How to ask a specific question without in the risk of "too localized"?

Question

One may met problems in document writing, some of them are specific that rarely to met by others. So questions asked for seeking resolutions for these problems are often marked "too localized". But if one actually want a resolution, how does he ask to avoid the risk of "too localized"?

Answer 1

Questions that work well on the StackExchange sites are ones for which a resolution is well-defined. If this is not the case - when the question is too broad (leading to no specific, concrete answer) or when a question solicits subjective commentary (quite often although not exclusive to questions tagged best-practices) - the risk of being closed as "too localized" increases.

That being said, I often find that questions closed as "too localized" are from users who do not engage the community during the process of the Q&A. They post a question, mostly without MWE, making it difficult to reproduce. After the question is posted with some conversation through comment from community members, the OP is no-where to be found.

One should always (no exclusion, in my opinion) put yourself in the shoes of the community, when asking a question. Also - and this is important - be ready to clarify some things that are asked in comments. If clarification or feedback seems lacking, a question quickly goes stale and are likely to be closed as "too localized."

It's been mentioned over-and-over that posting a MWE^* helps, and it does. To some extent, it shows that the OP has put in some effort to isolate the problem so that discussions around a solution is much more focused and doesn't waste time of the contributors. Include pictures as well (even if you have low reputation; our "janitors" will embed them for you tout de suite), since they relay information very well. Yes, while the SE network provides valuable input from experienced community members, they most likely have jobs keeping them busy outside of this realm, making their time spent here valuable.

If all else fails, and you're still unsure, there's always the chat room where active members discuss everything from document classes to ducks, packages to programming and configurations to cricket. Visit it, start a discussion and see if your problem can be solved there. You might soon be prompted to ask the question formally on the main site, and can even provide a link to the chat as a frame of reference.

Most of the above, sadly, excludes the few that have less than 5 reputation on this site. But the reality is that not everyone can be accommodated equally...

_{^* By MWE here I mean that you should show what you have tried in order to replicate your work/problem. Sometimes the MWE is not working, but that better than nothing.}

Answer 2

If you can manage to demonstrate the problem you’re struggling with within 20 or 30 lines of a fully compilable Minimal Working Example, your question most likely won’t be closed as too localized.

What makes questions too localized is not including a full document or including several problems in one question (this might trigger other closing reasons as well). I don’t think we closed questions very often just because the asker is attempting something really obscure – if only they stepped away from the problem enough and tried to ask the question in an abstract way.

Imagine I was trying to typeset my latest musical composition: a song whose lyrics consist of chemical compounds and syntactic structure trees. For some reason, however, I can’t get the packages abc (for music), chemfig (for the chemistry magic) and qtree (for the trees) to work together. If I create a nice MWE clearly and reproducibly showing the problem occurring, this question would not be closed, even though I’d probably be the only dunce on this planet who ever needs this weird combination of packages.

For some hints on how to arrive at such concise code, check out I've just been asked to write a minimal example, what is that?.

Answer 3

Already you had good answers here. I don't know if this material is already covered. Just for record and helpful to other newbies.

My POV is based on: How not to ask question or atleast first solve by ourself approach. Assuming we are NOT Working on a TeX related project at the last minute.

Nevertheless don't stop posting any wierd TeX/LaTeX question with MWE effort because stupid ones become very popular and might be turn clever over time

1. Learning/Adapting using Advanced Search, somewhat helps not to ask a too localised Q or duplicate. On bugs related to TeX/LaTeX editors verify bug tracker and Q on bug reporting for . eg: Tag based search like : Best-practices tags and big-list tags are worth reading

2. Read the documentation and TeX/LaTeX books at http://ctan.org/search. It may seem stupid, but worth the effort because everyone becomes expert by Practice(more-filenames.tex),Reading(more-reading.pdf) and Role Reversal (Try answering TeX.SX Q's or post comments).

3. Working on MWE using Buildingup and Hacking Down and Remember .log file contents matter the most Learning debugging is very important to be self reliant. See Common errors and The joy of TeX errors.

4. Good(Sharp/Specific) Title, Perfect Tagging,Thorough Background Search/effort with minimal working example (MWE) along with instant OP comment support increases chances for a quick/accurate answer with heavy upvoting to Q and A.

5. Short Answer to Q: If possible try broadening the scope of the Q to benefit the larger audience. Everyone wants to learn from Q or from A including yourself, so you need to feed the fodder. BTW you can always show discontent and open the closed question with modifications. feel free to chat to resolve them, there are good TeX and LaTeX friends availabe.

6. Don't forget the Mantra Upvote Q immediately and often, Downvote with comments and Accept answer lately(24hrs).

Finally don't despair everyone makes mistakes and learn from them during the starting days. One day everyone will have a chance,but never ever give up trying.