Tags for natural languages help in content categorization and searching of posts [closed]

Question

This Stack Overflow post How to describe the closest enclosing element? has been tagged with just english. A tag has to be self-sufficient and it should tell something about the question. Looking at the above post, english seems useless. Should it be removed? I think it should be burninated.

However, tags related to other languages seems to improve content categorization and help in better searching of related posts, for example:

• hindi x 75
• arabic x 596
• hebrew x 251
• farsi x 21
• persian x 71
• tamil x 45
• turkish x 60

Should we keep them or remove them? I think they should be left as-is.

Update:

english has been burninated.

Answer 1

Language tags are certainly useful.

• Some questions are language specific. People asked for examples: trivial examples are changing the plurality of a word or anything relating to natural language processing.
  In fact, the language may be required for these questions.
• Languages have their own Unicode blocks. They can have tags.
• Languages have different encodings. As much as we love utf 8, sometimes you're dealing with an old file, or a legacy system.
• Internationalization and localization are just features.
  I don't need to get i18n and L10n right to write a Turkish site if that's the only language I'm supporting. I just want my language to work.
  ^{Here's a per-language tool: gibberish-to-Hebrew translator - (Hebrew site)}
• font, rtl, utf-8, etc can replace the language in many cases, but sometimes they assume you already know the cause of your problem.
  For example: Suppose my Japanese site "isn't working" on a Kindle Touch. Did I mess the encoding? Am I using a wrong font? Does the reader even have any Japanese fonts? If not, would a web font work? Is there a third-party solution to get it to work?
  When asking a question, I probably don't know any of these.
• Some "language features" are unique, or at least appear unique to their owners.
  In Hebrew we have nikud and optional vowels, for example. If I had problems with these, I wouldn't bother searching for a general solution.
  Unicode collation algorithm do make these problems appear simple, but at the end require per-language implementation.

I agree that a more general tag is better, but it is a mistake to ban or remove these tags, including English.

Answer 2

Yes.

It also seems like a good indicator of a lot of off topic or not constructive stuff. Make sure to flag/vote to close before removing the tag.

Answer 3

The tag wiki is also completely useless. It tells us about the history of the English language. Interesting stuff, perhaps, but not what I come to SO to learn. It should tell us when the tag is appropriate to use on Stack Overflow: what the point of the tag is.

English is a West Germanic language spoken originally in England, and is now the most widely used language in the world. It is spoken as a first language by a majority of the inhabitants of several nations, including the United Kingdom, the United States, Canada, Australia, Ireland and New Zealand.

It is the third most commonly spoken language in the world in terms of native speakers, after Mandarin Chinese and Spanish. But it is more commonly used as a second language than any other, which is why its total number of speakers -- native plus non-native -- exceeds those of any other language.

ISO Codes

• ISO 639-1: en
• ISO 639-2: eng
• ISO 639-3: eng

Links:

• English language on Wikipedia
• List of countries with where English is an official language
• List of countries by English speaking population

Answer 4

In most cases i'd say the spoken language tags misused, but there are some cases where it would be appropriate.

For example for cases surrounding localization issues, it would be prudent to list the actual language you're having issues with.

For example if I had a regex to match a given pattern, which works with English input but doesn't with Hindi input, then for the sake of being easier to find later by someone else with the same problem - the spoken language tags become relevant.

My vote is for emphasizing that the spoken language tags should only really be used for localization issues where the language of text actually affects how the code behaves.

Edit: Better example may be when using things like String.ToLower() in C# which uses the current culture, and different cultures have different rules on which characters change when doing upper <-> lower case conversion. In which case the language and/or culture in question would be relevant to the question (even if the answer is the same for different cultures)

Answer 5

I say burninate them all, as they don't really serve any purpose on SO.

We need a burnination signal in here.

Failing burnination, there's few enough that we could clean them up manually.