What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 129 Stack Exchange communities.

Consider:

None of those questions really have a good answer to, 'How do I generate an XML sitemap in PHP?' Often they already know about online generators, but perhaps their site is not publicly accessible and they still need a sitemap. In any case, I feel it's been asked often enough that we could give a bit better solution than closing it as not a real question or closing it as a duplicate of something with really lousy solutions.

Update: I also noticed that there are honestly two questions that have been linked together as duplicates when they are not.

  1. How can I generate a sitemap in PHP?
  2. All my content is stored in a database, how can I generate a sitemap? (Oh, btw, I'm using PHP).

These are too very different questions when it comes to solutions, but they are linked as duplicates. The first requires a lot of code, using a crawler, HTML parser, URL resolver, a downloader. The second simply requires a bit of code to know how to generate the URL for each record in the database.

Questions:

  • Is it just too vague to answer on StackOverflow?
  • Should we sort through the duplicate questions and properly link them base on whether content is stored in a database or not?
  • Should I attempt to outline an algorithm for question 1? I've already started but I'm not sure if I should give more detail or not. I may be trying to help a lost cause.
  • If I wrote a viable solution, which question would be the best place to post it as a solution?
share|improve this question

3 Answers 3

I was one of the close voters on that question. Basically, my take was that the user provided no context, demonstrated no effort, and apparently did no research (89,000,000 results, at least one or two thousand are probably useful, no?). Questions like that are impossible to answer objectively within a Q/A - what this user needs is a book on PHP projects, design patterns, also research on existing XML site map generators, research hosting solutions, etc, etc, etc. All this from a 5 line question with no details and a PHP tag?

I look at this as "give me the codez" multiplied by 1,000. The question boils down to something broad and unanswerable like "How do I make Google?". Looking at this , my interpretation is that a question ought to at least be about a specific problem; this particular question is about several topics that aren't specifically "problems" - nothing has been tried, no failure has been encountered, no obstacles are known and described except "I don't know how to make a large, potentially complicated project". What are SO users supposed to answer that with?

So my own answer to your first question, "Is it just too vague to answer on StackOverflow?" - Yes.

share|improve this answer
    
"I don't know how to make a large, potentially complicated project". What are SO users supposed to answer that with? That becomes a valid question if they add the tag algorithm. As a side note, I flagged the question as not a real question myself. –  Levi Morrison Nov 17 '11 at 20:17
1  
I've seen your comments to the effect that you don't find a problem in SO being a magical code generator; fair enough. However, throwing an algorithm tag on a poorly formed "give me the codez" statement is a miserable excuse for a question - it turns the Q/A process into an outsourcing project resource. We're here to help, not do it for them. "A process or set of rules to be followed in calculations or other problem-solving operations, esp. by a computer." - Needs a problem. When we're talking about SO, there is also a scope - the problem cannot be simply "I don't have the codez". –  Chris Nov 17 '11 at 20:21
    
Just to clarify my earlier comment, I do believe asking for an algorithm and asking for teh codez are very different, but my point was that unless they mark it as algorithm, it almost always sounds like give me teh codez. –  Levi Morrison Nov 17 '11 at 20:48

I provided the possible duplicates to that (now closed and deleted) question. The question had three votes for Not A Real Question when it got my attention. I had the option between closing and closing with a possible duplicate. I decided to close with a duplicate so the OP had some pointers to follow over just having a closed question. Given that there is lots of results for sitemap generator and the OP failed to point out any specifics, I still think it was fair to close and delete this question.

Regarding your Questions:

Is it just too vague to answer on StackOverflow?

I dont think it's too vague. It's rather too broad. The OP is not having specific questions about the task, which makes it somewhat unsuited for StackOverflow. The OP cannot expect someone to write the full program here.

Should we sort through the duplicate questions and properly link them based on whether content is stored in a database or not?

If the questions ask about these specific implementation details, then yes. However, the generator part of a sitemap generator should not care whether results are coming from a database or a crawler. This is separate responsibilities. How to crawl a website has been answered as much before as how to fetch results from a database. So it might make more sense to point the OP to duplicates of such questions instead.

Should I attempt to outline an algorithm for question 1? I've already started but I'm not sure if I should give more detail or not. I may be trying to help a lost cause.

Since the question has been closed and deleted now, it was indeed a lost cause. Nonetheless, I liked your answer that outlines a possible approach. Broad answer for a broad question. That's about the best an OP can hope for in an answer. But like I said in the previous paragraph already: how to write a crawler has been answered before. How to generate XML as well. So, it really should be two closevotes.

If I wrote a viable solution, which question would be the best place to post it as a solution?

Go through the existing questions on that topic. Pick the one you think is the most valuable and supply your answer there. Hijack the question. Create a canonical. The better your answer, the higher the chances future questions will be closed against it.

share|improve this answer

Given that a lot of people are having the same problem, it's fine to write an answer.

The answer doesn't have to be a book or complete source code. Your idea of just summarizing the algorithm is perfect. The principle should be that not too much more work should go into the answer than went into the question.

Other users can then edit and embellish your answer over time as if it were a wiki. This is the spirit of a "canonical answer" which identical questions can be closed as dupes of.

In situations like this I suggest choosing the "most generic" question in the range to answer. Feel free to edit the question itself for clarity, grammar, spelling, etc.

Other questions which are identical should then be closed as duplicate.

share|improve this answer
    
I updated my post because of something I noticed, it might be worth looking at again. –  Levi Morrison Nov 17 '11 at 19:15

You must log in to answer this question.

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