9

Suppose that I want to tag a post with , , and , and is not a tag that presently exists.

There are some cues that a tag may be new in terms of autocompletion, but it would help to know if a tag is new by seeing its background in a different (background) color, especially if autocomplete is slow and I type quickly.

I know there is some information conveyed by color already and possible collisions, but it would be nice to have a reddish color for a tag, or yellow, if the tag is new (and, in most cases, something I don't want to carry forward, even if I have the rep for it).

My main point, which I have tried to explain and have not had anyone else either acknowledge or contest, is typing at speed. With the feature I propose, network latency is less significant and you can type tags at speed.

With existing side-effect-of-autocomplete functionality, you need to stop one time per tag, and if you are entering a new tag, you wait until you call a timeout to decide autocomplete is not loading. With the feature I propose, possibly with a separate greyish color for a tag waiting on latency to mark as existing or new, you can type tags at speed, vi style, and you need stop at the end, and even that one latency is less than the latency of time-out decision on your end that autocomplete is not going to load.

There is a substantial difference for typing at speed.

  • Not that there might not be utility in this anyway, but wouldn't checking for the non-existence of the tag you're entering suffer from the same latency issues as the autocomplete? For maximum effect it'd probably be necessary to highlight the tag post-edit as well so you could go back and correct the tags if that wasn't what you wanted. – Tim Stone Jul 26 '14 at 14:29
  • It would not dodge latency but latency would be less significant. If you're entering five tags, you have to stop typing at speed five times to see if it comes up on autocomplete. If the feature I suggest is added, you do not need to stop; you need stop at most once, at the end, for the last tag. It makes a difference for touch typing / typing at speed. – Christos Hayward Jul 26 '14 at 22:10
  • Ah, right, that's a fair point. – Tim Stone Jul 26 '14 at 22:11
5

While I agree with Chris's answer that there might not be a use for coloring new tags on the question-asker's end, there might actually be use for it on the viewer's end; specifically, in the case where a new tag was created that looks like it makes sense, but is actually covered by a pre-existing tag.

Example: This would be useful for when someone with 300 rep (150 on a beta site*) creates a new tag for a topic that is already covered by another tag, simply because he doesn't yet understand how all the tags on that site works -- this would highlight the tag to those that know this tag's topic has been covered already, and replace it with the older tag.


*While I do understand and agree that beta sites' privileges need to be given out at lower rep to encourage involvement and allow for moderation earlier on, one answer with a score of five is not necessarily enough to understand how tagging works on a given beta site.

  • I'm new to Meta.SE ....should this suggestion have been a new question? – Shokhet Jul 27 '14 at 12:51
1

I don't think there's a need for this.

If the tag doesn't exist then it doesn't autocomplete at all so you can see quite easily that it's new.

Here I'm entering "new-tag" into the box and it doesn't appear in the list:

enter image description here

  • 2
    Yes, I was aware of that, but sometimes I type without waiting for autocomplete. I do not see how to distinguish between autocomplete not loading yet because with network conditions it's slower than my typing, and autocomplete not loading yet because it will never load. What you mention works if you slow down and wait for the autocomplete to load (or wait long enough that it probably will not ever load). What I propose works for typing at speed which does not wait for autocomplete. – Christos Hayward Jul 26 '14 at 13:53
  • One side note about your example, which I find is significant because it is your own, cherry-picked example for why my proposed feature is not needed: what you gave doesn't demonstrate that new-tag doesn't exist yet; I cannot tell from the screenshot whether new-tag doesn't match anything (though the new-tags-page result has information scent), or whether new-tag is below the fold on matches for new-* tags. All your example shows is that new-tag is below the fold for new-* matches, not that it is not anywhere. Now you can get a more definite "No" if the "above the fold" must be all there is... – Christos Hayward Jul 26 '14 at 14:02
  • ...or you wait a while and nothing pops up, but your cherry-picked example in fact does not show how one can tell there is no new-tag option. What I'm proposing works for typing tags at speed and not waiting on autocomplete. What you're pointing out in existing features does not. – Christos Hayward Jul 26 '14 at 14:04
  • @Jonathan We also (on some sites, including SO) show a warning when you're about to create a new tag. I unfortunately don't have a screenshot link handy cause I'm on mobile. But, we have the ability to check with the user whether they want to make a new tag or if they made a typo or if they want to search for an existing one instead. – Adam Lear Jul 26 '14 at 23:15
  • @Jonathan Found it: meta.stackexchange.com/a/233897/155160 – Adam Lear Jul 26 '14 at 23:19

You must log in to answer this question.

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