See the first revision of this answer, quoting completely:

I'm not really sure if you can do it in a terminal, but with my iMac I used tabbing instead.

(please wait, I'm writing the post...)

The next revision posted 9 minutes later (non-ninja-editing) of the answer is fine and of good quality in my opinion.

How ludicrous is this? Is it okay to do this or a really bad practice? Are your objections bound to a time frame? I mean, obviously, an answer in this state for more than an hour is unacceptable, right?

Obvious advantages of doing it; badges, an extra upvote you might catch, etc.

Running the search on this to see how often this is done is really hard - (valid) comments are showing up and loads of noise. I did see it before, though. For the same reason I couldn't find any metapost on it here or SO.

I have mixed feelings about this, so I'm asking your opinion.

Oh, the MSO crowd loves quick and dirty answers. For the specific issue of immediate non-answers that are edited into something halfway decent, try searching for “fastest gun in the west” (FGITW) and “edit window”. Thankfully smaller Stack Exchange sites tend to favor quality more. I hope Ask Ubuntu falls on the side of quality. –  Gilles Jan 9 '13 at 21:12
@Gilles Ahhhh of course!! How could I miss to use the very obvious "Fastest Gun in the West" as my search query to find opinions about it? holding sarcasm sign and smiles gratefully for the should-be-an-answer comment. –  gertvdijk Jan 9 '13 at 21:31
I think it is good when question have two parts and second part need more research –  Tachyons Jan 10 '13 at 15:38
I guess I do not agree with majority here, but as long as the person really is typing it up I would appreciate it. It has happened to me a few times that I was creating an answer for 10 minutes just to see that someone finished nearly the same answer in parallel. Basically meaning that one of us could have spent the time on something else. –  Dennis Jaheruddin Jan 14 '13 at 13:25
@DennisJaheruddin you'd probably be better off commenting on the question with something such as: "I'm typing an answer using FOO method " so people know that you're working on an answer using a certain method. –  Oyibo Feb 7 '13 at 22:14
add comment

4 Answers

up vote 24 down vote accepted

We can make this really simple and boil it down to this:

If it doesn't answer the question, it doesn't belong in the answer section. Don't post it there.

However there are some approaches that are legitimate that you can't really stop:

  • Iterative improvement is something we should encourage but only because it makes the answer better. But deliberately leaving detail out for the sake of speed is idiotic.

  • In the same vein, expanding stubs (eg originally asking the user to restart apache2 and then replacing that with instructions on how to restart apache2) improves the quality of the answer and should be encouraged. But similarly, leaving things out deliberately is silly.

  • Throwing something out when time is urgent. I'll be honest, I've never seen a question that really warrants somebody posting an answer to make sure they will or won't do something in time. A real-world analogue of where this might be useful is talking somebody off a bridge. It's critical to shout "don't jump" immediately and you can work on the reasoning later on.

    ... there is an argument this can all be done through comments.

So please don't do this here. MSO is not a template for other SE sites. If you catch somebody doing this please flag for moderator attention! Don't bother leaving messages, we'll deal with them directly. Once we've pulled out enough teeth that we're happy they won't do it again, we'll release them back into the general population.

I have a strong feeling that this is the answer. Just needs some upvotes. –  gertvdijk Jan 10 '13 at 13:48
It seems such placeholder answers may not be so approved of on MSO either. –  Eliah Kagan Feb 5 '13 at 21:22
add comment

The way I see it, AskUbuntu despite being good for promoting friendly competition to increase answer range and quality through vanity rewards such as badges, shouldn't be tolerating this sort of methodology as it promotes speed answers rather than quality answers.

Either the user:

  • Wants to inform the user that they'll have an answer soon in the case of hypothetical questions which need solving within a certain time-frame which is very rare.

In which case it would be better to post "I'm typing up an answer it will be completed soon." in the comment section below the question rather than a "please wait I'm writing the post..." in the answer section.

Or the user is doing it for reasons related to themselves:

1. Doing it to obtain various badges such as: Enlightened or Revival

Although in the case of the latter there wouldn't be much chance of other users posting an answer right before them if noone's answered in 30 days.

2. Making sure the answer they have worked towards gets the credit it deserves, in the case for example that another answer is written in the mean-time and the OP accepts without looking at theirs which could be of equal or higher standard.

3. They want to inform other people capable of answering the question that the question is already answered.

In case 1. Wanting to obtain badges:

The badges are a minor incentive and should be used as motivation to answer more questions rather than asking one question quickly. Quick answers will most of the time not lead to higher quality responses, in the case you've pointed out the answer was substantially improved, however this could not always be the case.

In case 2. They want credit for an answer they are researching/typing:

It is worth simply commenting and notifying the OP of the question afterwards regardless if someone else has already answered:

  • Either the answers will be similar in which case they can add theirs as an edit and they will have contributed to making an even better answers together.
  • Or they will be different solutions which will both benefit the community and the OP will be free to either accept their answer and/or upvote for effort.

In case 3. Wanting to inform others that the answer is already answered.

This is the worst case scenario in my opinion even though it could have stemmed from a good intention, doing this has several issues:

  • The question comes up as already having an answer in the questions list reducing the chance people will give it a look.
  • If the person writing the answer doesn't give a correct answer or one which doesn't solve the OP's problem, people will have overlooked it in the meantime despite potentially having more helpful ideas.
  • The question may not get as many answers, taking an example of an unanswered question where 3 people work on finding a solution at the same time, they will post their solutions not knowing others are working on it and so provide 3 answers for the same question. If someone answers "Typing an answer now" 2 of those people could not bother and it would reduce the variety of answers obtained by the OP and maybe the quality too if the person posted an answer in a rush.

The counter-argument to the 3rd point of part 3. just above would be that:

"You could have 3 people typing exactly the same thing and it could waste time better spent answering other questions".

That could happen in which case you'd have 3 answers which confirm each other. However if you only had one person working on the question and they misunderstood the question or posted a solution which didn't work, then the OP wouldn't have any alternative answers to turn to, I view the case of sometimes having 3 answers confirming the same reasoning being a better option.

Disclaimer: This is but an opinion on the issue feel free to comment/counter-argument any points made, I could easily have missed something out/been completely wrong about something.

add comment

I mostly agree with Oyibo, but I make a different distinction: This is less about fast answers vs. quality answers, and more about real answers vs. fake answers.

An answer should answer the question. At least partially. If it doesn't, it shouldn't exist. If it answers the question starting at some point in the future, that's when it should start existing, too.

If an answer is quick and short but it's a real answer that might help, it's OK to post it and then subsequently improve it. This is OK because it would be OK to post it even without the later improvement.

However, if an answer is just a placeholder for a future version of itself, that's not a correct use of the site.

  • If someone must announce the intention to answer a question, a comment can be used for this purpose. This occasionally makes sense when people are discussing closing the question.
  • It's often quite appropriate for a question to have multiple answers, even if they illustrate the same technique (but present it in different ways). A question does not belong to whoever manages to post first.
  • If someone accidentally posts an answer before it's developed enough to answer the question, they can delete it, finish working on it, then undelete it.
  • An OP can change which answer is marked as accepted. This happens with some frequency. Getting one's answer accepted isn't very important, but it also doesn't depend on who posted first nearly as much as people tend to think. Or, to put it another way: if it depends on who posted first, then no answer was really clearly better than the others.
Far more succinct than my answer, the real vs. fake criteria clarifies better why it should be flagged automatically. +1 –  Oyibo Jan 10 '13 at 13:56
add comment

It is very ludicrous. I'd downvote it and I expect others would also.

add comment

You must log in to answer this question.

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