4
\$\begingroup\$

Looking at the last 3 pages of tags (sorted by popular), we can see a whole bunch of tags which are used on 5 questions or less. All of the questions with 0 questions (page 9 and some on page 8) are all synonyms of other tags, so we don't have to deal with them. However, that still leaves almost 2 pages of tags which are used very sparingly. Notably, a large number of these tags are in reference to specific languages (such as ). I appreciate the desire to make it easier to find questions on certain topics, but the tagging system isn't designed to be hyper specific.

Therefore, I propose: we sort these tags into three "piles": keep, synonym and delete. For those we decide to remove, trusted users can edit them out of questions, and the system should remove them in due time. However, this discussion is a two-parter:

  1. Let's decide which low-usage tags we want and which we don't
  2. More generally: should we remove all tags with 5 or fewer questions that refer to a specific language?

I choose 5 simply because I consider that a reasonable threshold. Feel free to suggest your own thresholds if you disagree (please be reasonable).

To be fully clear, the tags that would be removed if the answer to 2. is "yes" would be:

, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,


Aside from the language specific tags, there are a whole number of tags with not many questions that could either be made into synonyms of more popular tags or just removed outright. To fully catalogue the tags with 5 or fewer questions, here they are along with my opinion with what to do with them. Feel free to suggest alternatives for any and all tags:

Synonyms

Burninate

  • . I don't see this being a helpful tag to distinguish between questions
  • . One challenge, closed as a duplicate
  • . My personal view is that this should be a meta tag, but I'm open to disagreements
  • . Two challenges: One language-specific for git on +2 and one, not language specific, on -9
  • . One closed challenge on +1, and one challenge about reverse engineering. I don't think the tag is particularly necessary
  • . Unless more challenges from 100 Little Keg Exercises are posted, I don't think this is especially necessary. Either way, it can be merged with
  • . None of the challenges tagged are intrinsically linked with astronomy, rather inspired by it. Personally, I can agree either way with this one.
  • . 3 of the 5 challenges are closed, and the other two are tips where would work in substitute.

Keep

All other tags with 5 or fewer questions not mentioned are already synonyms.


I appreciate that this is a lot of tags, and that you might completely disagree with my opinions on any and all of the tags I've mentioned. I'd love to hear the community's views on what we should do with these tags. Whether you'd like to discuss all or one, please post either a comment or an answer, so we can clean up these tags.

And, as a reminder:

We can remove tags from questions, and they'll get auto-removed. Mods can burninate those which are too widespread to easily remove, and certain users can vote on duplicate/synonym tags.

\$\endgroup\$
3
  • \$\begingroup\$ Genetic algorithm (a machine learning algorithm) has nothing to do with bioinformatics. For cards, my bet is card-games (one out of three already has it; the other two are about shuffling a deck of cards, and some other challenges related to card shuffle are already tagged as such). \$\endgroup\$
    – Bubbler
    Commented Aug 31, 2020 at 0:34
  • \$\begingroup\$ @Bubbler My mistake, I only briefly skimmed the tag Wikis for genetic algorithm and bioinformatics. I agree card games would probably be a better fit than permutations \$\endgroup\$ Commented Aug 31, 2020 at 0:36
  • \$\begingroup\$ Also, automaton is about finite-state automaton, which has nothing to do with cellular automata. I think it should be kept as-is, or renamed to "finite-state-automaton" or such. \$\endgroup\$
    – Bubbler
    Commented Aug 31, 2020 at 1:09

1 Answer 1

3
\$\begingroup\$

We need to be more careful with language tags

I have a feeling that deleting a language tag for having less than X questions/challenges with it (whatever the value of X is) is not the right way to go. Rather, I want to propose a few challenge types that do warrant a language tag.

Keep the tags which have...

  • a well-received challenge which only accepts answers written in that language, or
  • a well-received challenge whose core is highly inspired by a unique or interesting feature of that language

This is just a rough proposal, and I'm open to suggestions about the wording or details. The rationale is that Read out C variable declaration, The cell at the end of the rainbow, and Sugar free syntax, as well as most s and some language-feature-based s, are all valid use cases for a language-specific tag in a general challenge.

Specifically, questions or or challenges do not warrant keeping the language tag.

\$\endgroup\$
3
  • \$\begingroup\$ I think of the tags listed above this would save klein, lost and jelly. And looking at them I do think that they are all worth saving. (Although I might be biased since both Klein and Lost are my languages.) \$\endgroup\$
    – Wheat Wizard Mod
    Commented Sep 2, 2020 at 20:52
  • \$\begingroup\$ To me, cubically, x86-family, and golf-cpu also pass the test (in addition to the three you mentioned), and ruby is around the border. \$\endgroup\$
    – Bubbler
    Commented Sep 2, 2020 at 23:44
  • \$\begingroup\$ Going by these criteria, we could also get rid of the basic. The tags pyth and befunge would be on thin ice as well. \$\endgroup\$
    – Sara J
    Commented Sep 19, 2020 at 18:35

You must log in to answer this question.

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