-5
\$\begingroup\$

There have been a fair number of architecture questions, and I expect most of them have been subjective to some degree, as many tags, e.g. game-design, will be.

Is there an existing policy to discourage best practices questions in general? Is there a policy to discourage architectural questions in general?

If so, Isn't this is a severe blind spot in the knowledge we try to impart? Is this really worth enforcing when architectural concerns are so crucial to the way we implement projects? From experience teaching, coding details are considerably less important in the long run than the ability to teach overarching approach / philosophy / methodology. Many of us have had answers in this vein from teachers, professors, lecturers who gave us lessons that would go on to last a lifetime.

I'm a firm believer in the concrete, and SE encourages questions that can certainly be answered, which is good. But to kill questions like these off the cuff, seems supremely short-sighted, assuming that's the policy.

\$\endgroup\$
7
\$\begingroup\$

Is there a policy to discourage architectural questions in general?

No, we have the text "engine architecture" in the "What topics can I ask about here?" section on the 'on-topic' help page. There is also a healthy tag.

Is there an existing policy to discourage best practices questions in general?

Yes. There was a discussion here on meta, which linked to a discussion on meta.se.

Essentially, the consensus is that "best practices" question are often too broad, that "best" is subjective, and that askers often ask for best practices that would apply to their situation.

The issue too broad "best practices" questions is highlited in the "What types of questions should I avoid asking?" section of the 'don't ask' help page:

Your questions should be reasonably scoped. If you can imagine an entire book that answers your question, you’re asking too much.

Yes, one could give an answer to these broad questions, but would they cover it enough? Would all the aspects of the question be answered, for the user that asked the question and the future visitors?

Isn't this is a severe blind spot in the knowledge we try to impart?

I don't think so. The scope of our site is to help users with actual problems, not create tutorials for them, or write books on a topic for them. There are other places to ask for that kind of knowledge and share that kind of knowledge (e.g. reddit, gamedev.net, personal blogs, etc.).

Is this really worth enforcing when architectural concerns are so crucial to the way we implement projects?

I think so. gamedev.se is a StackExchange site, and it has been built around a specific model. I understand that you're eager to help others, but we still have to follow that specific model. If you want to change that model, you may want to escalate and ask on meta.se.

\$\endgroup\$
  • 1
    \$\begingroup\$ Thanks. To me it feels very inorganic / discontinuous that we can ask & answer certain types of development questions here, but not others. And I'm aware of the background behind that, because there are a lot of genuinely-too-subjective questions out there (which one might argue is itself a subjective judgement). But it does feel as if the baby gets thrown out with the bathwater, so to speak. \$\endgroup\$ – Engineer Feb 27 at 18:58
  • 1
    \$\begingroup\$ With that in mind, I think there are some constructive steps we can take when a potentially useful question crosses the "too broad" threshold in the current community consensus: 1) Try editing the question down to ask a concrete question about one specific practice - eg. instead of "what are best practices in UI" maybe "how can I decide when to refresh or update UI components" - the user can always ask a second question about the next aspect they want more info about. 2) Create a developer blog post discussing your recommendations, and link it in the comments as a useful resource. \$\endgroup\$ – DMGregory Feb 27 at 19:02
  • 1
    \$\begingroup\$ @DMGregory I agree with option (1), and P.S. if that question needed to be narrowed down further, advice in that regard would have been helpful. I'm not as happy with option (2) as you are, as it scatters knowledge far and wide instead of keeping it centralised here. Link rot is real. \$\endgroup\$ – Engineer Feb 27 at 19:07
  • 1
    \$\begingroup\$ Hm. I've taken a look through the architecture tag and at this very moment I am seeing quite a few questions that should not be open, if this is the policy. That is why I feel there is something of a double-standard on this, as architecture questions I've addressed in the past have not been treated similarly. \$\endgroup\$ – Engineer Feb 27 at 19:20
  • 1
    \$\begingroup\$ @Alexandre Vaillancourt: So, in the absence of a response to my last comment: as I knew would be the case, those other questions are left open, yet this one remains closed. No-one will mention or lift a finger to look at those others and query why these double standards exist on SE. But when someone comes to challenge that status quo, woe betide them for going against general public opinion. \$\endgroup\$ – Engineer Mar 2 at 7:51
  • 3
    \$\begingroup\$ There is a significantly higher volume of questions than community members who can, and do, participate in close-voting. It's very common that questions "slip through the cracks," especially older questions (because when policy on what is/isn't on-topic changes, nobody usually bothers to go back through the thousands of older questions to compare them against the new rules). If there are questions that you feel should be closed, please vote to close them or bring them to a moderator's attention. \$\endgroup\$ – Josh Mar 2 at 15:50

You must log in to answer this question.

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