86
\$\begingroup\$

It has come up several times recently (more than usual) that PPCG differs from most of the other Stack Exchange sites in that it's not a Q&A site. People don't come here to ask a question because they have a problem, people come here to solve recreational challenges. The most valuable contribution to this site isn't people using their free time to solve others' problems, but using their free time to think up interesting challenges for the enjoyment of others. This entire post rests on the premise that we are not by any definition a Q&A site. If you want to contest that, please start a separate meta post.

So the actual problem is that the entire site (and more importantly the onboarding experience for new users who don't look past all the standard texts and don't know about PPCG's "special" role in the SE network) is saying "this is a Q&A site, get help here if you have a problem and show what you've tried". I think that's why the "beta" label is more appropriate here than on any other ungraduated site. And given recent developments maybe it's time we actually talked about where those discrepancies are to see if something could be done about them.

Of course, every time someone brings up the "we're not a Q&A site discussion" that is always accompanied with "but SE isn't going to pour any dev hours into custom code for PPCG, so tough luck". And while it's probably true that we won't get large amounts of customisation unless this community becomes a seriously big player, there are some simple things like changing some standard texts that might not be much effort and would already greatly help improve the overall experience here. Plus, SE most definitely isn't going to spend any time on features that were never requested because we thought they wouldn't ever implement them anyway.

So the idea of this post is to discuss how the site could be improved for the way PPCG as a community uses it, and in terms of how we're not a Q&A site. The idea is to do this in the form of potential feature requests. Each answer should be one suggestion for an improvement to the site. These can be discussed in the comments and if they get some support they can get their own which can then be linked from here. By doing this "staging process" for feature requests relating to the Q&A issue, I hope that this post can act as a hub for them and we can post them in a more organised/coordinated fashion that increases the chances of us getting any of these things at some point in the future.

Note that this isn't about sorting out the gritty details of every single potential feature request like specific phrases and things like that. Those can be discussed when a corresponding feature request is posted later on. This is about suggesting which parts of the site could be improved (and roughly how), as a community brainstorming session, so we can see how they might work together and focus on pushing for the feature requests that seem most promising and urgent.

Please do not repost things that have already been asked about on meta. Existing feature requests related to the Q&A issue include:

\$\endgroup\$
  • 5
    \$\begingroup\$ Interestingly, we have a handful more "questions" than Puzzling does (22 pages). It's also in beta, of course, but it means PPCG isn't unique in how it uses the site. \$\endgroup\$ – Draco18s Feb 9 '16 at 22:15
  • 2
    \$\begingroup\$ @Draco18s Right, we're well aware of our friends at Puzzling (we even share a moderator), but the number of non-Q&A sites on SE is like a drop in the ocean compared to the "traditional" Q&A sites. \$\endgroup\$ – Alex A. Feb 9 '16 at 22:55
  • 5
    \$\begingroup\$ Taking @Rainbolt's feedback into account, perhaps it would make more sense to restrict this to new ideas for improving the site. \$\endgroup\$ – Alex A. Feb 9 '16 at 23:14
  • 4
    \$\begingroup\$ @Rainbolt this is specifically about improvements regarding the problem that we're not a Q&A site but are using Q&A software. This is specifically not tagged feature-request but discussion, because those actual feature requests are still supposed to be posted separately and linked to from here. The main idea is to collect all these closely related ideas in one place to refer to. I'm pretty sure this will ultimately have an expiration date automatically, when either SE goes about implementing some of these things or says "not going to happen". \$\endgroup\$ – Martin Ender Feb 9 '16 at 23:25
  • 9
    \$\begingroup\$ @Draco18s, Puzzling started out as Q&A and drifted a long way away from its original scope. It's a shame, because it was a more interesting site in the first couple of months than it is now. \$\endgroup\$ – Peter Taylor Feb 10 '16 at 8:27
  • 2
    \$\begingroup\$ Not an answer as this already has a meta post, but since, like Code Review, we're not a Q&A site with only short snippets, Should we increase our post character limit to allow substantial stack snippets? \$\endgroup\$ – trichoplax Feb 12 '16 at 4:12
  • 7
    \$\begingroup\$ Question - let's assume that you folks were to be set to "graduated" status right this week. Would that be alright, or would you folks prefer that one or more of these textual-quick-fix-maybe-sorta deals were addressed before we started flipping switches? Keep in mind that privilege levels won't immediately change, so there shouldn't be too large a change to warrant concern, but thought I'd check in first if these were more of "Things to look into as part of the graduation process" or "Things to look into prior to graduation". \$\endgroup\$ – Grace Note Feb 22 '16 at 18:26
  • 3
    \$\begingroup\$ @GraceNote I personally assumed these would be things that happened after graduation (maybe around the time we get our design) and not before. \$\endgroup\$ – a spaghetto Feb 22 '16 at 18:35
  • 7
    \$\begingroup\$ @GraceNote I personally don't quite agree with Mego's order there. Of course, we don't want all of these things to be done before the beta label goes (I'm not confident that we'll ever get some of them, since they'd require substantial work on your part, and we're still a rather minor site in the network, I suppose). However, I think some things like the tour page and help centre are actually more important to the impression this site makes than getting fancy colours. Seeing how out of place some of these texts are they make the site seem more unfinished than the beta design. [tbc...] \$\endgroup\$ – Martin Ender Feb 22 '16 at 18:42
  • 5
    \$\begingroup\$ So, maybe all of this doesn't have to happen before you flip the first switches, but it would be great if some of these things could be taken into consideration as part of the graduation process. I don't see the graduation myself as that incredibly important that everything else on this list can wait. So if we could get some of those things while we're graduating that would be great, even if it did push back the graduation by a few weeks. I feel like once we have our colours people might also stop caring about these changes to a certain degree. \$\endgroup\$ – Martin Ender Feb 22 '16 at 18:44
  • 4
    \$\begingroup\$ Alright. There's the graduation process (enables elections and community ads, removes beta label, allows migration paths pending approval) and then a separate design process (actual visual design, plus increase of privilege levels). If I understand it right, the former would be fine to proceed with, but we'd want to work on reviewing the stuff in here before starting the gears on the latter, yes? \$\endgroup\$ – Grace Note Feb 22 '16 at 19:15
  • 2
    \$\begingroup\$ I also don't agree with Mego. To me, having the site text make sense in the context of what we do is more important than actually graduating. Sure, the regulars are used to the generic Q&A style text, but it's certainly not obvious to new users how everything translates to our site. \$\endgroup\$ – Alex A. Feb 22 '16 at 19:15
  • 7
    \$\begingroup\$ @GraceNote I'd be fine with that. Seems like a good compromise between acknowledging that we're graduating and giving these requests the right priority in the graduation context. \$\endgroup\$ – Martin Ender Feb 22 '16 at 19:16
  • 2
    \$\begingroup\$ @GraceNote For the record, I second Martin's last comment as well. \$\endgroup\$ – Doorknob Feb 22 '16 at 20:59
  • 4
    \$\begingroup\$ Hey Martin! I agree with all the suggestions you and others have listed in the answers (excluding few exceptions on the second page). But as of March 29 2017, I don't think any of the suggestion has been implemented (I have'nt checked through all the answers; perhaps a few have been). So, is there any likelihood that the suggestion will ever be implemented? \$\endgroup\$ – Arjun Mar 29 '17 at 12:25

35 Answers 35

86
\$\begingroup\$

Can we get a custom Tour page?

The most important part of the onboarding experience for new users (especially those who don't arrive via other sites from the network) is the tour page. Moderators can customise the first paragraph, but in the case of PPCG this is hardly enough, because the rest of the page is strongly misleading:

Ask questions, get answers, no distractions

Get answers to practical, detailed questions

Focus on questions about an actual problem you have faced. Include details about what you have tried and exactly what you are trying to do.

Ask about...

It would be great if a) those misleading bits could be removed and ideally b) replaced with some important information that applies here instead. The usual process for new users on a Q&A site is that they join the site to ask a question and then start answer others as well when they feel confident about it. Around here, we prefer users to start with solving challenges so they can get a feeling for how challenges work around here before writing their own. (Yes, CH, you are an exception to that...)

\$\endgroup\$
  • 2
    \$\begingroup\$ With the unanimous support for this, I've turned it into a separate feature request now. I'll slowly do so for other popular ideas here, but not all at once, so that we can focus on each one separately. \$\endgroup\$ – Martin Ender Feb 15 '16 at 18:05
  • 46
    \$\begingroup\$ no distractions Anyone who has ever sneaked a peek at PPCG while doing something else knows this is a lie. :P \$\endgroup\$ – ETHproductions Feb 22 '16 at 16:30
72
\$\begingroup\$

Built-in support for scores and leaderboards

The vast majority of PPCG challenges involve minimizing or maximizing the score of a solution, usually the byte count of a program. As it currently stands, users are encouraged to start their answer with a header of the form

## Language, N bytes

These headers have no special meaning to the SE platform, and an often-requested feature is the ability to sort answers by their score. We have developed a leaderboard Stack Snippet that parses these headers and prints a pretty leaderboard, listing the overall ranking and winners by language. This works, but is not optimal, since new users may be unaware of the required header format, and only a fraction of our challenges actually include the snippet (it has to be copy-pasted from the Meta page, and the QUESTION_ID variable needs to be reset after the question is posted). Another workaround is given by a userscript (thanks @ETHProductions for mentioning it), but new users, unregistered visitors and those who can't install browser additions will not benefit from it.

What would be great is some built-in support for attaching scores, language names and/or other metadata to answers (some of which could be disabled for questions and unconventional challenges), and the ability to sort and/or filter answers based on the metadata. This would probably take much more effort to implement than most of the other requests here, and the snippet and userscript do provide workarounds, so it's by no means a critical feature.

\$\endgroup\$
  • 6
    \$\begingroup\$ There is also automatic leaderboard support in the PPCG redesign userscript, if that helps any. \$\endgroup\$ – ETHproductions Feb 9 '16 at 23:17
  • \$\begingroup\$ @ETHproductions Thanks, I added a mention to the post. \$\endgroup\$ – Zgarb Feb 10 '16 at 1:15
  • 4
    \$\begingroup\$ The way I'd love for this to be implemented is to have separate "areas" for challenges and questions. We'd keep a regular SE instance for non-challenge posts like our tips, questions asking for golfing advice (and could then probably also broaden our scope to questions about designing challenges and esolangs) - and in addition to that have a separate area for challenges, which would probably be a modified SE instance that supports metadata for answers, doesn't allow answers to be accepted (and in the long run maybe has software support for things like CnR...). One can dream... \$\endgroup\$ – Martin Ender Feb 10 '16 at 8:14
  • \$\begingroup\$ However, there is always the argument of h1 versus h2... \$\endgroup\$ – LegionMammal978 Feb 10 '16 at 12:09
  • 1
    \$\begingroup\$ This feature request has been brought up in the context of creating a friendlier golfing environment for people who don't enjoy the use and competition of golfing languages. \$\endgroup\$ – Martin Ender Sep 26 '16 at 7:56
  • \$\begingroup\$ If you can get built-in support for the language name as metadata on the answer, I think it would be great if that also enabled people to earn language-specific tag badges. I'd much rather have my tags reflect the languages I most commonly use than the fact that I'm answering code-golf questions. \$\endgroup\$ – James Holderness Dec 18 '16 at 16:41
62
\$\begingroup\$

Remove the "you already have an answer" pop-up

When you post an answer on a question where you have already posted an answer, a pop-up appears asking you if you really want to do this and encouraging you to refine your existing answer. While this makes perfect sense for Q&A sites, it doesn't make any sense for PPCG since answers are generally unique and standalone -- you shouldn't integrate a new solution into an existing one. It's also quite annoying (especially when posting on catalogs).

\$\endgroup\$
  • 22
    \$\begingroup\$ This pop up could be particularly misleading for new users, who might then think they are encouraged to post solutions in several languages in one answer. \$\endgroup\$ – trichoplax Feb 25 '16 at 2:51
60
\$\begingroup\$

s/question/challenge/g

This is pretty big and this might actually be a lot of effort on SE's part, but I think the overall impression of this being a Q&A site would be greatly reduced if the site didn't use the word "question" everywhere.

With a few exceptions, all "questions" on PPCG are "challenges". To post a new challenge you need to click on the "ask question" button. There are help centre pages about "how to ask questions". And you can get a list of "unanswered questions".

Changing a single word across the site might seem like a trivial change for so much effort, but I think the effect on the user experience would be quite amazing.

(That said, I realise that the biggest problem isn't even the amount of places that need changing, but that the word "question" is actually tied pretty strongly to the software, e.g. search queries use is:question and inquestion:???, URLs use question or q, etc. This is probably a utopic idea but I wanted to put it out there.)

\$\endgroup\$
  • 9
    \$\begingroup\$ I wouldn't say all questions are challenges. tips are on topic and calling them "challenges" seems odd. \$\endgroup\$ – Downgoat Feb 9 '16 at 23:06
  • 19
    \$\begingroup\$ @Downgoat "With a few exceptions..." I'd rather make the few non-challenge posts we have seem out of place with the terminology than the >95% of challenge posts. \$\endgroup\$ – Martin Ender Feb 9 '16 at 23:10
  • \$\begingroup\$ Also, s/answer/submission/ would help with the same problem. (If we start changing stuff this deep, it wouldn't probably hurt.) \$\endgroup\$ – PurkkaKoodari Feb 17 '16 at 10:40
  • 1
    \$\begingroup\$ @Pietu1998 That's true, but I wanted to focus on one thing at a time, and calling a submission to a challenge an "answer" to it doesn't seem to be quite as much of a stretch as calling a challenge a "question". \$\endgroup\$ – Martin Ender Feb 17 '16 at 10:42
  • 12
    \$\begingroup\$ I thought of a simple way that might make this possible. If there's Stack Overflow in Spanish and Portuguese, we might be able to get a custom language file with the desired changes. \$\endgroup\$ – PurkkaKoodari Feb 25 '16 at 12:11
52
\$\begingroup\$

Fixing the Help Centre

This has the same problems as the tour page. On one hand it's not quite as urgent because it's not as prominent as the tour page, on the other hand the problem exists on a much bigger scale here. I have a 2-3 page list of things that could use changing. Some of them don't apply to PPCG at all and are actively misleading. Others could just use some additional usage guidance specific to PPCG (e.g. that all challenges need a challenge-type tag; that edits should not be used to improve the score of an answer; etc.). I'll include the full list if this is turned into a feature request, but some of the more obvious bits:

\$\endgroup\$
  • 4
    \$\begingroup\$ Deleted my answer because you're just too good at this. +1 \$\endgroup\$ – Geobits Feb 9 '16 at 22:45
49
\$\begingroup\$

Modify the trigger for low quality posts

Occasionally, very short answers to code-golf competitions written without explanations will end up in the low quality posts queue on the ground that they are too short and are mostly composed of a code block (I just reviewed one, minutes ago). Sure adding an explanation would make a post of higher quality but no explanation does not necessarily means low quality either. Longer answers that are still composed of only a code block without explanations will not trigger this, which put us in an absurd situation where a longer answer to a code-golf competition is considered of higher quality than a shorter one. In the context of a Q&A this trigger makes sense but in the context of a competition linked to the length of the answer it obviously does not.

\$\endgroup\$
  • \$\begingroup\$ This is an issue that have been discussed earlier but i don't think we talked about changing that before. \$\endgroup\$ – plannapus Feb 10 '16 at 9:22
  • 39
    \$\begingroup\$ Even worse, if a post is auto-flagged as low quality and the user edits their post, the flag is automatically marked helpful and the Community♦ user automatically downvotes. Serious ಠ_ಠ. \$\endgroup\$ – Alex A. Feb 12 '16 at 5:57
  • 12
    \$\begingroup\$ @AlexA: WTF? How is that supposed to be a useful feature on any site? If that's really how it works currently, I think it deserves a feature-request on meta.SE to disable it globally. \$\endgroup\$ – Ilmari Karonen Feb 23 '16 at 19:43
  • 4
    \$\begingroup\$ @IlmariKaronen There was one and apparently it was "fixed," but we did some testing here and it seems not to be. \$\endgroup\$ – Alex A. Feb 23 '16 at 19:51
  • \$\begingroup\$ @AlexA. Could you link to the mother meta post? \$\endgroup\$ – Martin Ender Feb 27 '16 at 12:03
  • 5
    \$\begingroup\$ @MartinBüttner Meta.SE post. Supposedly the downvote by Community is supposed to be revoked if the post later gets an upvote, but some testing has shown that this may not actually be happening. \$\endgroup\$ – Alex A. Feb 27 '16 at 20:25
  • \$\begingroup\$ @AlexA. If you've tested that, it might be worth bringing that up on that MSE post? Or making a new bug report on MSE? \$\endgroup\$ – Martin Ender Feb 27 '16 at 20:31
  • \$\begingroup\$ @MartinBüttner Yep. I'll have a chat with Dennis, who orchestrated the test, to get all of the details. \$\endgroup\$ – Alex A. Feb 27 '16 at 20:33
  • 1
    \$\begingroup\$ @AlexA. I wonder how much rep has been lost due to that! \$\endgroup\$ – noɥʇʎԀʎzɐɹƆ Mar 30 '16 at 23:55
  • \$\begingroup\$ I recently noticed I've only clicked "Recommends Deletion" once, for a dupe. \$\endgroup\$ – NoOneIsHere Aug 10 '16 at 19:14
  • 6
    \$\begingroup\$ A more sensible trigger on PPCG would be for answers that don't contain a headline (so that it can be edited in) or don't contain a code block (because those are almost always spam or answers that should have been comments). \$\endgroup\$ – Martin Ender Sep 24 '16 at 10:10
44
\$\begingroup\$

Don't steer new users into posting challenges

The SE interface points first-time users to ask a question. The Ask Question button is bolded and the tour is largely about asking. Answering is more tucked away. This makes sense for Q&A where users go for help. But on PPCG, posting a challenge without having participated rarely goes well.

Our site design should foremost direct users to

  • Answer challenges, and give them guidance on how to do so
  • Post challenges in the Sandbox for feedback rather than directly on the main site
\$\endgroup\$
  • 6
    \$\begingroup\$ Having a good solution for this would be great. Just the other day I thought it would be neat to have some "find a challenge to solve" feature (beyond the search capabilities we already have). Although I think they're working on new question-discoverability features for SO already, so those might be very helpful for us as well. \$\endgroup\$ – Martin Ender Jun 26 '16 at 11:52
  • 1
    \$\begingroup\$ @MartinEnder Really the starting rep should be a min of 5. So new users can post on the meta. My first question ended up with 2 meta posts by me (I got enough rep to do that from my challenge) and 2 closings. I stuck through that but it was kinda crazy that I needed that much work as a new user to get one challenge down. \$\endgroup\$ – Christopher Jan 26 '17 at 21:54
  • \$\begingroup\$ @ChristopherPeart as far as I'm aware PPCG has no rep limit for posting on meta any more. \$\endgroup\$ – Martin Ender Jan 27 '17 at 0:47
  • \$\begingroup\$ Thanks for the info. \$\endgroup\$ – Christopher Jan 27 '17 at 0:54
44
\$\begingroup\$

Better Close Reasons

This has now become its own separate feature request.

Mods can edit a few slots of custom off-topic reasons. However, the majority of close vote reasons are fixed and cannot be changed by us. While the names of the reasons largely make sense here, their descriptions don't:

  • Duplicate: "This question already has an answer here:", "This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question."
  • Unclear: "Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question."
  • Too broad: "There are either too many possible answers, or good answers would be too long for this format. Please add details to narrow the answer set or to isolate an issue that can be answered in a few paragraphs."
  • Primarily opinion based: "Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise." (This one in particular I don't think we need at all, because we've got that covered with lack of objective winning criterion.)

The notice boxes and tooltips on questions that are closed have similar problems. We also might want to think about whether there's a custom close reason that doesn't really fit under the "off-topic" banner.

\$\endgroup\$
  • 5
    \$\begingroup\$ We're willing to look into this, but you need to flesh out what you'd want for close reasons. I'd start with a new feature request similar to what was done here and get buy-in for new reasons and we'll look into it. No guarantees though. \$\endgroup\$ – Taryn Jan 25 '17 at 19:18
41
\$\begingroup\$

Remove the Accept feature

This feature is likely way too integral to how SE instances work for this to be possible without forking the software, but I don't think the concept of accepted answers is useful on PPCG.

The most fundamental pillar of this community is that every single question needs an objective winning criterion. That means there's really no reason to give a human the ability to choose a best answer, because the only thing that can really happen is that a) the author chooses the wrong answer or b) at some point the accepted answer will be outdated. The only real benefit is that the best-scoring answer can easily be found, but built-in support for scores would make this obsolete, too. Additionally, the best-scoring answer isn't necessarily the most interesting (e.g. because some language had a built-in for the challenge), so pinning it above longer but cleverer (and higher-voted) answers isn't very beneficial here either.

Additionally, people regularly emphasise to users who are discouraged by golfing languages that they should view challenges as separate competitions within each language. While the green checkmark is only worth 15 rep, it still seems to matter quite a lot to people and is counterproductive to giving the impression that the overall winner doesn't matter as much.

(Side note: a few badges would have to go with that. Scholar, Enlightened, Guru, Populist, Tenacious and Unsung Hero all depend on an answer being accepted.)

\$\endgroup\$
  • \$\begingroup\$ What about time-limited challenges like cops and robbers? In them acceptanse means the competition is over. \$\endgroup\$ – Qwertiy Feb 10 '16 at 13:09
  • 1
    \$\begingroup\$ @Qwertiy Sure, but there is still no reason to leave accepting up to a human because the criterion determining which answer should be accepted is entirely objective. \$\endgroup\$ – Martin Ender Feb 10 '16 at 13:11
  • \$\begingroup\$ Populatity contests do not have objective creteria. What if I want to make a populaty contest using a set of questions where each new question appears after previous competition is finished? In this case accepted answer frezes the winner at some moment, but woting continues going. \$\endgroup\$ – Qwertiy Feb 10 '16 at 13:15
  • 10
    \$\begingroup\$ @Qwertiy "Populatity contests do not have objective creteria." hey, yet another reason that popularity contests should be off topic. ;) \$\endgroup\$ – Martin Ender Feb 10 '16 at 13:29
  • \$\begingroup\$ I don't think so :) By the way, do you have enought rep to view deleted content on SO? What about idea with making subsites for some perposes? In case of CodeGolf, what about splitting the meta into few subsites? I think, I'll post a question sometime. \$\endgroup\$ – Qwertiy Feb 10 '16 at 13:33
  • 4
    \$\begingroup\$ I would love for this to happen (more than any other feature listed here). \$\endgroup\$ – Nathan Merrill Feb 10 '16 at 21:41
  • \$\begingroup\$ There's a possible issue here, which is that submissions that don't actually solve the question are often posted, so accepting the best compliant solution is sometimes of value. This is especially relevant in older questions, and in questions with more answers. \$\endgroup\$ – isaacg Feb 12 '16 at 8:18
  • 3
    \$\begingroup\$ @isaacg We have a fairly clear consensus on what should be done with "submissions that don't actually solve the question" (delete them). Especially when this applies to the leading answer, I'm sure people would pay attention. \$\endgroup\$ – Martin Ender Feb 12 '16 at 8:20
  • \$\begingroup\$ @MartinBüttner I'm thinking of submissions like this: codegolf.stackexchange.com/a/6142/20080 which are very popular joke submissions. \$\endgroup\$ – isaacg Feb 12 '16 at 8:24
  • 1
    \$\begingroup\$ @isaacg That one in particular is probably now officially invalid with orlp's latest meta question, but I get what you mean. Still, even in that case I'd rather trust the community's judgement on whether an answer should be valid or not than a single user's. \$\endgroup\$ – Martin Ender Feb 12 '16 at 8:33
39
\$\begingroup\$

Tag badges for challenges

Questions (challenges) aren't counted towards tag badges. Seeing how much emphasis we put on the fact that high-quality challenges are the main contribution someone can give to the community, it would be nice to reward people who post a lot of (well-received) challenges about certain topics.

This could be implemented by simply counting challenges towards the normal tag badges, but I think it would be nicer (and more interesting), if there was a second batch of tag badges which only counted challenges. These would probably need different thresholds from the answer badges... 20 challenges in any tag except is a lot.

\$\endgroup\$
  • \$\begingroup\$ You may want to make that a separate feature request. \$\endgroup\$ – programmer5000 May 19 '17 at 22:32
34
\$\begingroup\$

Remove "This question has more than 30 answers already" pop up for challenges

It doesn't make sense to restrict the number of solutions to a challenge. In some cases there may be solutions in more than 30 languages, and multiple solutions per language are encouraged.

We have a small proportion of non-challenge questions, but I don't see a problem with them not having this pop up.

\$\endgroup\$
  • 2
    \$\begingroup\$ Most non-challenge questions are tips. Despite not being challenges, those also have a culture of the answers contributing to each other, rather than competing, and of splitting up multiple different tips into different answers. So I'm not sure the warning would make much sense for those questions either; the "you have already posted" warning definitely doesn't, the "make sure this isn't a duplicate" warning might make sense but it's borderline. \$\endgroup\$ – user62131 Feb 17 '17 at 18:53
33
\$\begingroup\$

Better Bounty Reasons

This has now become its own separate feature request.

When starting a bounty, people here almost always use either "reward an existing answer" or "draw attention", occasionally "improve details". That's because none of the others make any sense here (and the descriptions of these could also be improved):

  • "Authoritative reference needed Looking for an answer drawing from credible and/or official sources." This is simply not a thing here.
  • "Canonical answer required The question is widely applicable to a large audience. A detailed canonical answer is required to address all the concerns." For challenges this doesn't make any sense either. It might be useful for the odd question about golfing tips, but that can probably be covered with "improve details".
  • "Current answers are outdated The current answer(s) are out-of-date and require revision given recent changes." Not a thing here - with our current rules, new languages/features aren't even allowed to be used.
  • "Draw attention This question has not received enough attention." Yep, this works. (Although I'd rather read "challenge" in that sentence.)
  • "Improve details The current answers do not contain enough detail." The idea of this works in a broad sense, but we're not really looking for "detail" in answers here. Maybe this could be reworked into a bounty reason to improve the score (overall winner, or language-specific solutions).
  • "Reward existing answer One or more of the answers is exemplary and worthy of an additional bounty." This works verbatim. :)

We could also think about whether we could use other bounty reasons here. E.g. bounties are often offered for bonus/side challenges. Maybe we could get a specific bounty reason for that?

As a slightly more elaborate feature request (which will probably go in a separate post eventually), it might be good to have software support for our "bounties without deadlines".

\$\endgroup\$
  • \$\begingroup\$ I'm not saying this will definitely happens, but if you post a new feature request similar to what was done here we will look into it. \$\endgroup\$ – Taryn Jan 25 '17 at 19:18
32
\$\begingroup\$

Vote button tooltips

Whenever someone asks why their challenge or answer was downvoted (or whether they should up- or downvote some other post), the best response is often "well, what does the tooltip say", since on other SEs it sums up reasons for voting pretty well. They don't really work here though:

  • The question upvote says: "This question shows research effort; it is useful and clear." Apart from "clear" these aren't necessarily qualities we're looking for in a challenge. "Interesting" and "well-specified" are a lot more relevant. Correspondingly, the downvote tooltip is: "This question does not show any research effort; it is unclear or not useful."
  • The answer upvote says: "This answer is useful." Not at all what we're looking for. We tend to upvote answers that are interesting, clever, show effort or are educational (but not in the same sense that an answer on a Q&A site is educational - this part is pretty much entirely optional here and usually goes beyond what the challenge actually requires). (And of course, the corresponding downvote tooltip is "This answer is not useful.")
  • Accept button: "Click to accept this answer because it solved your problem or was the most helpful in finding your solution." I didn't have a problem to begin with and I didn't need help with finding a solution. Something like "because it solves the challenge and achieved the best overall score" would make a lot more sense.
\$\endgroup\$
  • 1
    \$\begingroup\$ I think that research effort is something we should look for in questions. It's always advisable to write a reference implementation so that you can spot flaws in your spec: that's research effort. In fact, I would go so far as to say that a poor spec almost always implies a lack of research effort. \$\endgroup\$ – Peter Taylor Feb 10 '16 at 8:31
  • 1
    \$\begingroup\$ @PeterTaylor I did consider addressing that, but figured there are very good and simple questions that people can write up without actually having to look up anything anywhere (e.g. Zgarb's latest "Hollow out a matrix"), so I wasn't sure whether that would qualify as research effort or not. So while you're right, maybe there's a better way to phrase it than "shows research effort" that would apply to hard and simple problems alike? \$\endgroup\$ – Martin Ender Feb 10 '16 at 8:33
  • \$\begingroup\$ As I know, these texts can be updated as localisation. But I think it would be better to ask a sepatate question for the phrases before. \$\endgroup\$ – Qwertiy Feb 10 '16 at 13:07
  • \$\begingroup\$ @Qwertiy I'm pretty sure all of these will get more detailed discussion when they are posted as separate feature requests. \$\endgroup\$ – Martin Ender Feb 10 '16 at 13:11
  • \$\begingroup\$ @Qwertiy I tried clarifying that in the question now. \$\endgroup\$ – Martin Ender Feb 10 '16 at 13:14
32
\$\begingroup\$

Increase the "Ask Question" Privilege Threshold to +2

I should note that I'm not convinced that this is a good idea but I figure I'll mention it to see how others feel.

On Q&A sites, the point of posting a question is to get help, to ask a question and get an answer. There shouldn't be a barrier to asking a question. However, in the context of creating and posting challenges, there are many more things to consider.

For this reason, it may be a good idea to require some familiarity with the site in order to post a challenge to the main site. The easiest way to do that would be to require potential challenge authors to have earned at least 2 reputation on this site before being allowed to post questions. This would require one upvote on one answer or one accepted edit. Note that "on this site" implies that an association bonus wouldn't count toward this privilege.

This would also help reduce the number of blatantly off-topic questions we get.

\$\endgroup\$
  • 22
    \$\begingroup\$ I wouldn't even mind it being +10 and if it would disregard association bonus. It's extremely rare that someone posts a decent challenge without ever answering one. \$\endgroup\$ – Martin Ender Feb 14 '16 at 22:32
  • \$\begingroup\$ I agree, and am aware this is Programming puzzles, but would this weed out those who are not programmers from posting challenges (since they'd likely not be able to answer one)? I guess a user could still suggest an edit, but then we're forcing users to go sift through pages of challenges to find something worth editing. \$\endgroup\$ – Tas Feb 15 '16 at 4:05
  • 11
    \$\begingroup\$ -1 I'm against this one. My first activity on this site was posting a question (besides coming into chat), and considering how much emphasis we place on challenges, I don't think we want an additional barrier against potential challenge-writers. \$\endgroup\$ – Nathan Merrill Feb 16 '16 at 15:18
  • 4
    \$\begingroup\$ @NathanMerrill At least subjectively the signal-to-noise ratio in challenges posted by people who never participated in one is a single-digit percentage. While there are some notable exceptions, I do think this would greatly reduce the noise we get. If someone is actually determined to post interesting challenges, I don't think posting one measly answer is going to hold them back. \$\endgroup\$ – Martin Ender Feb 16 '16 at 18:09
  • 3
    \$\begingroup\$ -1 here. While not every new user can post good challenges right away, why make them jump through hoops if they can? For a concrete example, take a look at this question, the very first site action of a new user. Now, if this now-prominent user had been presented with a banner/popup saying "sorry, you can't post yet because you haven't posted other things", would they have stayed? I don't know, but I do know it would have been a great loss to the community if not. \$\endgroup\$ – Geobits Feb 28 '16 at 5:06
  • 4
    \$\begingroup\$ TL;DR: While we're not exactly a Q/A site, we are still a Challenge/Answer site. Don't handicap half of that by placing arbitrary restrictions on it. \$\endgroup\$ – Geobits Feb 28 '16 at 5:08
  • \$\begingroup\$ I disagree. Some newbie can create a great challenge, so keep it as usual -1 \$\endgroup\$ – Akangka Mar 8 '16 at 11:29
  • \$\begingroup\$ It could be a way to lead new users to use the sandbox. In the sandbox the challenge would get at least one upvote. So making it +10 is not that much. \$\endgroup\$ – Hedi Dec 3 '16 at 22:44
  • \$\begingroup\$ @Hedi: Upvotes on Meta don't give reputation, though. \$\endgroup\$ – user62131 Jan 28 '17 at 1:23
32
\$\begingroup\$

A bounty-like feature for challenges

Sometimes, challenges are so well-designed and/or have taken so much effort, that I want to give the author a bunch of my rep for the great service they've done the community. Currently, the best/only way to reward a challenge is to put a bounty on it to attract more answers (and maybe some more votes because of the additional attention) and to share it off-site e.g. on reddit or Hacker News.

It would be great if the software provided a way to reward our most dedicated challenge authors more directly.

I'm saying "bounty-like" because it probably doesn't make sense to put them up for 7 days and say "this will go to the best challenge which meets criteria X, Y, Z". These would either be indefinite bounties ("first well-defined challenge that pulls off an audio-to-image processing task") or could just be given to any challenge on the spot (maybe also featuring it, like a normal bounty would). Of course, one would have to carefully think about whether and how this system could be gamed.

\$\endgroup\$
  • 1
    \$\begingroup\$ This doesn't seem like something that would only benefit PPCG and Puzzling. I would imagine many other communities would like some way of rewarding users who wrote a particularly nice question. Perhaps a way to have the rep awarded immediately and the same "Featured" flag get set so the challenge is more visible for 7 days. Rather than a predictive bounty, this would necessarily be after-the-fact (a reward rather than a posted bounty). \$\endgroup\$ – AdmBorkBork May 24 '16 at 20:53
  • \$\begingroup\$ @TimmyD I think it was asked on Mother Meta and declined. \$\endgroup\$ – Martin Ender May 24 '16 at 21:27
  • \$\begingroup\$ Puzzling is currently discussing this. \$\endgroup\$ – Bobson Jul 1 '16 at 18:20
  • \$\begingroup\$ @Bobson So are we. \$\endgroup\$ – Martin Ender Jul 1 '16 at 18:20
  • \$\begingroup\$ @MartinEnder - Ah, there it is. I thought there should be one, but I didn't see it. Thanks. \$\endgroup\$ – Bobson Jul 1 '16 at 18:26
30
\$\begingroup\$

Modify the placeholder search bar text

If you look at the top right corner of the page, you'll see a search box with the placeholder text "Search Q&A." Even outside of the context of deemphasizing the Q&A terminology on our site, I think this could be better phrased as "Search the site" or similar. So while we're requesting that Q&A-specific text be changed, we might as well request this as well.

\$\endgroup\$
  • 1
    \$\begingroup\$ If this bothers anyone a lot, you could always use the design userscript c: \$\endgroup\$ – Downgoat Feb 10 '16 at 1:09
  • 1
    \$\begingroup\$ I agree this would be an improvement to SE sites in general (currently it could suggest that the network is being searched instead of just the specific site), but it's even more important here on PPCG, even if they don't change it network wide. \$\endgroup\$ – trichoplax Feb 10 '16 at 1:13
  • 1
    \$\begingroup\$ I'm not saying this is 100% possible, but if you post a new feature request similar to what was done here, we will look into it. \$\endgroup\$ – Taryn Jan 25 '17 at 19:16
30
\$\begingroup\$

Stop automatically protecting challenges

Right now, a question will be automatically protected if it gets three or more deleted answers. This is a good move for Q&A sites, but the nature of PPCG makes this problematic:

  • PPCG tends to get an unusually high number of answers per challenge, making the odds of a few deleted answers much higher.

  • Self-deletion because of missing a detail or edge-case in the challenge is very common.

  • Lots of amazing answers come from new users.

Additionally, spam is relatively rare and efficiently handled by the community when it comes up, so protecting has less upsides too.

Because of this, right now the community consensus is to always unprotect challenges once they are auto-protected. This is a pain, and it would be nice to not need to do anymore. If a particular question is unusually problematic, it could always be manually protected, but I imagine this would be rare.

To quote Dennis,

I've had to unprotect Is this number a prime? three times so far. If I hadn't done it, the creator of this answer (new user, hasn't written another post before or after that one) wouldn't have been able to post his stellar Hexagony primality tester which earned 90 upvotes and a juicy +500 bounty. Yes, this is a rare gem, but it surely was worth the 22 answers that had to be dealt with without protection.

\$\endgroup\$
  • \$\begingroup\$ I was about to ask this. \$\endgroup\$ – Rohan Jhunjhunwala Feb 9 '17 at 20:11
  • 9
    \$\begingroup\$ I'd add to this that oftentimes deleted answers from established users get the challenge protected (which makes no sense at all) and that the handful for challenges that deserve to be protected can be handled by the users with that privilege. \$\endgroup\$ – Dennis Feb 9 '17 at 20:28
29
\$\begingroup\$

Support for indefinite bounties

I don't know if any other site is doing this, but it's quite common here to promise a bounty without actually starting it, because some are asking for additional effort on a challenge that can easily take more than 7 days or are not even known to be possible. In fact we're keeping a list of them on meta and it's fairly long. I don't see this practice going away and it's a nice source of extra(-hard) challenges.

It would be helpful if there was software support for this, so a) we wouldn't have to keep track of it manually (I'm sure there are some additional such offers buried in our challenges that aren't in that list yet) and b) it would give more exposure to this extra challenges, because not everyone is aware of the linked meta post.

Just to clarify, I'm not proposing that bounties on PPCG should generally be without a deadline. This should be a completely separate feature (which wouldn't even necessarily be tied to a single challenge, as some offers on that list show), and the regular 7-day bounties should remain unchanged. I also don't think these should get mixed with the regular bounties in the featured-challenges tab. They could have their slightly less prominent own spot, and should be retractable (e.g. in case someone decides to solve their own challenge). Or maybe - as for normal bounties - the capability of refunding rep should be left to moderators, so people can't game the system easily.

\$\endgroup\$
  • \$\begingroup\$ It can be simulated by putting bounty description in your challenge text. \$\endgroup\$ – Akangka Mar 8 '16 at 11:24
  • 1
    \$\begingroup\$ @ChristianIrwan right, but the issue is that a) the bounty might be by someone else than the author of the challenge and the author might not like to have this edited into their challenge but more importantly b) it's still not possible to find all of these bounties unless we stick to a fixed, searchable format (and even then it would be essentially invisible to new users). \$\endgroup\$ – Martin Ender Mar 8 '16 at 12:14
28
\$\begingroup\$

Collaborative posts with equal rep sharing

Occasionally 2 or more people will work together on writing a challenge or a solution. Currently all the reputation goes to the person whose name it is in. I'd love to see the ability for the post owner to add 1 or more additional owners so that all rep for the post is split equally among them.

Anyone else would still be able to edit as normal, without a rep reward.

I've seen both challenges and solutions that were a group effort, and I'd like to see more of both.

\$\endgroup\$
  • 3
    \$\begingroup\$ As I think about it, this gets more and more complicated. Consider a post with 3 owners getting an upvote. Do each of the 3 users get 3 rep, or does one of them get 4? What if it gets 3 upvotes? Does it then somehow figure out that each person should get an extra rep point? What if somebody un-upvotes the post? What about downvotes? \$\endgroup\$ – Nathan Merrill Sep 8 '16 at 1:35
  • \$\begingroup\$ @NathanMerrill very good points. I'd still love to see this, but I understand it would need careful design, rather than just switching on an option. \$\endgroup\$ – trichoplax Sep 8 '16 at 13:41
  • \$\begingroup\$ I guess I'd probably give all the rep to the initial author, than have a periodic (daily?) assignment of excess rep to the additional authors. Then any odd amounts that can't be split evenly will just wait until the next assignment. \$\endgroup\$ – trichoplax Sep 8 '16 at 13:43
  • 1
    \$\begingroup\$ Maybe it would be better to only allow this if all authors are included at the time of posting, to avoid other problems with deciding which parts of the rep are to be shared and which are not. \$\endgroup\$ – trichoplax Sep 8 '16 at 13:45
  • \$\begingroup\$ @trichoplax Or perform a rep-recalc for the whole question if more users are added. \$\endgroup\$ – wizzwizz4 Sep 18 '16 at 18:27
  • \$\begingroup\$ @wizzwizz4 that would be ideal, but if a recalc is a technical hurdle then I wanted to suggest the deliberately simpler version in case it has more chance of being implemented. \$\endgroup\$ – trichoplax Sep 18 '16 at 21:55
  • 3
    \$\begingroup\$ @wizzwizz4 I think that would make people reluctant to add further authors since it would mean that they could lose rep in the process. One possibly working solution (which would require score meta data) might be to have a special kind of suggested edit to improve an answer's score. This would need to be approved by the OP themselves and it would grant the editor 10 (say) instead of 2 rep. This would not result in long-term rep gains for the editor but it would still be an improvement over the current situation. \$\endgroup\$ – Martin Ender Sep 22 '16 at 7:56
  • \$\begingroup\$ Closely related suggestion. \$\endgroup\$ – Martin Ender Sep 22 '16 at 7:57
  • \$\begingroup\$ Would it be possible that upvoting a comment gives reputation? This way everyone can reward helping comments. Another way would be that only the post owner upvote would give reputation. \$\endgroup\$ – Hedi Sep 25 '16 at 10:03
  • 1
    \$\begingroup\$ @Hedi I don't think that's a good idea because comments are also used for querying questions and answers \$\endgroup\$ – Blue Dec 3 '16 at 13:38
26
\$\begingroup\$

Feb 2017 - See this feature request.

This little blurb from Downgoat's PPCG Design Userscript:

Before you post, take some time to read through the forbidden loopholes if you haven't done so already.

That appears right above the Post your answer button when the script is enabled. Here's what it looks like as I type this:

enter image description here

I think a little recommendation (for main only, of course) to read this site's README.md is a good idea.

\$\endgroup\$
  • 3
    \$\begingroup\$ There are no guarantees, but if you post a new feature request similar to what was done here, we will look into it. \$\endgroup\$ – Taryn Jan 25 '17 at 19:17
25
\$\begingroup\$

Rephrase "Share your knowledge, Q&A style"

When you post a question on any site in the Stack Exchange network, you're given the option to immediately add an answer to "share your knowledge, Q&A style." That's nice. However, as is the point of this entire post, we're not a Q&A site, and thus this text makes no sense in the context of our site. When you think about it, posting a submission to a challenge is kind of like sharing knowledge, so if you answer your own challenge then you are in some sense sharing your knowledge. But in the context of our site, that phrase still sounds kind of awkwardly out of place.

\$\endgroup\$
  • 3
    \$\begingroup\$ While we're on the topic of the "Ask Question" page, the sidebar title is "How to Ask"... \$\endgroup\$ – Martin Ender Feb 18 '16 at 7:51
  • 4
    \$\begingroup\$ @MartinBüttner halp how does ask challenge \$\endgroup\$ – Alex A. Feb 18 '16 at 20:14
21
\$\begingroup\$

A way to refine questions before we can answer them

The sandbox feels rather hacky. On all other sites, questions are refined until they make sense but here, we have a problem when existing answers are already posted. A question should be able to go from a draft state to a final state. I don't have a particular idea on how to implement it, but the simplest for me would be to close (manually or automatically) the question as soon as it is asked ("waiting for approval and feedback"). Questions would have more visibility. When we are ready to answer, we reopen it.

Related: PPCG Sandbox Viewer

\$\endgroup\$
  • \$\begingroup\$ Heh, I was considering writing a similar answer today. I suppose the cleanest way would be simply to have software support for the sandbox as it is right now, and then have a way of declaring a challenge ready which would "migrate" it to main. It would solve a lot of challenge quality problems if every challenge could only be posted to that before being sent to main after a few people have signed off on it. (If this was mandatory it would however also be necessary to separate challenges and non-challenge posts completely, because it doesn't make sense to post the latter to the sandbox.) \$\endgroup\$ – Martin Ender Feb 19 '16 at 13:27
  • \$\begingroup\$ @MartinBüttner My personal problem with the sandbox is the inner platform effect: many questions inside a question (btw, could meta be used as a sandbox?) Whatever we choose to do, the place where questions are sandboxed need to be clearly visible and reachable, in particular for newcomers. If that was integrated within the site (with buttons and everything), it would be nice. \$\endgroup\$ – coredump Feb 19 '16 at 13:35
  • 3
    \$\begingroup\$ I think instead there should be a draft tag. \$\endgroup\$ – ASCII-only Feb 21 '16 at 20:59
  • 5
    \$\begingroup\$ @somebody What prevents anyone from answering a challenge tagged with draft? It would also mess with how SE does the HNQ, in that drafts might get there if they do get answers, and challenges are a lot less likely to end up there once the draft tag is removed because they are already a lot older then. (I know some people would love PPCG to be excluded from the HNQ entirely, because they also amplify the apparent popularity of some challenges that the core community here doesn't really like, but they are also our primary source of traffic and new users.) \$\endgroup\$ – Martin Ender Feb 22 '16 at 7:45
  • 1
    \$\begingroup\$ I disagree with making the sandbox mandatory. Many long-time users (myself included) often will post a question without it ever appearing in the sandbox because they feel it doesn't need community critique. That said, I am in favor of requiring newcomers post to the sandbox, and if that doesn't happen, making the sandbox much more visible when posting. \$\endgroup\$ – Nathan Merrill Feb 23 '16 at 15:35
  • \$\begingroup\$ Maybe add it as an additional review queue? But have it so a question isn't posted until it goes through (and have a privilege so your challenges don't have to go through it) \$\endgroup\$ – Blue Feb 24 '16 at 19:09
  • 3
    \$\begingroup\$ @First I'm not sure how the challenges by experienced authors would be negatively affected by having to go through the sandbox. If the handful of regular challenge authors who "don't need the sandbox" would have to use it but also all the inexperienced authors who need someone to look over their challenges would post better challenges then I think that's a very small price to pay. \$\endgroup\$ – Martin Ender Feb 24 '16 at 22:21
  • \$\begingroup\$ @MartinEnder It may be worth looking at how the Rosetta Code site handles this. Tasks initially start off with a draft status, and at that stage they are listed separately from "approved" tasks, but users are still actually encouraged to answer them. However there's a clear understanding that those early answers may need an update as kinks are worked out in the challenge and the requirements change. I think it helps ensure that the task is somewhat understandable and feasible by the time it makes the main list. \$\endgroup\$ – James Holderness Dec 18 '16 at 16:13
20
\$\begingroup\$

Allow the sandbox to be permanently featured

Most sites don't have or need a sandbox. Since we have "challenges" instead of "questions" it is very important to have a sandbox and to make sure that it is easily seen by new users, so that our challenge-quality can stay high. So, we have the sandbox featured.

However on other sites, it doesn't make sense to have a single meta post permanently featured, and so it is automatically un-featured every month. On one hand, mods can re-feature it every month, but this certainly isn't the best solution. There is right now a history of almost 200 edits on the sandbox. It would be nice if we could just have it always featured instead.

\$\endgroup\$
  • 9
    \$\begingroup\$ Honestly, I don't think this is a really big deal. I'd rather just have an actual sandbox supported by the software instead of our hacky solution via a meta post. \$\endgroup\$ – Martin Ender Aug 13 '16 at 15:33
  • 4
    \$\begingroup\$ @MartinEnder While an actual sandbox would be better, I think this is a much more realistic request. \$\endgroup\$ – Mego Aug 13 '16 at 22:19
  • 2
    \$\begingroup\$ @Mego Yeah, but I mean I don't think it's sufficiently important to waste any dev effort on it even if it wouldn't take a lot. \$\endgroup\$ – Martin Ender Aug 14 '16 at 13:26
  • \$\begingroup\$ @MartinEnder I think that all questions should go into the sandbox automatically. If they have at least, say, two votes by three days, they can go into the main question section. However, I'm quite new to this site, so there's probably some logic behind not having this. \$\endgroup\$ – 0WJYxW9FMN Jan 22 '17 at 12:22
18
\$\begingroup\$

Remove the "more than 30 answers" pop-up, but only on the Sandbox.

When you post an answer on a question that has more than 30 answers, you get a pop-up asking you if you're sure you want to do this. This is fine for most questions but it doesn't make much sense in the context of the Sandbox, and I fear could scare new users away from posting their challenges in the Sandbox.

\$\endgroup\$
  • 20
    \$\begingroup\$ I don't see why we'd to keep this pop-up on challenges either. \$\endgroup\$ – xnor Feb 16 '16 at 4:32
  • 1
    \$\begingroup\$ I'm not sure if this is tweakable at a per-question level, but if not, surely there wouldn't be an issue with doing it on the meta site if we find reason to keep it in place on main. Meta questions rarely go beyond 30 answers (there are only four right now, including the sandbox). \$\endgroup\$ – Geobits Feb 16 '16 at 5:08
  • \$\begingroup\$ @xnor On catalogs it could be useful. \$\endgroup\$ – a spaghetto Feb 18 '16 at 22:38
  • \$\begingroup\$ @xnor That's a good point. I've added a separate answer for that one. \$\endgroup\$ – trichoplax Feb 22 '16 at 0:46
18
\$\begingroup\$

System to Suggest Improvements (golfing help)

An important part of this site is suggesting ways to shorten existing answers. However, there isn't a good way to do this. Comments are generally the accepted method, but that poses a few problems:

  • New users can't make suggestions as they are unable to comment. This can (and has) cause(d) them to instead post a trivial variation of another answer because they can't suggest the change.
  • If the author does accept your improvements, you do not receive any reputation for your hard work; the best you can get is a shout-out such as "EDIT: thanks to <username> for saving x bytes!" added to the post. This could again lead people to post their own answer
  • it is hard to describe some changes with a comment; often it requires pasting in the old section of code and the new code, taking up more of our valuable time and precious space on our screens.

The other option would be to suggest an edit, but that seems to have been rejected by meta (with good reason).

I would personally love a "suggest improvement/golf" feature, as it makes up the majority of my activity on this site.

This feature would require the author to accept it; otherwise things could get messy.

\$\endgroup\$
  • 2
    \$\begingroup\$ This sounds like a nice idea provided that the original author's approval is necessary. Otherwise it comes with the same problems as suggested edits. \$\endgroup\$ – Martin Ender Mar 30 '16 at 22:15
  • \$\begingroup\$ @MartinBüttner good point! I added that to the post. \$\endgroup\$ – Cyoce Mar 30 '16 at 22:22
  • 2
    \$\begingroup\$ This would also be interesting in terms of completely collaborative answers. That's something that doesn't usually happen on a regular Q&A site, but for a tricky challenge it's not too uncommon for several people to work on a single answer together. \$\endgroup\$ – Martin Ender Mar 31 '16 at 9:41
18
\$\begingroup\$

Allow answers to be tagged, just like questions are

Unlike most sites, where questions are fairly specific, questions on PPCG are highly general; when we ask a question, we're really asking a range of related questions ("write a short Jelly program to do X", "write a short Python program to do X", etc.). As such, we get a range of different answers, and many of those answers need to be treated differently.

It makes sense that we could have a tag badge for answering questions in, say, Perl or Brachylog. The problem is, the site is looking for a tag on the question (because the answers don't have tags), and the question obviously won't have a tag specifying the language the answers are in. Allowing tags on answers to count towards the tag badge, in addition to tags on the corresponding question, would make sense. (On a side note, the tag isn't functioning anything like SE likely intended, because the tag is on almost every question; I'm not convinced that this is a problem, but it's likely at least worth being aware of.)

There are other tags, separate from language tags, that I can potentially see on answers. (For example, Anarchy Golf, another golfing site, makes heavy use of a "cheat" tag which is used to identify answers which meet the letter of the specification but not the spirit; cheating at golf is something of a genre of its own. Over here we let the genuine answers stand, and argue over whether the cheating answers count, but always count them as part of the same competition; answer tags might help solve this problem). Likewise, a tag could indicate the sort of algorithm that was used (e.g. recursion, or brute-force). These are all tools that are useful when searching for answers with particular topics, and for seeing what kind of answers people normally give.

This would also help with automatic syntax highlighting (at present, syntax highlighting is something that has to be added manually in a fairly tedious way, because we don't have language tags to suggest highlighting rules like other languages do).

\$\endgroup\$
  • 2
    \$\begingroup\$ Yes please. My idea for this so far was to add certain meta data fields for score and language etc to answers (I think there's a separate post here for that). But being able to tag answers would be really neat in addition to that (for choice of algorithm as you say). \$\endgroup\$ – Martin Ender Feb 17 '17 at 19:13
  • \$\begingroup\$ We could start putting [tag:python] or whatever in answers. That would be a lot of answers that aren't up to date, though. \$\endgroup\$ – mbomb007 May 14 '18 at 19:01
14
\$\begingroup\$

Start all questions (challenges) as initially on hold

(This is a specific proposal for implementing this general idea.)

In general, questions get a lot more attention when they're posted on main than when (and if) they're posted on Meta. This includes both answers, and improvements to the question itself. If we're lucky, the question gets fixed in the sense of being made better via edits, rather than being fixed in the sense of being unchangeable by answers which rely on loopholes in it (which normally makes the question unsalveagable even if it could have been good).

Because a question is more of a community collaboration here than it is on other SE sites (I know a very common pattern is post question → close → edit and reopen), it would make sense to have time to perfect the question before answers are allowed.

This would ideally be implemented in the SE software, but it's possible to do an inferior version without changes: you could get a bot with diamond moderator rights (or even a ton of bots which merely had the reputation to close posts) to close every question/challenge as soon as it was posted (probably unless it has as a tag), placing it on hold for a few days (after which it will naturally lapse to closed). The main awkwardness in this would be what close reason you'd use ("too broad" is probably closest but doesn't really fit).

This would create a lot of extra work for the reopen queue, but that's unlikely to be a problem given our current balance between users and reputations.

\$\endgroup\$
  • \$\begingroup\$ I'm not a fan of this. I'd much prefer a separate location to find in-development questions to help improve. Questions that are still being worked on would get lost in the activity list. \$\endgroup\$ – Nathan Merrill Dec 2 '16 at 22:00
  • 2
    \$\begingroup\$ @NathanMerrill Not true at all. Edits to closed questions still bump them to the top of the active questions list. Reopening a closed question does the same. \$\endgroup\$ – Mego Dec 2 '16 at 22:02
  • \$\begingroup\$ Edits are much less frequent than normal question activity, especially because we prefer questions to be sandboxed for a couple of days. Furthermore, even if it is on the active list, I'd have to look for on-hold questions, and hope that they are on-hold for sandboxing (and not actually on hold) \$\endgroup\$ – Nathan Merrill Dec 2 '16 at 22:05
  • \$\begingroup\$ @NathanMerrill A custom close reason would let you know it was on-hold for sandboxing, and a review queue would make them easily visible. \$\endgroup\$ – Mego Dec 3 '16 at 6:23
  • \$\begingroup\$ But then we'd also want to remove all comments when the on hold status is removed (and our HNQ presence would be non existent). While this solution may be simpler than others to implement, it seems far worse than our current sandbox solution \$\endgroup\$ – Nathan Merrill Dec 3 '16 at 10:26
  • \$\begingroup\$ @NathanMerrill Removing comments isn't hard, and not having a presence on HNQ wouldn't necessarily be a bad thing. We get a lot of drive-by upvotes on not-great challenges because of HNQ. \$\endgroup\$ – Mego Dec 8 '16 at 5:15
  • \$\begingroup\$ Right, none of these issue are impossible to overcome. However, the list of disadvantages: (appears in the list of current questions, must have rep to access to review queue, no HNQ even for most good questions, mods need to remove comments). Are the advantages that much better that they outweigh the disadvantages? \$\endgroup\$ – Nathan Merrill Dec 8 '16 at 13:18
11
\$\begingroup\$

s/answer/???/g

Complimentary to Martin Buettner's s/question/challenge/g request, the word answer is quite universal and integral to each SE site... except us, for all the same reasons noted in Martin's proposal.

To the end of consistency and sanity, this should probably change alongside, to... ? Well, I'm not sure what, but surely "Solution", "Response," etc are better alternatives.

\$\endgroup\$
  • 1
    \$\begingroup\$ I'm all up for rewording this, but much like the s/question/challenge/g request, the use of the word "answer" is all over the place on SE, so if it's changed it should be reworded consistently everywhere and not just on one button, because that might just increase the confusion even more. \$\endgroup\$ – Martin Ender Feb 29 '16 at 17:04
  • \$\begingroup\$ @MartinBüttner there, now it's a corollary to yours \$\endgroup\$ – cat Feb 29 '16 at 17:11
  • 4
    \$\begingroup\$ I tend to use "submission" or "solution" in my challenges. I'd support s/answer/solution/g. \$\endgroup\$ – Mego Feb 29 '16 at 19:39
  • 2
    \$\begingroup\$ I don't like "solution". It doesn't make sense with king-of-the-hill, cops-n-robbers, or answer-chaining questions. \$\endgroup\$ – Nathan Merrill Aug 22 '16 at 15:53
  • 7
    \$\begingroup\$ I agree. "Submission" seems more generally applicable than "solution". \$\endgroup\$ – trichoplax Aug 29 '16 at 2:49
10
\$\begingroup\$

Adding a "bytes" sorting order to the solution section of each challenge

The built-in leaderboard function might be too "clumsy" for the people at SE HQ to implement. This might be a cleaner way of fixing the same problem.

\$\endgroup\$
  • 1
    \$\begingroup\$ Related. I suppose if general meta data for answers would be added, then it would be simple enough to sort by that meta data as well. \$\endgroup\$ – Martin Ender Jun 23 '16 at 7:41
  • \$\begingroup\$ Note that this is already partially covered by the ppcg design userscript. It is a leaderboard on the question though, but it shows it can be done. \$\endgroup\$ – Rɪᴋᴇʀ Jun 23 '16 at 14:16
  • 7
    \$\begingroup\$ I disagree. This would give an even larger advantage to golfing languages than we have currently. How can we tell users that the language doesn't matter if all "normal" languages get relegated to the bottom of the page (or the next page)? \$\endgroup\$ – Geobits Nov 1 '16 at 17:44
10
\$\begingroup\$

Include Top Askers in tag wiki side panel

Currently the side panel in each tag wiki page shows a list of Top Answerers, which for a Q&A site makes perfect sense. However, here it would be good to also recognise our challenge authors.

I'm not recommending removing the Top Answerers list - it's also very useful to see who is answering for a given tag. The Top Askers would simply be another list of 5 users directly below the first list.

\$\endgroup\$

You must log in to answer this question.

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