Proposal to lock the Language Showcase as historically significant

Question

I recently revisited the following classic challenge: Showcase your language one vote at a time. The gist of the challenge is as follows:

1. An answer is a set of program snippets, each with a unique length.
2. The maximum length and maximum number of snippets are given by the answer's current vote total.
3. As an answer gains votes, it earns to right to contain more/longer snippets.
4. popularity-contest the highest voted answer will win

Also notably, there is no restriction on what the code does. You are allowed and encouraged to post "anything worth showcasing." Additionally, answers are not required to contain any code at all (and indeed, they aren't allowed to until the answer has an upvote).

I believe the Language Showcase, despite its popularity, is not within the current scope of this site, and thus it should be locked as historically significant. For reference, the historical significance lock reason is shown below:

This question exists because it has historical significance, but it is not considered a good, on-topic question for this site, so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed.

In my opinion, if such a challenge were to be posted today, it would be closed. This challenge places a length restriction on code, yet does not provide any goal as to what the code should do, other than to be interesting.

Examples of similar challenges which have been closed

I'm referencing these early in the post because they are the basis for a lot of the existing meta literature.

1. The Tweetable Mathematical Art [pop-con] challenge involved producing an interesting image with a code length restriction. Similar to the Language Showcase, the goal was simply to make an interesting image. There was little restriction on what made an image valid. This challenge has since been closed as too broad and locked as historically significant.
2. Paint the Mona Lisa in 1kB of code [pop-con, deleted] involved producing an image with a limited amount of code. The goal was "visual resemblance of the output image to da Vinci's masterpiece." This is a stronger "goal" than either the Math Art or Language Showcase challenges, yet this challenge was closed as primarily opinion-based, and then deleted by a moderator.

Relevant meta posts

1. Guidelines for posting and closing popularity contests (12 Jan 2016)

   This post was the first to introduce the term "objective validity criterion" as a requirement for all challenges. An objective validity criterion lays out the minimal requirements that each answer must meet. One important question is: does the Language Showcase have an acceptable validity criterion? I believe that the simple requirement, that snippets be of a certain length, would set the bar way too low as far as validity criteria are concerned. This opinion is backed up by this meta answer, which deals primarily with the Mona Lisa challenge. This answer argues that the Mona Lisa challenge does not have an objective validity criterion, and, even if you considered "output an image" as the criterion, then the challenge would still be considered too broad.

2. The state of the popularity contest tag (30 Jan 2016)

   This this thread, the community decided to rewrite the pop-con tag description, leading to the meta post I describe below. The accepted answer points out the importance of updating the tag info given the recent evolution of site guidelines, particularly regarding validity. For reference, the Language Showcase challenge was asked an entire year prior to this meta thread, as evidence that site guidelines have changed significantly between then and now.

   The second-most-upvoted answer states that "the specification needs to define what makes something worth upvoting" and that "popularity-contest does not make your challenge exempt from the standard rules of what is on-topic here."

3. Updating the [popularity-contest] tag info: suggestions (1 May 2016)

   This is the thread in which the community rewrote the pop-con tag description. A couple very relevant phrases are the following:

   Things that MUST be included in a popularity contest...

   A clear specification of the goal that must be achieved. Questions like "do (this) the most creative way" should be avoided. Creativity should be the tool, not the goal.

   ...Qualities which should be avoided in popularity contests...

   Asking to solve a vaguely defined task in any way that the entrant wants (this will probably make your question be closed as too broad).

4. We need to take an official stance on code trolling (3 May 2014)

   Although code-trolling is just a subset of popularity-contests (a subset with its own, entirely unrelated issues), the accepted answer makes an important point that popular doesn't imply on-topic. I believe this is very relevant given the popularity of the Language Showcase, which currently sits at +419/-12.

   ...But just because it's popular that doesn't mean it's a good kind of question for this site. Even "because it may generate an amazing answer from time to time" doesn't mean it's a good kind of question for this site.

   "But this is a community-driven site! If it gets so many upvotes, the community wants it, and it should stay!" Or should it? The StackExchange network prides itself in its high-quality content. In any case quality over quantity...

   ...There are both list-type and fun questions on StackOverflow which are massively popular and have amazing (and even useful) answers. Nevertheless, they are strongly discouraged and new similar questions will be violently closed and deleted.

---

Need for broader action?

This thread deals mainly with the Language Showcase challenge, although there are potentially several other challenges which fall into the same category. A large number of existing pop-cons predate the above meta conversations by a year or more. There does not appear to be much consistency in which challenges are open and which are closed. For several of the same reasons outlined above, I believe many of these challenges are long-overdue for their closure.

Answer 1

Problems I see with locking the Showcase

• We have practically nowhere else to post what we can post on the showcase. You can always put them in the language's readme on GitHub, but I prefer the showcase because it's a ubiquitous place to post interesting snippets and tutorials. You can find a snippet about pretty much every language there, rather than having to look at the repo or google to find a tutorial for each language you want to learn about.

• We would lose a lot by not having a common place to explain and advocate for new languages. The showcase currently serves as a place to really introduce languages in a very interesting way, revealing one feature at a time, and also to receive feedback on the language. An alternative to this would be to post a tutorial in the language's repo and make it easy to find; I believe this should always be done anyway.

I guess the bottom line is:

The showcase is not an optimal solution, but it's the best we have for now.

I feel like the showcase currently stands at an awkward half-ground between what you could put in a readme and what's on-topic here in PPCG. But until we have a better solution for showcasing languages, I think it is best that we don't lock the showcase; it would cut off the best option we currently have.

This is only my opinion; it is by no means definitive. Feel free to comment with your own opinion. Also, I'm writing this while tired, so please don't mind if it isn't very high quality.

Answer 2

The showcase is part of a unique category of pop-cons that should be on topic.

Before I explain this, consider for a second why we banished code-trolling, underhanded and now are considering eliminating popularity-contest: site quality. During the great code trolling apocalypse, Doorknob and Martin made excellent arguments against it that more or less resolved around "even though it's popular it spawns very low quality content."

Since then, I feel a lot of issues resolving around popularity contests is the fear that poor popularity contests will spawn poor loop-holey answers that nobody likes but nobody can really argue against. Although our standard loopholes have cut down a lot of the dumb answers like pulling things from the Internet, when you have subjective contests it's inevitable that people will post answers that don't actually do what the challenge author intended but can squeak by on technicalities and subjective arguments.

Consider the Mona Lisa question. Here's a screenshot if you can't see deleted questions:

Per our rules on valid answers, an answer that does not demonstrate sufficient effort to solve the challenge is invalid. In the case of code-golf, this is usually pretty easy. But let's take a look at the Mona Lisa spec and consider how one would categorize an answer as involving sufficient effort:

• Is under 1KB
• Doesn't violate any of our site meta rules
• Looks like Mona Lisa

The last rule is obviously the problematic one -- we have no objective way of saying whether something is similar to the Mona Lisa, and therefore no objective way of ruling whether an answer is invalid. When the Mona Lisa question was first proposed, I argued that I could post an answer that outputs an apple and spend all day arguing that an apple is exactly the same as the Mona Lisa and we all live in the Matrix. Obviously this isn't desirable -- indeed, this is exactly what we want to avoid.

Now let's take a look at the showcase. Although there are currently some wording issues that make it seem more subjective than it is, this is all that is required to meet the spec sufficiently:

• The language you choose meets our requirements for programming languages.
• For every upvote you get, you must have a code snippet of length n in that language where n is the number of upvotes at the time of the upvote.

This is very minimalist but perfectly objective. The content of the snippet does not matter for whether an answer is valid, just that they're there. As a result, there isn't much room for bad answers: if they were posted, they could be swiftly deleted since they violate objective rules.

Granted, the showcase is an unusual challenge. But it is definitely not the only one that has this property:

• Programming Languages Through The Years
• Images with all colors
• Make your language unusable
• Play a song for me

Compare these popcons to ones with subjective validity criteria such as:

• What? No error?
• Sort a list and write some english!
• Weirdest obfuscated Hello World
• Generate an understandable sentence

(Note that I'm only really considering challenges from 2013/2014 onward -- anything before that was back when we didn't have enough of an audience for it to really matter.)

The difference is very striking. The former have almost no such loop-holey answers, and most importantly any that were there were quickly deleted. The latter have lots of controversy and some answers that are pretty cheaty and just in general low quality. I could list a lot more sample questions but I think the point stands.

With minimalist objective validity, leaving the subjectivity for the voting, it's clear that we can have fun high-quality pop-cons with high-quality answers and an objective way to cull the low-quality answers. Since this is the only concern with pop-cons in the first place, it makes little sense to nuke these questions. If our rules don't allow us to have high-quality popcons like this, then something is wrong with the rules because they are hindering us, not helping us.

Answer 3

Closing this challenge is pointless even if out-of-scope

This challenge is still opened and…

…it still receives interesting answers showcasing new languages. I still receive upvotes semi-often on that challenge suggesting that people are still interested in seeing languages snippets and explanations.

What closing this challenge would do:

Absolutely nothing beside stopping new languages from being showcased. I haven't seen any recent new answers of very low quality to that challenge (though I'm sure there are some at times).

The only thing closing this challenge would do is satisfy the few people who feel like this challenge is out-of-scope.

But as Mego said in a comment, there is nothing wrong with having out-of-scope challenges kept open if they are liked, attract answers that still interest people of the site, and don't attract much low quality content, like this challenge.

Answer 4

No, don't close it, don't lock it

If it ain't broke, don't fix it.

Unless there are severe quality issues, don't let bureaucratic mumbo-jumbo get in the way of a challenge that many people are clearly entertained by on a site that (I believe) is, at-the-core, about entertainment — about enjoying programming.

Answer 5

The showcase is too broad

As I've said in chat yesterday, I like the showcase, but it is too broad for our site:

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

Both apply here.

• Without any tangible goal – apart from showing off what the language can do – the answers can do literally anything. The only restriction is that the byte count is limited by the answer's score, but that doesn't really restrict what can be done in an answer, only when it can be done. Like with Make the most useful program within 100 characters, which was closed in 2014 for the exact same reason, there are simply too many possible answers.

• Good answers to a popularity contest are the popular ones (I assume), which right now stand at 355, 164, 155, 139, and 121 votes for the top five. All these answers are allowed to contain more than one hundred snippets, pressing against the 30,000 character limit that applies to all answers on Stack Exchange. The format explicitly encourages adding more and more snippets to the same post, so yes, good answers are too long for our format.

That makes discussing whether the showcase fits our evolving vision for popularity contests and the associated and controversial hurdles (e.g., objective validity criteria) rather pointless, since it doesn't make the showcase any less broad.

Showcases are list questions

We're not the first site that ran into this specific problem: we've created a resource that, while certain useful and appreciated by many, doesn't really fit into what we do here. The showcase asks what can your language do in x or less bytes, making it a list question (yes, the have their own tag on Mother Meta) and not really a programming contest.

While not really fitting for the site, it would make no sense to destroy these resources. One kind of list question that (mostly) survived on Stack Overflow is the The Definitive X Book Guide and List family, which have a wiki lock:

This question's answers are a collaborative effort: if you see something that can be improved, just edit the answer to improve it! No additional answers can be added here

While sending a clear message that this is not a typical thread and preventing the addition of comments, answers, and close/delete votes, the answers can still be edited and voted on.

Sadly, that doesn't solve anything in this case, since the current answers are already too long. I'm also not sure if that would automatically make the question and its answers community wiki (which would be the only sensible option if we went this way).

The showcase should be locked

This is a textbook case for a historical lock:

1. The post is Off-Topic or Not Constructive, and
2. The post is stellar, in spite of its off-topic nature, and
3. There are a large number of views, upvotes and inbound links on the post, and
4. The post is contentious; e.g., it has been closed and reopened at least once, or deleted and undeleted at least once

Check, check, check, and check.

I like the showcase, I really do, and many of its answers are educational and entertaining, but making it a programming contest doesn't quite work and doesn't really benefit the showcase.

• The one vote at a time gamification means I can't add snippets when I want to, which will result in less snippets being added in the long run. In addition, answers in verbose and non-popular languages may never get the required votes to start showing off interesting language features.

• The specification discourages creating more than one answer in the same language, while the showcase, as reading material, would benefit from all users contributing to all languages they can contribute to.

• There's no way to view a single answer at a time, and the answers are huge; the first page alone weighs over five megabytes. Just trying to scroll past the accepted answer is painful on some devices.

• The 30,000 character limit effectively prevents achieving the goal of showcasing as many language features as possible. While none of the answers seems to have hit the character limit yet, four of the top five also haven't been updated in the last 20 months.

TL;DR

The showcase is not a good fit for PPCG, and PPCG is not a good fit for the showcase. Let's lock the thread as historically significant and explore alternative ways of showcasing languages.

Answer 6

No, Unlock and Reopen, but…

As @Calvin's Hobbies and others state, this community isn't about strict rules and bureaucracy, but about fun as long it is good content. However, we do not want (or at least I think we do not want) more questions of this nature posted. I think the solution is to unlock and reopen, but post a mod message on the question saying something similar to the historical lock dialog. This way, we can have our cake and eat it too!

Answer 7

Showcasing languages is not on-topic

Here is Martin Ender's now very-highly voted Mathematica post in its entirety as of when it was posted.

Mathematica

Factoid

Mathematica can perform symbolic manipulation, so variables don't need values to work with them.

Note that it contains no code, as required by the rule "Since all answers start off with 0 votes, answers will initially contain 0 code snippets." I wonder if it would have stayed that way if not for Martin's reputation with Mathematica preceding him, the language being well-known, and its being posted in the first 15 minutes. To get off the ground and allow interesting snippets, an answer must be voted on based on basically no content. This is a popularity contest in the pejorative sense of the word.

More than being a do-anything-creative pop con, what makes the challenge out of scope is that it showcases languages, not programming. Many of the snippets are standard features of the language, used exactly as intended. Some are just a built-in. The poster is not finding clever and unexpected uses, but simply listing cool things the language has to offer.

A popular, many-featured, or unusual language will get more votes. Yes, a knowledgeable writer can do better with interesting snippets, but it's still not their coding skills that are being judged. The challenge doesn't require snippets to be golfed. Some are, to fit in the character limit, but others don't even bother cutting spaces once they have enough votes.

PPCG is about programming contests. People compete to write code that performs a task as efficiently or effectively as possible. Describing why languages are cool is not within the purpose of the site¹. It's out of scope not (just) due policies about pop cons and objectivity, but for the same reason that an SE about bicycles wouldn't take a question about cars. Or, more closely, a site hosting video game speedruns wouldn't take a Let's Play.

The whole spirit of "Here's something I find cool! What do you find cool?" feels out of place, like something out of a discussion forum thread. Keeping the showcase open wouldn't just require changing policies for pop cons, but fundamentally shifting what the site is about.

¹We have a place to explain why a language is interesting to golf in that I posted, but it's on meta and specifically about golfing.

Answer 8

It's not a grey area, it's off topic

Much as I love the challenge, it doesn't meet our current definition of on topic. I also loved the tweetable mathematical art challenge (and got most of my rep from my 6 answers there...) but it's not on topic so it was closed. Similarly to that challenge, I think the language showcase should be preserved, but closed. One way to do this is with a historical lock. This seems easier than transferring the information elsewhere.

Once closed, there will be an incentive to set up an alternative

There is clearly demand for a place to show off a language as the challenge is still active, so once this way is removed the community can focus on coming up with new, better suited ways to host showcases.

Demand to keep it open would require a scope change

If there is community support for changing what is on topic and redefining our scope, then it could be possible to make this on topic. Personally I think that would be changing scope for the wrong reason, and I can't think of a way of changing the scope that would permit this challenge without also permitting lots of challenges we wouldn't want here.

Answer 9

While the debate regarding the showcase's ontopicness is still open, the majority of the community seems to agree – as do we moderators – that it is too broad for our scope. As a result, I have closed and locked the challenge in question.

Unfortunately, that means that we do not have a place for showcasing languages right now, and we should seek one as soon as possible. Leaving the challenge open would only slow this process down and – as I have said in my previous answer – the showcase itself will benefit from a more suitable format.