0
\$\begingroup\$

Food for thought.

There are times when questions come along which demonstrate a student's query, and with that comes the student's ill fated attempt to pose an "on topic" question without much experience. As such many such proactive, inquisitive, poignant and, in my view, very worthy questions end up [on hold] by the strict rules of Code Review.

Thus we end up with a constant stream of ignored content, not due to blatant disregard of site policy, but rather due to the inexperienced phrasing of a legitimate attempt for a code review. Legitimate attempt to learn.

Code Review is orders of magnitude smaller a place of learning than Stack Overflow. Where we could really be on par. Do we teach, want to teach, or are we here for exercise?

Who do we serve?

All here that have experience writing code know well that Code Review is not where a professional would seek review, nor would one expect an experienced professional coder to use Code Review for anything but a place to exercise. What is this place about?

I suggest

I suggest we be a lot more pliable in regard to questions. That review includes an assessment of how a code related question is asked, with particular attention to the student's conceptual difficulties.

I suggest that a bad question can be corrected by good answers, that reviews not only address the code, but the methods, misconceptions, inexperience involved in asking for review.

I dare propose that good review could serve more than the few that want to exercise their "in the know". That good review is serving the many, many we seldom see learn how to ask to for review.

After all (RULE ONE OF LEARNING) there is no such thing as a bad question?

\$\endgroup\$
9
  • \$\begingroup\$ A passing comment: I notice that you seem to conflate question quality with topicality. I think there is merit to closing questions that would be worthy of answers on different sites. I personally agree that we could be closing less questions, but I also want to see some questions closed more aggressively. Could you try elaborating your point a bit more, maybe add a few examples to help trace your line of thought? \$\endgroup\$
    – Vogel612
    Oct 7, 2018 at 20:37
  • \$\begingroup\$ @Vogel612 I posted this simply to start a discussion. Quality and topicality. Quality? is that for review. Topicality? via association (tagged). Off topic "How do I?" and "Where is the bug?" clear and easy to redirect (if you struggle to redirect?). Yet, "A best attempt, I get A & B, but C eludes" is not off topic.Answers should help many, not just posters. I struggle to see a benefit for anyone but those involved for reiterating "dot your i's and cross your t;s" which seams to me to be the majority of CR content. Hence I ask who do we serve? I ask because it is far from clear. \$\endgroup\$
    – Blindman67
    Oct 7, 2018 at 23:37
  • \$\begingroup\$ Hi Blindman67, it's good to see you bringing up this meta. I think there's some material for discussion here so I am not sure why you deleted it. Yesterday was Sunday and we normally have much lower activity on Sundays (I myself didn't have time to look at your question then). \$\endgroup\$
    – Simon Forsberg
    Oct 8, 2018 at 10:31
  • \$\begingroup\$ @SimonForsberg I deleted reluctantly as I became aware after answering Vogel's comment that my position on where the line between on and off topic is, is not clear. But then it is the discussion's aim is to help clarify, nor does my stance have to be clear, that I hope will result in the process. So thanks for the nudge. :) \$\endgroup\$
    – Blindman67
    Oct 8, 2018 at 15:46
  • \$\begingroup\$ "As such many such proactive, inquisitive, and in my view very worthy and poignant questions" but still of low-quality (no research effort) or simply off-topic (we can't move that line, it would allow a lot of crap to be posted). Not all bad questions can be countered with a good answer. Some, yes. But it's not a magic trick. \$\endgroup\$
    – Mast Mod
    Oct 9, 2018 at 4:51
  • 2
    \$\begingroup\$ "After all (RULE ONE OF LEARNING) there is no such thing as a bad question?" That's simply not true. \$\endgroup\$
    – Mast Mod
    Oct 9, 2018 at 5:14
  • \$\begingroup\$ @Mast the saying relates not to the content or quality of a question, but rather how it affects the willingness to ask. Also... real "crap" is easier found on an angels arse than on a Stack Exchange site. The line we "allow" is very clear, spam, abuse etc... gone before you can blink. We allow all other content, good or bad, closed or upvoted, it remains forever. You advocate with a worthy goal that reducing the willingness to ask improves content by reducing the amount of bad content. All that does is choke the tube that feeds us. \$\endgroup\$
    – Blindman67
    Oct 9, 2018 at 9:55
  • 1
    \$\begingroup\$ @Blindman67: Few links to questions that brought you here could really help in understanding your view. The main purpose of CR should be: "I have this code doing this and that, it is working, I have done my best. Can I still improve something?" But questions that should be closed usually are "I am trying to solve... this is my attempt... can you help me SOLVE it?" The line between the two is not so clear for me, I sometimes answer bad questions that show enough effort (and/or first comment makes theoretical code concrete). \$\endgroup\$
    – user52292
    Oct 9, 2018 at 13:06
  • \$\begingroup\$ @firda It is not the content of any question, the validity or justification for closing I address. Rather reexamination of what does forced prevention of answers serve. Its effect on posters needs, quality of their reviews, and who, how, and what is this meant to benefit. Between SO and this site I have voted to close, 18 times 2016 then 3 times last, and once this year, while proving 1300 answers. The sites needs are important to me. Only a few close votes i gave had justification for quality and betterment of the site, the rest served no one. \$\endgroup\$
    – Blindman67
    Oct 9, 2018 at 15:03

2 Answers 2

Reset to default
7
\$\begingroup\$

To reduce noise, I've posted what I think your main points in the question is below.

Food for thought.

  1. Our standard means some well meaning peoples questions are off-topic.
  2. We have a stream of ignored content from well meaning people.

Who do we serve?

  1. Code Review is a place to perform exercises, rather than get their professional code reviewed.

I suggest

  1. We expand what is on-topic scope to cater to students with problems.
  2. There is no such thing as a bad question, because it can have good answers.
  3. Changing to this will serve the many, not the few.

First I'd like you to know that your question is filled with good intentions, but much like this question intentions don't define what is good or bad.

  1. It does. But in life you always have to discriminate, otherwise you'll bite off more than you can chew. And choking on your food isn't good for anybody.

  2. Assuming we follow your proposal, and open the flood gates to these hard to answer questions. They will still be ignored. Take Python's unanswered stats, currently we have a 30% unanswered rate in the past 7 and 30 days, with 10% overall. How would following your proposal help this?

    It wouldn't. This is because our answerers have a limited amount of time, and so adding questions that require more time will reduce the amount that we answer, and will cause more questions that would have been on-topic anyway to also not be answered.

  3. This is just flat out wrong. Some of our users created, contribute and support rubberduck, with a GitHub repository stared by 669 people at the time of posting, and 168 forks. We also have a tag, , to review changes made to this project on Code Review.

  4. This as I hopefully have explained before would cause a steep reduction in site quality. And I think would cause Code Review to collapse due to the help-vampire effect.
  5. You can post a bad question whilst meeting our standards. Our standards don't define if a question is good or bad, just if it reaches a level we've found over time to be acceptable.
  6. I alone have 'reached ~722k people', with only 469 answers. This is on average 1539 people per answer. The idea of Stack Exchange is to have answers that help more than just the asker and the answer, but also on-lookers that don't ask or answer.

I don't think following your plan is a good idea, as I think it'd destroy the site as we know it, and put us in the same position as Stack Overflow. Which will just destroy us with the amount of answerers we have.

\$\endgroup\$
8
  • \$\begingroup\$ Thanks for the answer, I was not aware that so many view a relaxation of what should not be answerable questions (or what is poor content) will only result in more poor content. I am not one that avoids superlatives to make a point, but your answer lays it on with "...open the flood gate...", "...Code Review to collapse...", "...steep reduction in site quality.", and "...will just destroy us..." .I believe this view is incorrect but without evidence or example for or against I can only ponder. -2 for the paraphrased butchering, +3 for a sharp knife's edge. All other points good. +1 \$\endgroup\$
    – Blindman67
    Oct 9, 2018 at 13:15
  • 4
    \$\begingroup\$ @Blindman67 I'm sorry that you feel I've butchered your question. If you feel anything in your points are missing, I will more than happily add them and respond to them. I also feel Code Review is currently in a state of decay, answer rates have reduced and we don't really communicate between each other as much anymore. I think a lot of us are feeling the effects of help vampires, and so adding more will only exacerbate the decay and ultimately lead to a collapse. For you to convince me it is a safe and sustainable change we'd have to at least be in a strong state. \$\endgroup\$
    – Peilonrayz Mod
    Oct 9, 2018 at 13:46
  • 4
    \$\begingroup\$ @Blindman67 Honestly, I think Peilonrayz presents a perfectly good view. The site has been decaying already, partly due to the amount of low-quality and/or low-effort questions. Lowering the bar is counter-productive. Code Review is a site where quality is more important than quantity. We don't need 200 questions a day. 10 good questions is much more preferred. \$\endgroup\$
    – Mast Mod
    Oct 9, 2018 at 13:51
  • 1
    \$\begingroup\$ @Peilonrayz yes well "hacked to bloody death with a rusty axe" was the superlative i thought would better drive my point,, but that was over the character limit;.Personal I prefer my words, but not an issue. On "help vampires" is a contrivance and deliberately divisive. I have been helping people via forums since the days of bulletin boards. The solution to these "vampires" is never feel that giving an offer of help via anonymous public platform includes an obligation of help. We create the help we give, Vampires suck what we need to live... \$\endgroup\$
    – Blindman67
    Oct 10, 2018 at 22:42
  • \$\begingroup\$ @Mast: "The site has been decaying already, partly due to the amount of low-quality and/or low-effort questions." If these questions get closed so effortlessly, how do they cause the decay of the site? If "answer rates have reduced", it is because what is left after we close the off-topic questions are all more or less the same question. I follow the C++ tag. 50% of answers are "don't do using namespace std". It gets boring. \$\endgroup\$
    – Cris Luengo
    Oct 24, 2018 at 7:36
  • 1
    \$\begingroup\$ @CrisLuengo Good quality and a question being on-topic aren't the same. Yes we get low quality questions that contain the answer "don't use using namespace std", but if we lowered the bar around what is on-topic it would only add more things like that, and will bring questions where we can't understand what the OP is doing. \$\endgroup\$
    – Peilonrayz Mod
    Oct 24, 2018 at 8:14
  • \$\begingroup\$ I didn't interpret this question as lowering standards or accepting crud. It could be possible to widen what is considered on-topic to allow for discussions about what a good program design could be. That might be interesting. Currently you need to implement a poor design to get ideas from this community of how to design it better. And I feel that a program with a poor design makes for a poor question, since it's so easy to critique, there is not I learn from reading such answers. \$\endgroup\$
    – Cris Luengo
    Oct 24, 2018 at 23:21
  • \$\begingroup\$ Maybe my interpretation ^^ is more about my own desires, rather than what OP actually wrote. :) \$\endgroup\$
    – Cris Luengo
    Oct 24, 2018 at 23:23
3
\$\begingroup\$

Who do we serve?

My humble opinion is: people wanting to improve themselves and put enough effort in it.

What is enough effort?

The first reaction to bad question usually is a comment describing what is wrong (providing instructions, link to help and/or meta questions), maybe asking questions to clarify etc.: attempt to help OP improve the question. Edited (improved) questions can be reopened, theese left bad should not be here.

What do we need to serve well?

I think that this is the important question! My answers:

  • Keep good reviewers here! That requires a lot of work from moderators to keep good questions available for the reviewers and bad questions out of sight not to bother them. BTW, my favourite search pattern is [c++] or [c#] is:question score:1 closed:no, guess why.

  • Encourage good questions, discourage bad questions. Good question is the one that can help many people (not just OP). Bad question is usually serving only one person - OP. I myself have sometimes difficulties to see the line, but I understand, that raising the gate too much would flood this site with bad content that will expel many people - reviewers AND people wanting to see some good code, good soultions to problems and good reviews that can help them improve themselves without even posting any question.

My answers to your suggestions:

  1. Comment with link to meta question is usually enough to help improve the question. See e.g. Frequently Posted Comments. Way to better automate this for everybody, use chat, or allow (short-lived) non-answer posts could be helpful, allowing question-quality related answers is not.

    EDIT to clarify: There would be conflict with no answer-invalidating edits to questions rule. Non-answer posts would have to be clearly marked as such, not to violate the rule. Comment with proper link is usually enough (create meta question/answer if there is no proper content for the link).
  2. There can be good answer to bad question, but there should be no bad questions. I know that I am crossing this line myself sometimes, but most often after at least attempting to help improve the question. Good answer does not make the question good, but can still help to make it good, but that needs editing the question to make it good!
  3. How would you find good answer/review if the first thing you see is bad question?
\$\endgroup\$
6
  • \$\begingroup\$ Attract raher that "Keep" but good point on reviewers. The quality of the teachers dictates the quality of education.. However your first point is a big part of what i see as the problem here. "Who do we serve?" "... people wanting to improve themselves..." Yes.. but it comes with a condition "...and put enough effort in it" which would be speculative. And quantified with "...enough... " subjective and thus unworkable in a diverse community. \$\endgroup\$
    – Blindman67
    Oct 10, 2018 at 22:06
  • \$\begingroup\$ Enough effort = make the question meet the standard, which is decided by the community by voting (up/down, flag, vote to close/reopen). Community is judge with jury. I myself have sometimes problem to see the line, as I said many times, but I at least know the purpose, the goal. My main answer to your suggestions is: use comments ;) \$\endgroup\$
    – user52292
    Oct 11, 2018 at 5:30
  • 1
    \$\begingroup\$ "I myself have sometimes" You are not alone.Vote up or down is perfect but 5 close votes prevents the community of 1000;'s from answering. Comments don't work. .For me the community is global and we serve the most needy, the many poor where 30 min internet access costs a day's wage, a long trip into town Time to code is shared at the overcrowded school's only computer and the difficulty of language, That is effort and preventing an answer because a person did not read the rules sends a very anti community message to the ones who need it the most. \$\endgroup\$
    – Blindman67
    Oct 11, 2018 at 8:30
  • \$\begingroup\$ Comments don't work? Why? There is some room for improvement (as I stated in my answer), but comments are good, good enough - short sentence with proper link - most often to meta where more detail can be found - create a question/answer if you lack the content. Community is big and there has to be some democracy (which there is and we seem to agree that up/down voting is good), but too big for direct democracy (that worked in small city-states, not today). And even if you invoke those poor, they have more time to think it through and will benefit from good content (not seeing bad content). \$\endgroup\$
    – user52292
    Oct 15, 2018 at 12:44
  • \$\begingroup\$ Closing questions have very negative effect on new users. This year 323 members had questions closed (Jan to 13th Oct 2018) with a mean rep of 70, and 93% did not post again. "Comments don't work? Why?" as there is no room for a good answer and too late to assure new users "closed" does not mean "Go away". As for community, 334 closed (on hold) questions had 1536 votes to close, 1 member contributed 60% of those votes, 7 members 80%, and 18 members 94%. Of 4008 members posting questions this year. 18 does not represent the community view. \$\endgroup\$
    – Blindman67
    Oct 18, 2018 at 8:09
  • 3
    \$\begingroup\$ @Blindman67 Your question and comments seem more and more like a rant than a constructive question. I'm going to ask you to thread very carefully. We're losing enough highly-active community members as is. Don't put the blame on them. \$\endgroup\$
    – Mast Mod
    Oct 18, 2018 at 12:38

You must log in to answer this question.

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