13
$\begingroup$

This thread is meant as a place to discuss the utility (or lack of it) for longer tags.

Currently, tags can have a maximum of 25 ASCII characters. That leads to interesting results such as differential-graded-algeb, synthetic-differential-ge, and computational-number-theo. While these are certainly marginal cases, they appear occasionally and are quite legitimate.

There is also an intrinsic use for longer tags, since they allow for more descriptive and specialized tags: for example, 'infinite-dimensional-manifolds' or even 'locally-convex-topological-vector-spaces' (these are probably not the best examples; perhaps you can suggest better ones).

In view of the above, I would like to ask people to give their input on the possibility of increasing the maximum length of tags:

  • How useful, or not, would this feature be?

  • What are the advantages and possible use cases?

  • What are the disadvantages?

  • Is it technically feasible?

  • Other opinions and comments...

In case this discussion leads anywhere and there is support for increasing the maximum length of tags (say, to 30 or even 40 characters), then this could be turned into a feature request.

$\endgroup$
8
  • 6
    $\begingroup$ Why do we need a length limit? $\endgroup$ Aug 23, 2013 at 0:21
  • 3
    $\begingroup$ @AndresCaicedo Well, part of me fears questions tagged approximate-ultra-character-amenability (this term has not yet entered the literature, but given the state of refereeing of my corner of the literature it can only be a matter of time) $\endgroup$
    – Yemon Choi
    Aug 23, 2013 at 4:11
  • 1
    $\begingroup$ Why not use high level tags with modifiers? For instance, the second two examples could be dg.synthetic and nt.computational $\endgroup$ Aug 23, 2013 at 15:29
  • 2
    $\begingroup$ @PhilipvanReeuwijk on the one hand I like the idea, on the other hand I am extremely hesitant to introduce anything in the tag-system that is not very self-explanatory; the prefixed top-level tags already create confusion. Also it could be bad for searching or at least for search engines (and then as a consequence for people searching). $\endgroup$
    – user9072
    Aug 23, 2013 at 19:35
  • $\begingroup$ @quid: I think a well-organized tree structure on tags would actually increase searchability (for humans; ample tag descriptions would hopefully take care of computer searches). Of course, I agree that enforcing consistent use of these multi-level tags could be problematic, let alone the feasibility of a complete tag overhaul... On the other hand, lots of very long, very specific tags without any organization like Yemon warns about aren't a great prospect either. $\endgroup$ Aug 23, 2013 at 20:34
  • $\begingroup$ @PhilipvanReeuwijk yes I think there is not much disagreement between you an me. As I do not want to populate this comment thread with my general remarks on tagging I write some things in a rough way in the chat romm for editing and tagging chat.stackexchange.com/rooms/10243/mo-editors-lounge $\endgroup$
    – user9072
    Aug 23, 2013 at 21:33
  • $\begingroup$ I am a little surprised by the first two tags listed by Ricardo, since I know that DGA and SDG are entirely standard and instantly recognized abbreviations. $\endgroup$
    – Todd Trimble Mod
    Sep 5, 2013 at 11:02
  • 5
    $\begingroup$ @ToddTrimble "instantly recognized abbreviations" I will admit my ignorance...I did not reread the post before reading your comment and then scrolled back to check what SDG means. (Now, this is far far away from the things I know about, and since we now have tag-wikis it could be alright to be a bit more generous with abreviations. Still, and in particular for tag-searching and suggestions, in general I think it is an advantage to have'full tag names'.) $\endgroup$
    – user9072
    Sep 5, 2013 at 21:09

2 Answers 2

2
$\begingroup$

On my opinion, we already have too many tags. This makes them less useful for all purposes. If we allow longer tags there will be many more tags... until the difference between the tag and the title will disapear, at which point tags will become completely useless. I think we should somehow encourage the use of fewer and more standartized tags. Like arxiv does.

$\endgroup$
2
  • 2
    $\begingroup$ I agree we have too many tags and basically with everything else in this answer. Yet, in my opinion this is however a bit orthogonal to the issue at hand. For example, I think it is reasonable to have a tag for computational number theory or transcendental number theory and would find it better could the be named, computational-number-theory and transcendental-number-theory instead of the current computational-number-theo and transcend.-number-theory An conversely, 25 characters still leaves huge room for any number of tags. $\endgroup$
    – user9072
    Sep 5, 2013 at 21:15
  • 2
    $\begingroup$ For more standardized: there used to be a prominent suggestion to use at least one of the top-level tags (those with two letter prefix, in fact correpsonding to arXiv categories). This was a bit lost in the software-upgrade. To encourage thos could be on the one hand achieved by adding it back into the documentation, but also via retagging questions. If you (not you personally but anyone) sees a question without such a tag high-up in the active list and there seems to be a fitting one (which is true most of the time), I'd say add one (in particular if you do not need remove one in process). $\endgroup$
    – user9072
    Sep 5, 2013 at 21:19
1
$\begingroup$

Let me give my answers on the specific questions:

  1. I think it would be somewhat useful to have slightly longer tags, there are some cases where natural tag-names are too long. For instance, computational-number-theory or transcendental-number-theory. Of course, one can then abbreviate these in some way but it seems odd to inflict this inconsistency on the tag-naming via an arbitrary limit.

  2. Immediated use-cases would be to 'correct' the artificially abbreviate tag-names we already have. These are not very many but there are some.

  3. Disadvantages could arise from longer tags taking up more space, this could be relevant in particular in the list of tags where (I think) the maximum possible length is taken into account for consistent layout (but this should not be a show-stopper). A non-disadvantage is the possibility of creation of too specific tags due to the increased limit; it seems a lot more likely people will 'force' the tag they want to create in the given length (resulting in a tag that is harder to parse) rather than to abandon the idea of creating the tag.

  4. I would assume there is no big technical problem here, except perhaps the layout-thing. But I cannot really know.

  5. Further comments:

    • It was asked in a comment why we need a limit at all, here my answer would be that some limit is definitely necessary to avoid a possible attack on the site, in theory somebody could copy-paste a large textstring in the tag-field and mess up the site even a couple hundred nonsense characters could be annoying (observe that most things on the site is somewhat limited). But to increase the existing one seems a good idea to me.

    • The idea of increasing this limit was brought up on occassion already (on meta.SO). It was not implemented but the case against it was not very strong either.

    • A thing to keep in mind is that tags are displayed in the browser-tab so having them too long is sort of inconvenient. However, the tag displayed is the most popular one (by number of prior usages) so that this should not be a big problem.

To sum up, I would consider it as a good thing if the tag-length was somewhat increased, maybe to 40; and if we can get strong support for this here, there might be a chance this gets done.

$\endgroup$
1
  • $\begingroup$ I answered this older question now as the question resurfaced 'to the top' and I thus was reminded of this (as opposed to explictly wanting to restart this discussion). $\endgroup$
    – user9072
    Oct 5, 2013 at 18:57

You must log in to answer this question.

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