I am talking about this answer. It doesn't work with XeLaTeX and LuaLaTeX, because it uses a package which is written by the PDFTeX author and is specifically for LaTeX and PDFLaTeX.

By the time I posted this comment, the answer has got 2 downvotes and zero upvotes. I can think of some possible reasons why it is downvoted:

  • It does not work with XeLaTeX and LuaLaTeX.
  • It does not provide a correct answer or sufficient information: the answerer should state clearly that the proposal only works with LaTeX and PDFLaTeX

For the first reason

Now: refers to a recent comment:

[...] Posting a solution that uses newer technology does not merit a downvote unless it fails to answer the question, i.e. ignoring the specified requirements. This site should encourage posting alternative solutions, some of which use LuaLaTeX or XeLaTeX. [...]

So, we should not downvote answers which don't work for some compilers except when those compilers are mandatory.

First reason is, therefore, not applicable here.

For the second reason

Okay, it is the answerer's fault. However, I edited the answers and added some information to it.

However, one must be in a very bad mood or one must be very rude to downvote such answer for such a reason. I have seen, on this site, hundreds of answers which are even less informative, they are not upvoted: okay, but they are not downvoted.

Second reason is, therefore, not probable.

So, why the downvotes? What is wrong here? And as usual, I want to add a comment:

@downvoters: Please add a comment saying what is wrong. I see no reasons.

This is some fact about the proposal: in Vietnam, in every 10 LaTeX users, 8 uses this proposal, about 1.5 uses the proposal in some documents and babel in some other documents, only 0.5 uses babel totally. I don't have the official satistics, but I am pretty sure this approximation is close. If TeX.SE were a Vietnamese-speaking community, the linked answer would get at least twice as many upvotes as the accepted answer.

In a TUGboat article, Hàn Thế Thành stated both answers as "correct". The proposal in the linked answer is "better for Vietnamese-only documents" and the babel proposal is "better for multiligual documents which use Vietnamese".

  • 3
    We can never really tell why people downvoted, unless they comment on why. But since there are comments saying "it didn't work for me", people might downvote on that basis. But really there's no way to tell. The person who noted that it didn't work with XeLaTeX or LuaLaTeX said they didn't downvote, so I don't think there are any grounds to conclude that someone downvoted because of the pdfLaTeX vs. the others issue. – Alan Munn May 10 '19 at 22:20
  • Voting is for content curation; it's to tell other users that the answer might not be useful. It doesn't matter if most of the userbase is from the US or Vietnamese, as long as the answer is incorrect/missing something, it's expected to be downvoted. The comment for the first reason doesn't contradict the second reason to downvote: yes, alternative solutions are good, but they should be stated clearly that they only work for specific systems. Otherwise, the readers will assume that it works for all systems when it's not. (Your edit has rectified the issue) – Andrew T. May 11 '19 at 11:45
  • i don't downvote but it seems not unreasonable that people who do do so on that, what is the point of posting an answer 2 years after another answer with (I think) essentially the same information. If there is any material difference between specifying utf8 to the vietnam package and specifying it to inputenc, then loading vietnam (as done in the existing answer) the new answer could have said what that is) But this meta question is really unanswerable and should be closed as "opinion based" as only the people who voted can say why they voted. – David Carlisle May 16 '19 at 11:29
  • I could ask the same thing about this answer of mine: tex.stackexchange.com/a/479410 – Henri Menke May 20 '19 at 4:45