Following specification vs. following language standards

The way I see it, this question follows The Six Subjective Question Guidelines:

  1. It inspires answers that explain "why". Options are given, so I need reasons.
  2. Long answers are expected which provide example implementations and reasons for designing them that way (like why XDocument does not follow specification).
  3. It's not a holy war topic, so constructive tone is expected.
  4. It invites sharing experiences, because there're many standards implemented both by Microsoft and third-parties.
  5. As per above, facts and references are expected.
  6. And finally, it's a very important practical question about designing libraries.

The way I see it, the question perfectly fits Programmers scope: conceptual questions about software development. It is about software architecture which is declared as on-topic by FAQ.

So, the question is: why was the question closed and how to fix it to get it reopened?

  • opinion based close reason matches Gorilla vs Shark scent of the question. Although, both specifications being "solid, widely adopted, time-tested, written by the best professionals", followed by "which standards and recommendations should be preffered? Why? In which cases?" probably makes too broad closure an even better match. One could easily imagine an entire book that answers such a question – gnat Feb 19 '14 at 6:39
  • @gnat I fail to see "Gorilla vs Shark": 1) it's a highly practical question 2) I've provided a specific example of HTML specification etc. Overall, if you think it's "too broad", I've provided a very specific example of HTML specification and listed three specific points of contradiction between the standards. A whole book to resolve these? Really? If "in which cases" is the root of the problem, I can remove it. I just wanted to give some hints on how to write more comprehensive answers if anyone feels like it. – Athari Feb 19 '14 at 7:46
  • per my reading, asking for all imaginable cases ("in which cases...") makes question too broad: "...Please add details to narrow the answer set or to isolate an issue that can be answered in a few paragraphs." Consider editing the question to focus on a specific case instead. Note that answerer capable of providing solid, authoritative solution for a particular case will likely be capable to refer readers to a more general guidance, if there is such – gnat Feb 19 '14 at 8:02
  • @gnat I sort of agree with you that the question sounds too broad, but the problem is, it's closed as "primarily opinion-based", not as "too broad". While I can address the breadth by removing "in which cases" question, it won't solve the primary reason why the question was closed. – Athari Feb 19 '14 at 8:12
  • in your case these reasons look somewhat "intertwined"; it may be that opinion banner is shown only due to weird vote split. It is quite possible that if your edit addresses one concern (too broad), this will suffice for reopen voters to consider other gone, too. You see, when question asks on how to resolve real, concrete, specific design issue, it's typically pretty hard to claim that it's polling for opinions – gnat Feb 19 '14 at 8:18
  • 2
    @Athari There's one detail missing, that would make the question a lot more specific: What's the target audience of your DOM implementation? – yannis Feb 19 '14 at 10:10
  • @YannisRizos Er... Do XML (or HTML) libraries have some specific "target audiences"? Is XmlDocument's audience any different from XDocument's audience? They're pretty much equivalent functionally, it's just API that differs (and performance maybe). I have no idea in what way it can affect target audience. I'm talking about a general-purpose implementation with reading, writing, modifying, querying etc. – Athari Feb 19 '14 at 10:38
  • 2
    @Athari Let's try this another way: Why are you building a DOM library at all in the first place? What's the problem you are trying to solve? – yannis Feb 19 '14 at 10:51
  • @YannisRizos Trying to solve the lack of it, or unsatisfying quality or performance of existing libraries, or them not satisfying some specific requirements, like preserving input on deserialize-serialize cycle, or matching behavior of browsers etc. I don't see how any of these reasons can affect naming conventions or exception safety recommendations. W3C specs cover HTML, CSS, XML, SVG and lots of other stuff. HTML here is merely an example. Search-replace "HTML" with "CSS" and the problems will be absolutely the same. – Athari Feb 19 '14 at 12:10
  • 2
    rev 2 looks good enough, voted reopen. To be fair it still feels somewhat broad, but I can't put a finger at anything concrete to justify for self leaving it closed – gnat Feb 19 '14 at 21:07

There's no way to fix the question.

Programmers site is FUBAR. It's about closing useful questions, not about answering them. If half questions get closed, the site is doomed. Just look at the front page. No one understands what this site is about. Those few who managed to farm enough reputation, enjoy to abuse their power to close every question which is even remotely opinionated or broad. And of course nobody bothers to explain how to fix questions to their liking. Because it's boring and absolutely impossible. Mods don't bother to visit Meta either. Most upvoted questions on Programmers stay open out of pure luck, or after prolonged open-close battles, because most of them are opinionated or too broad or both. Most upvoted questions on Meta are "What the heck Programmers is?"

Run, run, and never return.

  • After 5 downvotes, still no one bothered to answer. As expected. – Athari Feb 24 '14 at 19:51
  • okay I will answer. You appear to be overreacting. I cast reopen vote and I am also unhappy that your question stays closed, but apparently edits you made were insufficient to convince enough voters. If you strongly wish it to be reopened, give it another chance: edit it further (this will re-submit your question to reopen queue). Consider re-checking comments under your question, there may be ideas on how to further improve – gnat Feb 25 '14 at 9:55
  • 1
    @gnat It's impossible to edit the question to someone's liking if they don't explain reasons and just cast votes. THAT is the problem — I have zero information. I've fixed the question as you'd suggested, thanks for help. The only other person who bothered to comment is Yannis Rizos and he disappeared after a couple of comments without giving any practical suggestions. Maybe I'm overreacting, but the website does have problems: of 50 recent questions on page 2, 19 are closed, 22 are downvoted and 24 have zero answers. Most migrated questions are downvoted. That's failure in my book. – Athari Feb 25 '14 at 11:12
  • well, however hard, edit seems to be still the most viable option - it at least guarantees bump for more eyeballs, along with pushing question into reopen queue. I just found an MSO post explaining how meta posting has little effect in cases like this: Is it too hard to re-open questions by the community on SE sites other than SO? – gnat Feb 25 '14 at 13:31
  • ...btw, since you mentioned discussion with Yannis, it's worth warning that he will unlikely cast reopen vote, at least not when question has only 1-2 such votes. He's a moderator and at Programmers they try to abstain of voting close / reopen, except for fairly clear cut cases (and your doesn't look like that yet, sorry). Their votes are binding and they try not to abuse that power – gnat Feb 25 '14 at 16:04
  • @gnat If mods aren't confident enough to use their binding votes, I expect them to help, to explain, to actively participate on meta, to help new users, at least (that's what they all say in election pledges). Yannis asked a few questions and left. Well, thanks for the link. Meta discussion on SO does help, so I expected it to work on PSE too. But looks like I'm out of luck here. – Athari Feb 25 '14 at 19:43

You must log in to answer this question.

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