More and more I keep seeing folk post what I would think of as 'answers' as comments on the original question.
Copying attributes from one polygon layer to another? is one example - but I have seen many more.
I wonder why this might be. Is there a particular Stack Exchange reason or do folk think it isn't really a proper answer?
I find that if it's not posted as an answer then you can't have it up-voted in the same way, which makes it less useful.
No, it is not unusual, but it is a problem.
Some users don't understand the difference between an answer and a comment. The comment text box is the first thing you see after the question, so it's a nice convenient place to simply start typing. Comments even look somewhat like a traditional forum, so it's easy to fall back into those bad habits of interjecting whatever random thoughts come to mind about the topic at hand.
But more often than not, those users simply don't have the time, the complete story, or simply do not have the inclination to post a full answer — so comments offer the path of least resistance. I can understand their desire to offer what little help they can, but the problem is that they're not helping the system by spreading little bits and pieces of the solution wherever they please.
That's a forum problem. We don't do that here.
Comments were designed to simply ask for clarification about the original question. Comments are such a light-weight mechanism, that putting useful information in there can actually be somewhat harmful:
When someone asks a question, members of the community are supposed to propose answers. Others vote on those answers so, very quickly, the answers with the most votes rise to the top. You don’t have to read through a lot of discussion to find the best answer. Answers-as-comments breaks that entire structure.
Discussions in comments are no better than any bulletin board system on the Internet. Every time a new user comes along, they have to read through the initial post, then all the noise in the comments to get the whole story. You end up with valuable bits of the solution scattered among a bunch of noise and conversation. That's a condition we work very hard to prevent on these sites: The Chat Room/Forum Problem.
I understand the urge to interject "just one more interesting piece of advice" when you have something terribly interesting to say. The problem is that other users will emulate what they see on the site and soon a lot of problem-solving goes into the comments.
Every comment that does not belong adds one more broken piece to the site.
When someone asks a question trying to troubleshoot a problem, I find it better to ask them a question like "Have you tried X?" in a comment. If they haven't tried X, but then do so and reply that it solves their problem I will then post an answer ("Try X.") - unless I forget. Otherwise I would end up with multiple answers ... "Try X", "Try Y" and "Try Z".
Perhaps there can be a way to flag comments as an answer (instead of offensive, spam or unconstructive). Then, if a comment is flagged enough times as an answer, it is converted into one. Also, if a certain user has enough of their comments flagged as an answer, they get sent a message laying out the purpose of the comment system, or lose the ability to comment for some period.
Here are two more examples:
Dissolving two different buffers together so not overlapping using ArcGIS Desktop?
SE policy is that answers should be posted in replies. But IMHO there are mitigating factors that deserve sympathy.
Sometimes a respondent knows an answer but simply hasn't the time for a full reply; sometimes they think they might have a lead on an answer but aren't sure; and sometimes they think the question is so straightforward its answer can easily be found on the Web, so they provide an appropriate pointer. In all cases it's fair to assume the motivation to post a comment is due to a desire to help the questioner.
In the particular case you point out, the answer given as a comment is rather tersely described and it will not always produce a correct solution. It would be subject to downvoting if posted as a reply. Although you cannot downvote comments, you can always respond to them to point out where their deficiencies may lie.
You are correct that answers posted as comments are not as useful as full-fledged answers. Therefore, in situations where a comment is, or can easily be converted to, a full and accurate reply, often the questioner or other community members urge the commenter to do so. There's nothing the matter with someone else jumping in and offering the comment as a reply, either. For instance, if you're convinced the comment you reference really is a good answer, go ahead and post it as a reply--but be prepared for the consequences if you don't fully understand its implications!
Sometimes the question requires more explanation/information - most of the comments on the question are pointers not necessary answers.
Like this Question Migrating ArcGIS Server 10.02 to a new server requires a system requirement and future requirement to be successful answer. Time constraints also do play apart.
Well, if you need any clarification, you sometimes post a provisional answer as a comment ("but I need more specifics to be sure"). Sometimes that's good enough for an answer, but often it's not. And if you post a non-answer as an answer, you can get downvoted a lot. So personally, I just err on the side of safety and make it a comment if I'm not 100% sure that I understand what the OP's problem is yet.
