12

Very often we get questions which reference TeX code without any code being present.
In many cases, the answer requires a MWE for the answerer to continue on, but it hasn't got a MWE.

However, if one could review questions into requiring a MWE for obviously trivial things, would that be a good idea?

Her are my thoughts about how it could be implemented:

  1. A possible answerer see a question which would benefit from a MWE. In that case he flags it as "MWE required" and it gets to the review process.
  2. If it succeeds in the review (i.e. flag accepted) the question will be closed for answers until an edit has been made. Also an automatic comment will be attached linking to "What is an MWE?", "How to generate a good MWE?", "Why is a MWE required?". Furthermore, if the questioneer, adds a comment below his question, the reviewers who made the decision will get notified so as to take action on any requests, etc.
    Now there are two possibilities after the edit:
    - It is re-opened for answers (I lend towards this, as it lessens the burden for the reviewers)
    - It gets back in the review pile (to ensure a suitable MWE).
  3. If it succeeds, there will be a bar under the questions which is there to inform other answeres about the request for a MWE. Then they can decide whether they wish to answer or not.

Drawbacks

This review process has a couple of drawbacks:

  • The burden of the community increases, as a lot more reviewing takes place.
  • It might frighten some people away from this Q&A, if they keep getting "Needs an MWE" when they dont know about to create one. (however, this will be leveraged through comments with the user)
  • This will only work if the review process goes faster than any answerers post an answer, but just as a close(x) flag is apparent on the question frame, so could MWE required(x) /MWE(x) be.
  • A flood of MWE required could pose a potential stall for any answers on the site.

Benefits

Some benefits:

  • The quality of the questions will be raised, as they will be more specific, with examples.
  • Newcomers will immediately be introduced about how to ask questions via SE standards (or at least TeX.SE).
  • Will increase quality of answers as it already introduces some code which should be explained (instead of just one-line answers with no or limited explanation) (also please elaborate on this opinion)
  • Very simple questions could receive extra attention as it leads to more detailed discussion about the inner-workings of TeX (often a MWE example shows what the person was thinking rather than what the person would want). This is a future benefit for other having the same problem.
  • Also in regards of TikZ (which is a heavily active question topic) it will be clearer for answerers to see what the questioneer actually is doing (TikZ is so complicated that any drawing has nearly unlimited number of possibilities).

Even if the review was wrongly justified, a simple edit would bring it to an active state again.

I am very much in doubt whether this is a good idea or not, but at least I would like to hear your opinion on this.
I also think that the community on this site would not stall due to the above method. People are nice about things here (and of course the way it is presented to the questioneer should not be "DO THIS YOU ...", but more like "Users xxx,xxx,xx,xx and xx suggests that you provide a MWE. This will show them exactly what your question is about as there could be ambiguities... etc. "

1 Answer 1

Reset to default
8

In my opinion, there are two parties that get hurt in the process of asking for MWEs.

One is the people who have commented for a MWE just above a 14 voted answer (which is kind of funny if you think about it, happened to me a lot so no complaints!).

The other one is the people who ask almost identical questions without MWE but the question gets closed instead of an answer.

I'm a huge fan of our community attitude however the unanswered list is increasing with simply unanswerable questions. When the load is a certain threshold Answer the Unanswered sessions are not enough to neutralize the inflation.

Hence, I am not in the favor of implementing a complicated policy that would cause user-conflicts on how a particular rule should be exercised. Instead we can warn the user that the question will very likely be closed after a few weeks unless someone with free-time or rep hunger steps in and do the hardwork for them. But s/he (arrgh) can make it attractive with just a few lines of code.

TLDR; my opinion is that, instead of rules and complications, psychological pressure is more pragmatic for us.

10
  • I agree on many points, however, my main motivation for answering is providing an as complete answer as possible. I would love that the same questions does not pop-up several times and that simple mistakes actually turns out to be too localized questions. I would just love to make every question as generalized as possible so that any inquires are more likely to show up that question. I just feel that many questions would benefit a larger audience by forcing the questioneer to work out a MWE. As stated, I also think this would increase the explanations in the answers?
    – nickpapior
    Commented Feb 12, 2013 at 19:22
  • For the first person (commentees on questions for MWE are typically frequent users, they will not be scared away by this, I think that the experts in fact are more encouraged to produce general, long-lasting answers). For the second person: How would you educate those to post questions in a stringent and appealing manner? By having an automatic way one does not need to keep a notepad open with references for MWE-links etc.
    – nickpapior
    Commented Feb 12, 2013 at 19:26
  • 1
    @zeroth I agree. But I think we are rationalizing the users a little too much. What I mean by that is, the manuals of the packages themselves already would answer 60-70% of the questions anyhow. So there is a strange fun for answering things on TeX.SE. You know that it's in the manual, probably they know too but they don't RTFM. So that already hints us that there is an inherent do-it-for-me nature to many, say, mostly TikZ, questions. Let's call these group 3. What is striking is that had they attempted to come up with MWE, they would most probably solve the issue themselves. (ctd.)
    – percusse
    Commented Feb 12, 2013 at 19:40
  • 1
    (ctd.) The group 4 is totally unaware of even the simplest workings of TeX, MP, TL,MikTeX and it's a complete mess. And they have all the rights to be confused since more or less everything is called ...-eX and half of them are drivers the others distros etc. So asking a MWE seems not the right way to push in my opinion for these. So as you say group 1 doesn't care agreed. group 2 has no other option but spending time on the site. My first question was also a mess, and I've posted stupid answers too. That was a credit offered to me and I got the point, a little too slow :) (ctd.)
    – percusse
    Commented Feb 12, 2013 at 19:44
  • 1
    So in summary, I think we should not forget that every individual has a different way of asking and learning. My conjecture is that the more people spend time here, the more they start immitating. Hence things go into the right place slowly. If we happen to implement such review mechanism, I'm pretty sure we will be facing more and more flame wars and SO like hostility. It's already present from our current review system that the questions are closed way way faster than pre-review system. I can't even catch any questions to review to be honest. We have to be as open as TeX.SE was open to us.
    – percusse
    Commented Feb 12, 2013 at 19:48
  • 1
    For the notepad part, I'm using this app meta.tex.stackexchange.com/a/2476/3235 and works like a charm. I don't even do anything else then this for comments.
    – percusse
    Commented Feb 12, 2013 at 19:53
  • Great response! First section, I agree. This happens too much to benefit anyone. Which is right here the MWE review would catch the persons who would have things handed on a plate. Maybe, I have not made myself clear here, but it seems to me that the community have a pretty clear mind on what the level of a questioneer is. In those cases a review vote would of course not be followed through as it is obvious from the general community. I highly encourage us to NOT lend towards the SO framework (much too hostile). But, I think that this review process, by presenting it in a correct manner
    – nickpapior
    Commented Feb 12, 2013 at 19:55
  • (ctd) would inhibit a natural and consistent way of addressing the issues of lacking information to supply an adequate answer. In regards of the imitating, I think that people which are addressed would could be improved to faster get an answer would benefit. Here, the concern is that if a very elegant answer is given the questioneer would probably remember the very elaborate answer. I think that many people would be encouraged to stay around and watch if answers provide more elaborations than anticipated, no?
    – nickpapior
    Commented Feb 12, 2013 at 19:56
  • I HAVE SO MUCH NEEDED THAT JAVASCRIPT!!! I haven't found that, only the ctan script! Thanks!
    – nickpapior
    Commented Feb 12, 2013 at 20:00
  • @zeroth That's perfectly appropriate. I would love to see that. But rational theories often don't go well with irrational behaving users :) As long as it's fair and human-like, I'm open to try it out. But your second drawback very likely would apply here regardless of what we do. I can handle a TikZ question one-on-one user but otherwise I have to wait a couple of users to vote. maybe I don't exactly understand your system sufficiently enough to rotate in my head.
    – percusse
    Commented Feb 12, 2013 at 22:37

You must log in to answer this question.

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