What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 131 Stack Exchange communities.

It's an additional unnecessary burden to the honest stacker during suggested edit duty. A question presents some code then the minor editor adds a tag like 'for-loop' or 'forms'.

These tags are so general that they only have a handful of followers. They serve no useful purpose and are the cause of many too minor edits. Why not permit culling of generic tags?

share|improve this question
22  
The best thing to do is reject such edits, but there are too many who just approve them. –  Andrew Barber Jan 1 at 21:17
    
1  
Closely related to What is considered incorrect tagging? –  bluefeet Jan 1 at 21:25
10  
My proposal is to eliminate these tags. They do not serve any use as far as I can see and are detrimental to the site - as Andrew says many of these edits slip the net. –  suspectus Jan 1 at 21:31
    
@suspectus Part of the issue is that there are a lot of them, and it's not always easy to discern if a tag is actually useless or just being misused. If it's the former, then feel free to make a burninate request. If it's the latter, then feel free to edit them out as you see them. –  Dennis Meng Jan 2 at 2:01
6  
This seems to have become the preferred way to gain your first 2000 rep: apply BS tag edits on 5-year-old questions. And new users think this is acceptable (or even desired!) because others do it and get similar edits approved without problems. –  Pëkka Feb 2 at 4:41
2  
@Pëkka an artifact of robo reviewers more than the existence of dubious tags –  Kate Gregory Feb 2 at 14:06
    
@Kate but we need to set down rules for this either way. I've talked to a few of the users who suggested edits like these, and they always seem convinced what they're doing is somehow desired behaviour. Plus there are community rules that support that position. –  Pëkka Feb 2 at 15:40
    
@DennisMeng, if a tag is not a predictor of the set of people most lickly to answer a quesions, it may not be useful. Hopefull they are also stats on how often each tag is used as part of a search. –  Ian Ringrose Feb 3 at 22:26
    
@IanRingrose I agree, and that's part of why I almost always reject the tag-only edits I see; the tags that are being added aren't really good indicators. –  Dennis Meng Feb 3 at 23:28
    
Loosely related: meta.stackexchange.com/questions/124504/… –  ɥʇǝS Feb 4 at 0:32
    
Resisting the urge to retag this with [tag] [tag] [tag] –  Emrakul Feb 8 at 5:05

8 Answers 8

Question askers often add bad tags. I remember when we had questions with because of parsing at spaces. It sometimes makes an asker feel better to tag with the particular language keyword they're having trouble with.

  • removing a tag like that, because you think it's kind of meta, and making no other changes, would in my mind be "too minor" unless the post was perfect. (And statistically, it won't be.)
  • adding a tag like that is pointless. It's one thing to reassure the asker by letting them "add details" with the tagging system, but there's no need for anyone else to do it
  • changing a tag from something specific like to something more general like only makes the tag nondiscoverable for askers who probably don't yet realize that for, if, while and such are all part of a bigger category

I would leave the tags alone - they aren't hurting anyone - and reject edits that only add or remove them to posts that generally need help. I would probably even reject edits that only added or removed them to posts that were otherwise perfect.

share|improve this answer
9  
We still get someone creating visual and studio about once a month. There's an unfortunately good reason why I have a list of 179 tags I keep dead... –  Charles Feb 2 at 4:41
3  
But what about users who set out to earn their first 2000 rep specifically by adding for-loop et al. to any question they can get hold of that happens to be about a for loop? –  Pëkka Feb 3 at 22:06
1  
@Pekka Reject, reject, reject! If you spot them when they're just starting, @ them and tell them not to. If you spot them doing it again, flag and beg a mod to hand suspend them. No different than any case of reasonably systematic or mechanical minor suggested edits (eg typo fixes, removing Thanks etc) –  Kate Gregory Feb 3 at 22:13
2  
@Kate yeah, fair enough. But it's a lot of work and there are too many users to keep track of to make a difference. I think we'll end up needing to ban some of those tags simply because they get systematically misused by so many. –  Pëkka Feb 3 at 22:15
    
@Charles How about this: Here on meta, there are mod-specific tags, such as status-completed. How about you create one tag, such as forbidden-tag, and then make each of those 179 tags an alias of that one tag? It seems to me this would be a lot easier than having to monitor all those tags. The tag won't be used on any questions, it would only appear as the bottom suggestion for any tag (for having 0 tagged questions), so users won't see it pop up unless you're actually trying to add one of the forbidden tags. (also posted this idea here) –  joeytje50 Feb 4 at 0:40
3  
@joeytje50, you forget that many of these tags are low-traffic. You'd never get the required in-tag rep to synonymize it with something else. In the end, it's really not much better than just letting moderators blacklist tags directly instead of needing the dev team to do it. –  Charles Feb 4 at 2:07
    
@Charles I didn't know blacklisting was also possible. If those tags are already blacklisted, why do you still need to follow them? –  joeytje50 Feb 4 at 10:48
1  
@joeytje50, they aren't. We have had exactly one public blacklist addition -- [google] -- in about two years. –  Charles Feb 4 at 17:51
    
@Kate With my software engineering hat on it irritates me that this clutter accumulates. It's useless noise and misleads novice users. e.g. OP with C++ question parsing text for underscore chars labels post "underscore" - 106 followers. –  suspectus Feb 4 at 21:38

See also About unnecessary tags and the minor edits adding them for my list of low-quality tags.

I think the following changes might either: Help prevent tags like these from flooding the minor edit reviews, and wasting space in the 20 allowed reviews per day, or: prevent such minor edits from leaking through to be accepted. There might be some flaws in these suggestions, but perhaps it might help to just put this out there.

I am aware that the bounty on this question is looking for sources, and I only have one, but I think this might be useful to share anyway. I am not sure about if these are good enough suggestions to be feature-requests on their own though.

Restricting the usage of tags

Perhaps this is a slight bit radical, but I think this might just work. What if, before using a tag, it needs to have at least 2 followers? That would prevent the creator of a tag from using it before anyone else has shown any interest in it at all. It's not hard to get 2 followers if the tag is actually a useful tag, but pointless tags that are pointless won't get anybody else to use it, preventing the creator from using it too.

I can't think of any (significant) downsides to this*, since it would still allow anybody other than the creator of a tag to use a tag anywhere he wants - all he has to do is add it to his favourites if he is the second person to find the tag useful. After somebody else has shown interest in this tag too, the creator will then also be free to use the tag anywhere.

**See Troyen's comment for a downside to this suggestion for the smaller SE sites*

Ability to flag tag wikis for mod attention

As far as I could see, there's no option to flag tag wikis (I'd guess because they have to get approved by 5000+ rep users in the first place). I do think it would be useful to be able to flag these, so that we don't have to create list-your-bad-tags posts on meta.

Review audits for minor edits

Part of the problem is also that suggested-edit reviewers are letting these edits through. Currently, the only suggested-edit review audit I've seen is audits that were mainly insertion of gibberish. This seems to focus mainly on bot reviewers, since most humans that are paying attention would pick out most of those edits (not all, but most).

What I think is that perhaps if there are some audits that simulate edits adding very minor tags, such as the ones listed here, these extra audits would help the reviewers spot these minor edits more often, and therefore prevent these kinds of minor edits from being accepted so often.

More restrictions on tag creation

I have no idea why, but for some reason creating tags without needing them reviewed is allowed from 1500 rep, while editing tag wikis without needing it reviewed is restricted to 5000 rep. I think this allows for the creation of too many pointless tags way too easily. Raising the ability to create tags to 5000 rep too might help prevent pointless tags from being flooded.

Special case for approving/rejecting tag edits

Reviews for edits that only added or removed tags should be counted via the same system as foul balls in baseball: They do count towards the 20 edit reviews per day, but after those 20 edit reviews, you can still review them. Alternatively, they could just not count towards the 20 reviews per day, since they're so minor anyway.

More checks for commonly rejected/too minor-editing users

(not entirely sure about how this currently works, but I'm getting the impression things might be lacking)

I'm don't know how often this happens, but I think users whose edits get rejected should have their edit-suggestion rights temporarily revoked if they make too many of these minor edits. That would prevent the users that are mainly adding pointless tag wikis (like those who are trying to raise their reputation easily) from doing more harm than good with their edits.

Perhaps there could also be a list for moderators to see what edits were tag-only edits, to make it easier to find users who tend to make such minor edits a lot. Perhaps to discourage these massive tag-editing sprees, there could also be a reputation penalty that would revert the reputation gained on approved edit suggestions. This would be in response to Pëkka's concern:

Short on rep? Find 1000 questions, no matter how old they are, and enrich them with some bullshit tag like image or div. Get them rubber-stamped by three reviewers and enjoy your free rep! —Pëkka's bounty message on this question

Stop awarding +2 reputation for approved tag-edits

I think adding tags might just be a too easy way to get cheap reputation. Exactly like Pëkka said in the above quote, getting reputation is really easy if you just edit a bunch of tags. Basically every single tag-edit is too minor, except when the edit is something like removing completely irrelevant tags, or adding vital tags to a question.

The impression I'm getting though, is that most edits that only change the tags on a question are way too minor to warrant the +2 reputation. I do get the reputation for edits that actually improve something about the post, such as formatting or grammar mistakes, but in general, I think tag-only edits are just way too minor to be worth 2 reputation points (keep in mind, 2 points is the amount you lose with a downvote, to put it into perspective).

A list of forbidden tags

I'm not sure if this would be possible, but if it is, the following would probably be useful: Here on meta.SO, there are mod-specific tags, such as status-completed. How about we create one tag, for example "forbidden-tag", and then make each forbidden tag an alias of that one tag? It seems to me this would be a lot easier than having to monitor all the bad tags that should stay deleted. The tag won't be used on any questions, so it would only appear on the bottom of suggested tags to add to a question, and because of that users won't see forbidden-tag pop up unless you're specifically trying to add one of the forbidden tags.

**See Charles' reply to this for the better alternative.*

share|improve this answer
2  
The two-follower limit on your first suggestion would be a headache on other sites, especially newer sites or sites like Arqade and Board Games that require a tag with the game's name. –  Troyen Feb 4 at 0:48
    
I could see review audits with adding really obvious meta tags (at least hopefully obvious), but a meaningless tag on one site might be a useful tag on another site, so you'd have to find some way to seed the "bad tag list" efficiently and accurately across the network. –  Troyen Feb 4 at 0:49
    
@Troyen if there is a forbidden-tag like I mentioned in my bottom section, that would make it really easy to generate these kinds of bad minor edits. –  joeytje50 Feb 4 at 0:50

I don't think eliminating those is a good idea; some questions deal directly with the for loop, such as questions about expanding a for(a;b;c){d} into a while() loop.

The same thing applies with forms. There are questions that deal directly with forms, so the form tag is very applicable there. However, if you see questions with a for loop that have a edited in, reject the edit, roll it back, or edit the tags out

Here are some questions that I think are good usage of

  1. Loop through array in JavaScript
  2. Why is using "for...in" with array iteration such a bad idea?
share|improve this answer
9  
I've never seen an instance where a tag like for-loop or forms actually added value to the question though. –  Pëkka Jan 1 at 21:56
3  
These questions are still reachable through other tags, and would probably get observed through people following them –  Leeor Jan 1 at 21:58
6  
Yup, what @Pëkka said. I think the best way to look at tags is as things that potential people to answer will be following. Languages and major frameworks/technologies will be helpful there. Broad, non-specific terminology won't tend to. –  Andrew Barber Jan 2 at 13:46

I see people cram as many tags as they can into a question, stuff that are tangentially related whether or not they add value to the question. For example, beginners might conflate and when the question should only concern either of the two languages, not both. You'll see many "generic" tags like and just because the code has them, not because the question concerns them. The problem is exacerbated if the mistake is a typo, something obvious like "you need == instead of =" and so on. Replacing it with or or would not help. The unnecessary tagging occurs because the OP doesn't read the wiki or doesn't understand what the tag is for. And these more abstract tags are more confusing and less obvious than .

I don't think the solution is to get rid of the tags, but simply edit them out. Users with 2k rep (that's not a lot) can do it in less than 5 seconds. If the edit is in error, you can roll it back. Those without edit privileges should leave a comment instead ("you should use not , also remove and and "), as those edits would be considered "too minor."

share|improve this answer

Respected ALL, I think, you are trying to cure pain instead of the illness.

The problem is not why useless tags are added, but why there are useless tags.

For example, the is absolutely useless, because its meaning in different languages could be different. And some languages have such construction, but name it differently. There is the third reason, too, the most important one, but I'll say it at the end.

Not all rare tags are useless.

The rare tags belong to different categories.

Some of them are far-from-the-root branches on the tree of some technology. I would name them 'leaf tags'. As ->->. I think, such tags are simply excessive, independently on their frequency. If the question is about such linear layout taken from source and used in code, it immediately "needs" following tags:

, , , , ,

And we haven't yet even started to speak on the label we need on that layout!

What are these tags used for?

  • For defining "the circles" of the problem.
  • For searching for the questions I want to answer.
  • For searching for answers I want to know.
  • For praising of the member's knowledge on its profile.
  • For badges.

For praising and medals we need only root tags. Maybe some first-level branches.

For searching if somebody had the same problem that I had solved just now, on the contrary, the leaves of the technology tree are the most useful. But the search for a phrase is even more useful.

For searching for questions I could help with the middle level is more useful.

For searching for the answer on my problem, the SO search is too weak and people use google search instead. And it goes for headers, not for tags. And tags won't help, because when I have a problem I very often do not know yet (with exactness up to the leaf) where it belongs to.

And while we are defining the problem these leaf tags could be very useful, if only their use would automatically mean, without writing them down, the tags closer to the root, with the root itself, too. But we HAVE NOT any such system of hierarchical tags and it is not probable that such would appear.

So, at the recent state of things, the leaf tags are absolutely useless.

But the simple clearing the rare tags would be an error, because:

  • Some leaf tags are not so rare. has 2k following questions.
  • Some rare tags are not leafs. They could be the roots for some starting or not so wide used technology. They are very useful for some subcommunity, are searched for, etc. And large trees could grow of them yet.

So, the problem "what tags are bad" is not so easy. I propose the following:

The usefulness of tags should NOT be valued by frequency of use in questions, but by frequency of use in SO searches. If somebody looks for a tag, either by clicking on it or setting it in square brackets in the search field, at that moment and only at then one it is really used. And all unused should be cleaned off. Of course, new tags should have some time to get acknowledgement.

If there will be no useless tags, there will be no editions using them.

I am sure, that searches for are extremely rare. So, it will simply disappear, with the problem proposed in QA. And if not, if it IS really heavily used, then we are mistaking and it is a useful tag and it should be used. In edits, too.

share|improve this answer

I agree; I really don't see any value in the some of the following tags:





etc.

In most cases, users who tag their own questions with these tags are liking missing possible help/viewership from those following more applicable tags. In addition, they lend themselves to minor suggested edits that really do not add any value or clarity to the question.

My thought was these could potentially all be referenced from or , both of which seem to encapsulate these tags.

Is it unfair to assume that a sufficient knowledge of control structures implies a knowledge of them individually? If not, these tags seem excessive and unhelpful.

share|improve this answer
    
A person who doesn't know why their if or while is looping infinitely will never use a more generic tag like control-flow. Either it's ok to tag with the construct they're having trouble with or it's not. It's probably never right to add these tags if the OP didn't, but we can reject those edits when they're suggested. –  Kate Gregory Feb 1 at 16:20
    
Ah good point, new users wouldn't think of adding that tag. –  Joel Feb 1 at 16:22
1  
I would prefer that people focus on problem tags like [zoom], [pan] and the like, rather than tags like these that are clearly and specifically programming-related. –  Robert Harvey Feb 4 at 16:46
    
And in many languages the control structures simply don't use these words. Or use them a bit differently. –  Gangnus Feb 7 at 10:33

I don't know exactly if i am the right person to answer this question, but I truly believe out of my experience with Stack Exchange's group of sites is that they give, a freedom to novice or professional people to learn and above all get their queries resolved. It is a platform to ask genuine questions and get genuine answers by experts all across the globe.

Just recently for a couple of my questions I have been getting answers from experts in their fields, which helps me so much that I can't even express. So for people like me, the best understanding to the tags is something that helps people to identify the subject of the question and think if they have any idea about the subject, they can go ahead and increase their knowledge in the same subject.

Indeed a big Yes to avoid unnecessary stupid tags which do not create a difference to understand the label to the question like adding loop, for-loop, while-loop. I mean c'mon make it loops in Java or anything else that you understand make it precise, if you want the answer precise, don't confuse people with multiple tags which are not helping at all.

Adding such tags makes it pointless. Ya making the content of the question is more important and at the end for the people who put down questions, they should realize the fact that it is a whole big forum, so be specific if you need a right answer.

share|improve this answer

The OP identifies the problem with these tags as:

  • a burden for folks handling suggested edit reviews
  • very generally, that they are 'detrimental to the site'

In response to that -- burden or no, suggested edits aren't exactly piling up -- they seem to have a very short lifespan in the review queue.

As for being detrimental to the site -- that sounds ominous -- is it causing some serious issues there? (What are they?)

For my perspective -- I'd say at worst they're not all that useful; but on the positive site, they may be encouraging new users to get familiar and involved with the site, not unlike points and badges, winter hats, or what have you.

I'd say just leave them be.

share|improve this answer

You must log in to answer this question.

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