Are there any non-trivial examples of exact-duplicate answers to non-duplicate questions?

Question

From time to time, discussions arise about the merits of posting identical answers to (allegedly) related-but-not-duplicate questions. In almost all of these, someone brings up the same argument that there can be rather different questions that still have the same answer. Unfortunately, all the examples I've seen are trivial ("42", "blue", etc), and entirely lack the explanations that are at the core of SE.

So I have a challenge: present a question-question-answer tuple — whether on-topic on any current SE site or not, although they should otherwise generally meet standards — that satisfies these three requirements:

• Questions are fundamentally different beyond reasonable doubt.

  In other words, they are not duplicates. This must of course be thoroughly demonstrated, since this is challenging the basic assumption directly.

• The answer includes an explanation sufficient to teach anyone with generally similar problems how to address their problem as well.

  Pretty simple; an SE answer without explanation is fairly poor at best and on some sites will be deleted out of hand. There's no point in arguing for the existence of questions that can be kept open only by the promise of attracting terrible answers.

• The answer needs only cosmetic adjustments, if any, to be suitable for both questions.

  That is, surface details like variable names, specific gadgets, or certain minor points of interpersonal conflict may need to be adjusted, but without changing the meaning of the answer any more than any good editor would.

I used to think it was possible, at least in principle, to have a few rare questions with identical answers that should stay separate. Now I'm not so sure, but I can still be convinced. Here's your chance!

Answer 1

Here is what I believe to be an example. Eclipse is an editor that software developers use write code in many different programming languages such as Java, C, and C++.

Question 1: Eclipse: Auto-align "=" in assignments.

The question asks for a solution to format Java code in Eclipse as follows:

// Without formatting
private final String s1 = "10011";
private final String s2 = "01100";
private final String ones = "11111";
private final String zeros = "00000";


// With formatting
private final String s1    = "10011";
private final String s2    = "01100";
private final String ones  = "11111";
private final String zeros = "00000";

Question 2: Eclipse - "Align Field in Columns" C/C++

The question asks for a solution to format C/C++ in Eclipse as follows:

// Without formatting
int mRect = 0;
int mSquare = 0;
bool isTriangle = false;

// With formatting
int  mRect      = 0;
int  mSquare    = 0;
bool isTriangle = false;

These two questions are different because the first one asks about Java, the second one asks about C/C++. In Eclipse, each language has a different perspective, and perspectives have different features. This means that quite often a solution works for one language but not for the other languages.

For example, the following answer only works for Java:

Preferences → Java → Code Style → Formatter → goto Edit and check the Align fields in columns:

Here is an answer that works for both Java and C/C++, i.e. that answers to both questions at once:

You can use the Eclipse plugin columns4eclipse:

• gratis
• open source
• I've tested it with Eclipse 4.3 x64 and 4.5 x64 on Windows 7 SP1 x64 Ultimate

As a result, I believe that both questions should be kept open (as they are not duplicate) and that the proposed answer correctly answers both questions.

Answer 2

So I just ran into this issue and again it's kind of a tag issue really. Someone asks a webgl question which effectively has the exact same answer as an opengl question but TPTB pretty much never allow a single question to be tagged both webgl and opengl so someone searching for answer in one will never find the answer in the other

Of course in general webgl is programmed in JavaScript and OpenGL is programmed in C/C++ but the question can be one of how it works as in

Question: Why are my particles flickering

I have code like this

  float value = texture2D(someTex, float(intNdx) / someTexWidth)).r;

tags: WebGL, JavaScript

Answer:

Texture coordinates go from the left edge of the left texel to the right edge of the right texel so to use an integer index to look up the value of a specific pixel you need to add half a pixel

  float value = texture2D(someTex, (float(intNdx) + .5) / someTexWidth)).r;

And then

Question: Why does my math change depending on machine

code

  uniform sampler2D texA;
  uniform float texAWidth;  // width of tex A

  ... intOffset ... is an integer offset into texture

  float value = texture2D(texA, float(intOffset) / texAWidth)).r;

tags: OpenGL, C++

Answer:

Texture coordinates go from the left edge of the left texel to the right edge of the right texel so to use an integer index to look up the value of a specific pixel you need to add half a pixel

  float value = texture2D(texA, (float(intOffset) + .5) / texAWidth)).r;

Should one of those be closed as duplicate? They don't share tags. They have effectively the exact same answer.

Answer 3

• Q1: How do I convert response headers into a map? [Java]
• Q2: How do I verify a response contains a header key-value pair? [Java]
• A: Import a library like "x" and call response.getHeaders()

For Q1, A is accepted as the answer; for Q2, another answer that solves via regex was accepted.

For Q1, the solution does exactly what the OP asked, so it is considered the best (easiest) solution. While the answer does also solve Q2, the problem is that bringing in a whole library is overkill when something built into your programming language can do the same job (verification), so their is better (simpler) answer that cannot be applied to Q1. The two questions are fundamentally different (one is manipulation, the other is verification), but they can be solved the same way (with minor insignificant changes specific to the question). While all answers to Q1 could be used for Q2, not all answers apply to Q1, and the accepted answer won't work for Q1 at all.

Since the answers have different values/relevance to each question, they need to be duplicated so that they can be properly quality controlled.