29
\$\begingroup\$

Comparing the vote tallies on Proposal to lock the Language Showcase as historically significant before and after the Showcase of Languages was closed and locked leads me to the following conclusion.

  1. The community agrees the showcase is not an optimal solution. While it is the best/only thing of its kind on PPCG, there are issues that can and should be addressed.

  2. The community agrees in principle that the showcase is too broad by our current standards.

  3. The community does not want the showcase closed and/or locked.

Going by the rules, points 2. and 3. are contradictory, but exceptions for highly popular and useful resources that already exist and the community wants to keep updating are not unheard of. For example, after getting closed, reopened, protected, unprotected, locked, and unlocked more times than I can be bothered to count, The Definitive C++ Book Guide and List finally settled on a special kind of lock (Wiki Answer) that combines all answers into one, prevents adding additional answers, but neither closes the question nor makes the answer uneditable.

While we cannot do the exactly the same for the showcase (combining all answers into one would create a ton of issues and solve none), we could still create an exception that fits our case. By this, I don't mean simply ignoring any issues with the showcase being a PPCG challenge and the showcase itself, but coming to a consensus what this exception should cover, and how we can at least partially address those issues for the benefit of the site, the showcase, and the community.

For reference, these are some of the issues that surged. Note that not all of them are considered issues by all (or even a majority of the) members of our community.

  1. By our current standards, the showcase is too broad to be a .

  2. The showcase isn't really a programming competition, but a way of showing off features of programming languages. There is no problem to be solved and no sensible way to compare answers.

  3. The one vote at a time aspect slows down and hinders the addition of snippets. Verbose languages may need a substantial amount of upvotes before they can do anything interesting. Not being able to do something interesting may mean these upvotes will never come.

  4. Answers are owned by a single member of the community, the challenge explicitly discourages creating more than one answer for a single language, and editing/adding code to another user's answer is frowned upon.

  5. Many answers are quite long. Some are huge. The high amount of pictures makes the first pages (by votes) a strain for slow internet connections and limited data plans. The question itself is pretty big and contains a directory that is updated manually.

  6. Answers are capped at 30,000 characters. If we manage to get more useful additions to the top answers, that will become a problem.

So, can we agree on making a clearly defined exception for the showcase? If so, what should be the terms of that exception?

\$\endgroup\$
5
  • 1
    \$\begingroup\$ What is your basis for point 2? The highest voted answer to meta.codegolf.stackexchange.com/q/10457/26997 does not talk about broadness. The next highest voted answer says showcases should be in scope, thus they are not too broad (or we shift the broadness border to include them). The next two highest voted answers are aware of the scope issues, but indifferent to them. \$\endgroup\$ Nov 22, 2016 at 23:28
  • 1
    \$\begingroup\$ My answer arguing that is is too broad stood at +18/-9. Not a single comment (or answer for that matter) was contesting my point of view. That seemed pretty clear cut to me. \$\endgroup\$
    – Dennis
    Nov 22, 2016 at 23:39
  • \$\begingroup\$ What about the 25 people who agree with quartata that showcases should be on topic? They can't be out of scope if they're on topic. \$\endgroup\$ Nov 22, 2016 at 23:57
  • \$\begingroup\$ The showcase should enjoy sui generis exclusion. \$\endgroup\$ Nov 23, 2016 at 0:07
  • \$\begingroup\$ @Calvin'sHobbies, yes they can. codegolf.stackexchange.com/help/dont-ask lists being on topic as the first of a series of filters, and the showcase clearly fails the "Your questions should be reasonably scoped" test. \$\endgroup\$ Nov 23, 2016 at 9:00

6 Answers 6

25
\$\begingroup\$

Let's make an exception

This seems to be what the community as a whole wants. I don't think this is an optimal solution (and would personally to prefer to leave it as it is), but since the showcase is both too broad and dearly beloved by the community, it seems like a viable compromise.

Below, I list some potential changes to the showcase that could both make it a better fit for PPCG and benefit the showcase itself. These are my own, personal ideas; they are not fleshed out, and feedback, improvements, and additional ideas are welcome.

Make it a non-challenge question

We do allow these, and the showcase is by far more similar to questions than to our usual programming contests. This should encompass unaccepting the answer with the highest vote count, removing the tag, and adding a dedicated tag which makes clear that this will be the only question of its kind, at least without a preceding meta discussion.

This would also mean removing the one vote at a time rule; additional snippets could be added at will and would not be restricted to one snippet per character count. I don't think character count should be a limiting factor at all; it hinders showcasing intentionally verbose languages.

Make it a community wiki

This is a perfect fit.

Questions should be manually converted to community wiki when they are marginal fits or 'list of X' questions that contain enough value to avoid deletion.

The showcase should obviously not be deleted, but it is a 'list of X' question, and considered too broad or even off topic by some members of the community.

Converting the entire thread to community wiki means that all answers will cease to be the "property" of a single member of the community; all users that cross the required threshold are free to add snippets of their own, or modify and/or remove low-quality contributions of other users.

In case you're worried, this will not affect the reputation you have already earned.

Combine answers in the same language (whenever possible)

Ideally, we would have one page per language, but there's no way to make that happen using PPCG's Q&A format.

Consolidating multiple answers in a single language into a single post should be done whenever possible, both to make finding and reading about your language of interest easier and to avoid multiple near-duplicate snippets.

One issue with this idea is the 30,000 character limit imposed on answers. We could request the limit to be raised; this would also benefit answers in exceptionally verbose languages which tend to not fit into an answer even for the simplest challenges.

Use Stack Snippets to fold answers

Everything except the language name and a factoid could be folding by wrapping it into a stack snippet. That would condense each page into 30 stubs, which can be expanded with just two clicks. This serves multiple purposes:

  • You don't have to load images of answers you're not interested in reading or have already read. This is a huge deal on small screens or limited data plans.

  • Stack Snippets can be expanded to take up the entire screen, making long answers substantially easier to read.

  • Stack Snippets can contain more than just Markdown. An animated GIF is no longer the only way to show an animation, snippets could be fiddled with inline, etc.

Unlock and reopen

I'd rather keep it locked to make clear that this constitutes an exception to our rules, but a notice in the question body will have to do. There seems to be no fitting lock reason, the closest fit being be Wiki Answer (which prevents adding additional answers).

We should also replace the manually curated answer list with a Stack Snippet.

Keep searching for better alternatives

As before, I still believe that the showcase deserves to be a proper wiki, quite possibly outside PPCG, free from restrictions on character count, size, Markdown, JavaScript, pagination, etc.

While there are existing alternatives – e.g., Esolang could showcase the recreational programming languages – the community doesn't seem to consider it a sufficient replacement. But once we find one, we could still revisit locking the showcase.

\$\endgroup\$
12
  • \$\begingroup\$ Relevant MSE feature request for folding away answers. \$\endgroup\$ Nov 22, 2016 at 19:51
  • \$\begingroup\$ If this becomes the highest voted answer, which one(s) of the options you present will be the result of the exception? \$\endgroup\$ Nov 22, 2016 at 20:08
  • 7
    \$\begingroup\$ "This would also mean removing the one vote at a time rule." I'm absolutely in favour of this, but then the last restriction on the challenge is gone. Since the showcase is incredibly broad anyway, this probably won't matter at this point. But it raises the question, whether it should be replaced with some content guidelines for what exactly the answers should be (so that they retain their showcase character without derailing into a learnxinyminutes.com mirror). Should answers be encouraged to present snippets of increasing complexity? Should they focus on golfing? On weird features? ...? \$\endgroup\$ Nov 22, 2016 at 20:08
  • \$\begingroup\$ @Calvin'sHobbies I believe Dennis means to do all of the things he suggested. \$\endgroup\$ Nov 22, 2016 at 20:09
  • 2
    \$\begingroup\$ @Calvin'sHobbies Yes, indeed. It would be pointless to make it a community wiki if we're not going to unlock it. \$\endgroup\$
    – Dennis
    Nov 22, 2016 at 20:11
  • 1
    \$\begingroup\$ It just wasn't clear if this was saying "let's do all of these things as part of the exception" or "let's call it an exception and discuss exactly what that means in another meta post" or something. \$\endgroup\$ Nov 22, 2016 at 20:18
  • \$\begingroup\$ Doesn't dropping the one vote at a time rule make it much more broad, which was the complaint in the first place? Not to mention mess up the structure of all existing answers (or all new answers, depending on how you view it). \$\endgroup\$ Nov 22, 2016 at 20:23
  • 1
    \$\begingroup\$ The one vote at a time rule isn't really a restriction for golfing languages that can do a million things in a handful of bytes and makes showcasing verbose languages a lot more difficult. I guess it technically makes it broader but it wouldn't be by much, and I don't think that's an issue if we come to view the showcase as a wiki rather than a competition. \$\endgroup\$
    – Dennis
    Nov 22, 2016 at 20:28
  • 2
    \$\begingroup\$ A random thought after reading "the showcase is by far more similar to tips questions than to our usual programming contests": A per language question "Examples for golfing/using <language x>. Please post one example/feature per answer. ... " \$\endgroup\$
    – Laikoni
    Nov 22, 2016 at 20:34
  • \$\begingroup\$ If you make it comwiki, we won't get rep pings anymore, so how will we know when we get an up vote so we can add another snippet? \$\endgroup\$
    – AAM111
    Nov 25, 2016 at 15:48
  • 2
    \$\begingroup\$ @OldBunny2800 I'm also proposing that we remove that rule, so you'll be able to add snippets whenever you want to. \$\endgroup\$
    – Dennis
    Nov 25, 2016 at 15:52
  • \$\begingroup\$ Rather than using stack snippets, what about limiting each post to ~10 of the best snippets? This keeps posts self-contained, without taking up massive amounts of space... \$\endgroup\$
    – FlipTack
    Dec 4, 2016 at 11:00
12
\$\begingroup\$

Make a new category

If the site benefits from this challenge, it should benefit from others like it. What if someone posts another showcase challenge that would be just as popular and well received, but never gets a chance to gather a following because it gets closed immediately?

We should think about what qualities make such challenges valuable and create a category for them with specific validity criteria. At least then, all challenges abide by the same rules without special exceptions, even if only retroactively.

Of course, it's not easy to extract and delineate these qualities, but we should try. I don't make any proposal what this category should be -- I leave that discussion to the community. I only ask that it not be gerrymandered specifically to include the programming language showcase.

\$\endgroup\$
3
  • \$\begingroup\$ This is essentially what my answer was trying to do but slightly more generally. \$\endgroup\$ Nov 23, 2016 at 2:23
  • \$\begingroup\$ One problem with additional showcases is that they probably would wind up being duplicates of the original one. The difference between showcases and challenges is that the former don't have a specific goal, and you can't really put a big spin on show what your language can do. \$\endgroup\$
    – Dennis
    Nov 24, 2016 at 18:10
  • \$\begingroup\$ If they are, close them as duplicates of the showcase. \$\endgroup\$
    – AAM111
    Nov 25, 2016 at 15:42
3
\$\begingroup\$

I'd like to enumerate all of the options I see so far, along with their disadvantages: (Note that this is all under the assumption that the Showcase is too broad, )

  1. Close the Showcase
    • The problem with this is that it is a great resource to show off new languages
  2. Keep the Showcase
    • Due to the 1-byte-per-vote rule, new posts have little ability to show off (which is the whole reason we were keeping it open)
  3. Remove the 1-vote-per-byte
    • This invalidates all current answers. It'd be like removing the from a challenge: the answers would still work, but the goal they were aiming for is gone.
  4. Close the Showcase, and open a new Showcase
    • The biggest problem I see with this is that we've fragmented our language showcase: We'd have some really great snippets on the old Showcase, and other great ones on the new showcase.

I didn't consider options of "Redefine what we consider too broad/on topic", as we haven't really gotten a solid definition of what the new rules would entail.

This isn't a "We should do X", but adding to the discussion.

\$\endgroup\$
2
\$\begingroup\$

Reopen, don't change the rules

What I'm arguing for is:

  • Call the Showcase an exception to our general rules on broadness.*

  • Unlock and reopen the Showcase ASAP.

  • Don't change the rules or the answer format.

  • Do not make it a community wiki without solving the issue of "rep pings".

  • Keep searching for better alternatives if you feel that is necessary.

The Showcase is plainly a popular challenge that people want to stick around. Whether it would be the only exception ever or if a special tag should be made for it and stuff, I can't say. Mostly, I'm advocating restoring the Showcase to the state it existed in for nearly 2 years.

*Or call it a valid challenge based on the fact that there was never a clear concensus to close it in the first place, the difference is semantic.


Here I've gone through the major points in Dennis' answer and added my suggestions/desires:

Make it a non-challenge question

I'm not opposed to retagging or unaccepting the answer. I don't feel it's necessary but it's not a big deal, merely a semantic difference.

I do not agree with:

removing the one vote at a time rule; additional snippets could be added at will and would not be restricted to one snippet per character count.

This breaks a golden rule of PPCG of not changing a challenge once many answers are present. It may not invalidate existing answers, but it will make their structure confusing to anyone reading the updated challenge, and new answers will look different. Basically it will cause incoherence, not an untenable amount, but that's not something we tend to intentionally inflict on challenges.

Also, removing this fundamental rule of the challenge makes it more broad, which was the reason it was closed in the first place. So it seems a contradictory notion coming from anyone against broadness. I get that the desire is to turn it into a general showcase area, but updating all the answers to that metric is something that I frankly don't think anyone is going to take the time to do (and I don't blame them).

(I can understand if the issue is that too many edits to the Showcase are bumping it too much so it hogs active questions, but I'm not sure I've ever heard that as a complaint, at least recently.)

Make it a community wiki

I think this would be ok if this didn't stop the little +10 green rep pings users get when they are upvoted. Without those on this challenge, users will constantly have to monitor their answers to see if they can add a new snippet (assuming the votes rule is kept).

That's annoying for answerers and bad for people reading the answers, since there will be fewer updates.

More fundamentally, I believe a capitalist system where people are incentivized to update answers by rep gains is the best model for growing such a challenge. It's perfectly ok to desire rep in exchange for good content. That's how SE works. However, I could get on board with a community wiki iff it didn't remove the rep pings.

Combine answers in the same language (whenever possible)

This makes a lot more sense with a CW and the votes rule removed, so it doesn't apply much with my assumptions.

(Doing a quality spot check on answers might be a good idea though. And maybe some bits could be culled and such.)

Use Stack Snippets to fold answers

No. This is bothersome and convoluted. Clicking and waiting for stuff to load is not necessarily faster than scrolling. And everyone knows what scrolling is, not everyone is familiar with Stack Snippets used as "spoiler" boxes.

Unlock and reopen

Yes. This is desired by the community.

Keep searching for better alternatives

Sure, go for it. Raise more meta questions about the Showcase or showcases in general or broadness. Migrate things to a wiki. But until there's a plain concensus that the Showcase should be closed/locked/deleted/whatever, it should remain open as it has been for nearly 2 years.

\$\endgroup\$
7
  • \$\begingroup\$ Except that new answers aren't going to be upvoted as much. The FGITW problem is very strong on popcons in general, and especially on this one. \$\endgroup\$
    – DJMcMayhem
    Nov 22, 2016 at 21:48
  • 5
    \$\begingroup\$ The one-vote-at-a-time thing was fun and interesting back when this challenge was fresh and active, but at this point - if we want to keep this as a place for good (and hopefully improving) content - then this rule simply gets in the way of providing that content. \$\endgroup\$ Nov 22, 2016 at 21:49
  • \$\begingroup\$ @DrMcMoylex Sure but there still will be upvotes, since that's the whole basis of the challenge. \$\endgroup\$ Nov 22, 2016 at 21:54
  • \$\begingroup\$ @Martin I get that it has slowed down, but if we're ok with showcases, why not write some new ones and they might be just as popular? I'm far more comfortable with moving forward freshly with new challenges than ravaging old ones. \$\endgroup\$ Nov 22, 2016 at 21:56
  • 3
    \$\begingroup\$ 1. The changes I propose wouldn't invalidate any of the answers, so I don't think this breaks our golden rule, at least not in spirit. The structure of the existing answers may be a little confusing, but if they're community wiki, we should have no problems with clearing them up. 2. "Rep pings" are a non-issue if we remove the one vote at a time limitation. At this point, it doesn't affect some answers and adversely affects all others. 3. Finally, and I only bring this up because you wrote it in bold, it has been like that for X years is not a good reason to (not) do something. \$\endgroup\$
    – Dennis
    Nov 22, 2016 at 21:58
  • \$\begingroup\$ @Dennis Call me cynical but I'm not convinced people will continue to contribute as much if there is no rep to be gotten out of it. And I understand about the rule change but it seems really unfair to repurpose such a highly voted question that has been around so long. Forget what it means to me, those ~400 voters voted for the question as it was, not as what you want to turn it into. \$\endgroup\$ Nov 22, 2016 at 22:10
  • \$\begingroup\$ @Calvin'sHobbies People contribute on meta all the time, and there is no rep to be gained. I agree with Dennis - "that's the way it was" is not by itself a convincing reason to do anything. \$\endgroup\$
    – user45941
    Nov 22, 2016 at 23:29
-1
\$\begingroup\$

Obsolete the rule once for all

There are either too many possible answers, or good answers would be too long for this format.

One answer per language is not too many.

95 snippets per answer is not too long. Note that we have less than 200 answers with more than 95 votes on the whole site. The answers are not really expected to be that long.

Allowing doing anything is not too bad. Note that everyone can invent their own languages.

Making a single exception would just bring more arguments in these cases.

We may try to fix the wording. For example:

  • There are too many apparently competitive answers.
  • There are general answers irrelevant to the question.
  • The answers are supposed to be big projects that practically have to consider the quality measures not objectively defined in the question.

But I'm not sure if it works. In fact, we are not a QA site and many of the rules don't apply, and we did't seem to have consensus about what is too broad on our site in the first place. Someone should post another question about the exact terms if we really want to do that.

We have 3 problems now

  1. Is it a bad question?
  2. Do we close it?
  3. Do we make exceptions or change the rules?

I'd say it's not a perfect question. But we are not going to close every imperfect question.

I'm not exactly sure whether we should close it or not. But we don't have to make rules to cover every case. There are cases where we should absolutely close the question, and cases where we should absolutely not, but there are also cases that it might be better to leave for the community to decide. That is, they could be technically allowed but subjectively a bit broad. If we try to make it exactly clear, it's likely our rules will be filled with contrived situations and not the obvious cases nobody is bothered to argue with.

A general rule makes sense only if it also makes some differences to other questions. But we have no idea what a question similar to this one (but isn't a duplicate) would be like. If we try to decide whether we should make exceptions, the implication is, other questions (that we had no idea what they are) similar to this one are all on-topic / all off-topic. That's indeed something ain't broken.

One reason to lock the question is to not encourage similar questions. We don't even have an idea whether we should encourage them.

\$\endgroup\$
-5
\$\begingroup\$

My suggestion, purely based on personal opinion, is that each language in the Showcase should be migrated to its own question, with one community wiki answer, with the original Showcase having a link to each question for each language. The original Showcase question already includes a list of links to individual language answers, which could be changed to links to individual questions. This would mean that, if each of these individual questions were tagged with a (to be created) showcase tag, they could all be browsed easily without loading the whole answer. This would also partially solve the slow internet problem, since when going to the showcase, only the text of the question would be loaded, allowing someone to click on the link for the specific language they want, rather than loading the showcase plus all the images of the first page of answers (currently consisting mostly of mathematica, which has a high proportion of images to text). This could also partially solve the character count issue, since each individual language question could have extra answers added if the character count grew too large.

Summary:

Create a showcase tag.

Migrate each of the showcase languages to their own question, tagged 'showcase'

\$\endgroup\$
1
  • 6
    \$\begingroup\$ While your idea certainly has its benefits, making an exception for one question is very different from making it for 230+ questions. To put that into context, we currently have 6,590 question on main, so one in 33 questions would bear the showcase tag. \$\endgroup\$
    – Dennis
    Nov 29, 2016 at 20:23

You must log in to answer this question.

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