has the bandwidth to try to negotiate fixes for every design issue on a case-by-case basis, especially when they interfere with the coherence of the design system. So let us teach you the principles, so we can stop micromanaging, and let you do your jobs!


This proposal applies double to dev-related accessibility concerns, such as how to lay out the DOM or how to make decisions about client-side versus server-side; but it's hard to motivate those to a non-developer. I personally think those are more important than such trifling concerns as which pixels get painted what colour on the average user's screen.

5
  • 2
    You're suggesting to ask accessibility questions on their own platform, but you're also suggesting that many accessibility-related answers can't be trusted as they're not answered by true experts? I'm a little confused what you're wanting them to do then. Not to mention, many of their questions would probably be duplicates (e.g. if they asked about contrast ratios, that would be closed as a duplicate to the one you linked to). Do you think you could expound a bit? Commented Oct 16 at 21:05
  • 2
    Separately, I, myself, am a bit confused by the blockquote example. You're saying WCAG counts as an expert in the topic, but StackOverflow's blockquotes colors pass their AAA contrast test with flying colors, but that's still not good enough? While WCAG does a good job at laying down base rules for accessibility, they're obviously not the definition of it - do you think you could point me to other experts on the topic (articles, etc) that I could use to understand why the blockquotes aren't accessible? (Mostly for my own learning - I'm trying to understand accessibility better) Commented Oct 16 at 21:10
  • 1
    @ScottyJamison Accessibility-related answers from 2012 probably can't be trusted. Back then, people adapted images specifically to the iPhone's non-sRGB display, and HTML5 was two years from W3C standardisation. Nowadays, if someone's confidently wrong about accessibility matters, you've got good odds that someone else will correct them; this wasn't the case 10 years ago. SE's blockquotes aren't good enough because users struggle with them: that's the start and the end of the story. (If you find meta references, please link them here.) Commented Oct 17 at 0:15
  • 3
    @ScottyJamison However, now that I think about it: your question about Stack Exchange's blockquotes would be better asked on User Experience, where it could get more than a one-sentence answer. Commented Oct 17 at 0:41
  • 1
    It's kind of hard to say "Users struggle with block-quote colors, end of story" when users seem to struggle no matter what changes. Users have also complained about readability after a redesign increased contrast. But I'm also out of the loop - I haven't seen these blockquote-color complaints yet. Anyways, I posted the question, as recommended - feel free to discuss further there. Commented Oct 17 at 14:25
25

Better spoilers would be a massive usability improvement on Puzzling, the site I moderate. This includes:

6
  • 8
    But no ask for inline spoilers? ;) I'd really love it if they took a page from Discourse when it comes to spoilers. They have the fun ones that open and collapse plus inline spoilers, I think. Commented Oct 14 at 19:33
  • 2
    Oh, and their spoilers allow a "summary" that would let you describe the spoiler content in a more mini Q&A quiz-type format. Commented Oct 14 at 19:39
  • 1
    @Catija here, for you :) [honestly, I'd just forgotten] Commented Oct 14 at 19:43
  • 4
    Oh no, not spoilers in comments!!! :P Mini Markdown spoilers would also make them possible in Chat, I think? Anyway, this request would probably also have support from sites with plot-heavy Q&A, such as M&TV and Sci-Fi. Commented Oct 14 at 19:48
  • 1
    Also spoilers for chat. Commented Oct 14 at 19:50
  • 7
    This would also apply to any other site that uses spoiler tags, such as Role-playing Games (I imagine the Movies & TV, Scifi & Fantasy, Anime & Manga, etc. sites would appreciate it as well). Commented Oct 14 at 20:11
17

Let's try one of the Stack Apps ones:

17

Having indicators for what you've flagged would be appreciated

While the spam waves have somewhat gotten mitigated, it would still be nice to have the same functionality we have with closevotes here, and have a little * next to the flag link on any post.

12

Yay! I'm hoping this might be the sprint where you can take a look at Please, I want some more HTML - specifically for me, please enable support for the <details>/<summary> HTML tags. (I had asked for this in previous sprints without success). The original, declined feature request is at Please add the ability to fold blocks of code in questions and answers which has been re-asked in a more general form. The specific posts for <details>/<summary> are quite well received (+28/-2 for the older one and +18/-1 for the newer one at the time of writing).

12

It could be a post or maybe a sprint of its own, but there's a lot of parts of the site that are not keyboard accessible.

Over the past year or two, I've relied heavily on keyboard shortcuts as a basis for other tools to manage moderation workflows. I've M M 4 A seared into my muscle memory from spammer destruction. In many cases in dealing with profile spam, I really wish I could navigate/edit the spammer's profile image from a keyboard.

As far as accessibility goes - you'd be helping folks with visual impairment , or folks using assistive tools

Discussing this, there was also requests for keyboard shortcuts for review queues - and there's an existing request for that.

On the main sites - being able to pick a post for a flag or action better via the keyboard would be nice

For chat, likewise being able to pick a recent chat entry for a reply or flag via keyboard would be nice too. One of my current userscripts lets me hit : and up or down to select a message to reply to.

So basically could we look at what the community needs so they can more efficiently use the site without lifting their hands from the keyboard?

1
  • Regarding accessibility, WCAG 2.2 guideline 2.1 explains why keyboard accessibility is so important in the "Understanding..." outlinks. Commented Oct 16 at 6:05
11

enter image description here

The 'new' reply system relies on smaller (and different) fonts to differentiate the source of a reply. This can be difficult to read. Could we consider other means to do so? Using a colour block would be nice, and having the font less saturated makes it harder to see.

Could we re-visit this, and look at other ways to visually distingush a reply?

11

The following got marked in April following the March 2025 Community Asks sprint, but no other action has been taken as far as I am aware:

Protocol-relative URLs to embedded images are incorrectly being blocked

Wondering if this can be re-prioritised in this sprint?

11

Request regarding site design:

UI (site design)

UX (community contests)

Quite some sites lack their own design. We know it is hard for SE designers to a: design something unique for each of the many sites still needing a new look, b: capture what each community wants, especially if one does not happen to be active on said site.

SO why not outsource the problem? Make a community event out of it. Everyone wins:

  • SE does not have to figure out a multitude of site designs.
  • The community gets some fun design contests & of course new designs for the sites.

(For the managers/PMs still on the fence about it: That 2x engagement aka traffic for basically free: Once on meta at the event to make the design and then again on the site themselves as people are being made aware of them. It is a great opportunity to pivot and synergize™ the effects of a dynamic, light-hearted community first environment?)

Request regarding flags:

UI

When dealing with spam waves and near identical spam posts, one might accidentally open a post already flagged. This is annoying and time-consuming. But also easily fixable by having a small label Flagged symbolon any post one has flagged.

UX

Spam wave means there is a lot of very obvious spam, no need to open the post. No need to open it. So allow user with experience(!) to flag from the overview pages, but only for spam and moderator attention. The other flag types always require looking at the post.

Custom mod flags can get responses, you (the flag raiser) likely never realized, as the only place to find them is absurdly obscure. While at notifying users: Inform users about rejected edits, it irritates me much more than it should that there is simply no notification about it.

Request regarding spam itself:

UX

New users should do something (successfully suggest-editing a post, posting without being marked as spam, etc.) before being allowed to edit the "About me"/"Links" section. This would greatly reduce profile spam.

Reduce the visibility of profile spam by removing badge-related sites from indexing would remove spam from search results, discourage spammers, and even save some resources as fewer pages must be indexed.


This is a mostly carbon copy of my answer from June/September, which is itself a list of previous requests (that were upvoted but not implemented).

11

We currently have 2 and a half dialects of markdown on the network - commonmark (with additions) for posts, a comment specific subset, and a chat specific subset.

It would be nice to have them more similar - and a specific painpoint is I'm needing to remember to use <strike> </strike>for strikethroughs on main, and triple dash on chat. I also need to remember the other dialect of markdown used elsewhere where a strikethrough is a ~~

Here's what I would like to propose. Allow ~~ as an alternate to the three dash strikethrough on chat (for folks more familiar with the discord flavored strikethrough), and allow both as an alternative to using embedded html tags for the same on regular sites.

2
  • for posts, there's also the nuance of client-side library and server-side library (and SE/specific commonmark extension implementation on each side). markdown-it and markdig, if I remember correctly. Commented Oct 16 at 7:34
  • @starball as far as I know, any client-side/server-side differences in markdown rendering are still considered bugs, whereas those other differences were introduced intentionally (because it seemed like a good idea at the time). Commented Oct 26 at 0:14
10

The design of the Flag Summary page has had a significant usability bug since 2013: while it shows you all the flags you cast, it often doesn't show you who or what it was you flagged. Could this please be fixed/improved?

7

In the same way the post flag box was refreshed, could we also get those changes applied for flagging election nominations?

Current Proposed
/ /
For reference, here's the current post flag dialog:
Screenshot of the updated dialog for posts

It's a fairly minor change, but it would fix the flow of the text and improve the UX.

5

Improving the flagging UI:

Improving the sponsorship experience:

Improving some error message:

Improving the new Ask page:

Improving the user profile:

2
  • 3
    Note the "remove 5 second delay for flags" issue got marked status-deferred on Super User while SE tests the new anti-spam measures. Edit: oh, its your question on SU. Did you miss the update? Commented Oct 15 at 0:56
  • 7
    @Robotnik "We plan to revisit this in a month or two." I thought that'd be a good reminder to them + idk if they only consider it on SU or entire SE Commented Oct 15 at 1:05
5

It's not UX in the UI sense, but it's UX in the sense of having a decent homepage or recent activity view: Implement minor edits that don't bump questions to the homepage or recently active views. This would allow people to blast through the edit review queues without needing to pay attention to the (hard to see) age of the question (which may not necessarily align with an old question getting a new answer). It would help curators do minor curation without worrying about flooding the homepage with old questions or need to space out their minor edits.

5

Let me ask for the sticky links to Chat again;

Please make the links for Chat in the hamburger menu and the sidebar point to the last active tab.

I like to end up on the tab that I have had active last time that I visited Chat (i.e. sticky; which was usually my favorite rooms, currently called Starred). What I don't want is to always end up on All rooms.

1

Ah, bit of an old/controversial one.

Could we get the non sticky top bar back? It was a pretty controversial move back in the day. It takes up room on smaller displays, or if someone needs to zoom for accessibility reasons. It takes up space that could be used for content. It probably simplifies design.

It also makes full screen screenshots break in some contexts (which is sometimes useful for archival reasons).

The feedback then was to make it non sticky - and we never got a good reason other than everyone is doing it (when many weren't).

This feels like a win for both ease of use and accessibility - so 4 years on, could this be reconsidered?

4
  • you mean as an additional option? or as the only option? Commented Oct 16 at 12:01
  • Well SE doesn’t like options so the only one I guess Commented Oct 16 at 12:02
  • 2
    there are a lot of things I don't like about the top bar, but I userscript it to my taste. I take issue with top bars that try to be smart and hide themselves and then appear when you scroll up (and typically cause layout shift because... reasons?). thankfully SE's topbar doesn't do that. I just display:none; anything I don't use, set the placeholder text of the now much-wider searchbar to the page title, and made some text readability changes. I personally prefer it being sticky. aside- I think there are more suitable ways to archive webpages... Commented Oct 16 at 12:08
  • I can't really do that on a mobile device - and going to a 'not too smart', normal one's something that was/has been requested. Having it optional would be ideal, but apparently that's extra maintenance. Commented Oct 16 at 13:10

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.