Mar 2, 2019 06:03
@MartinSleziak The tag-excerpt option only. Blacklisting is mostly useful for repeat offenders: deprecated tags that get created over and over again. That problem is very rare on MO. —
François G. Dorais ♦ Feb 25 at 6:04
deprecated-tags It seems the MO moderators dislike the idea of blacklisting deprecated tag while they still exist.
14
TL;DR: I suggest to blacklist the deprecated tags that still exist on some questions, in order to prevent them from being added to new questions.
There are a few tags which are deprecated and should not be used at all, but still exist on the site.1 Ideally such tags should be completely removed,...
I will try once again to summarize what I see as a possible advantage. (Maybe this does not add too much to what I already said in that post.)
TL;DR: Blacklisting the deprecated tags would be an easy way to prevent these tags from being used in new questions. Even if there remain older questions in need retagging, at least the number of problematic questions would stop growing.
My main point is: Why should people do manually something which can be automated. (And retagging new posts with deprecated tags is a thing which has to be done regularly and users have to pay attention to it and check whether the tags appeared in new questions.)
I remember that when similar issue was discussed on Mathematics Meta (
Can we please have a “deprecated tag” feature?), Asaf Karagila seemed relatively keen on something like that. However, he became moderator since then but I haven't seen this done with any tag on that site either.
(I do not know how often the moderator teams of the two sites discuss with each other, but you maybe talk about this with Asaf.)
Of course, all of this is probably irrelevant if the moderators think that the tag system on MO is already beyond saving and any efforts to do something with tags are just prolonging the agony.
The situation with the deprecated tags on MathOverflow - as I see it - is the following:
There are many questions with deprecated tags, the number of those questions is not decreasing too fast.
If you follow the links above you can find also how the number looked before. For example, there were 1000 questions with deprecated tags
in August 2017. However, since
August 2018 the change in the count of such questions was very small.
So if this trends remains, it might take rather long time until the deprecated tags get removed. In fact, it's possible that they will never be removed completely.
And I can easily imagine situation that the number of such question might be increasing if a few of the most active editors leave the site or become less active.
For example,
during May 2018 the number of questions with deprecated tags actually increased.
Since there are (at the moment) only very few users who edit questions and pay attention to tags, if they did not have to pay attention to new questions with deprecated tags, they might be able to find better use of their time.
And whoever follows the tag edits on MathOverflow can see that there are some users who make an effort to remove the deprecated tag from new (or bumped) posts - although sometimes a post with a deprecated tag goes unnoticed.
The situation is diffent on Mathematics - there are enough users on that site who are willing to devote their time to improve tagging. For example, after it was decided that
the (algebra) tag should be divided into (abstract-algebra) and (algebra-precalculus), enough users helped with this and the tag was removed relatively quickly.
As François G. Dorais pointed out
in an older discussion, MathOverflow is a smaller site, so older questions cannot be edited so fast as on Mathematics. (Not to mention that some of the users object to almost all edits of old questions, or even to almost all edits.)
Feb 1 '14 at 22:16, by
François G. Dorais I agree with @quid. The MO turnover rate is much slower than MSE. For 165 edits, it would take up to two months to do without causing ruckus.
Still, I think that the difference between number of users willing to help with editing tags on MathOverflow and Mathematics can be clearly seen. (Of course, maybe I am biased - since I am more active on Mathematics.
The only instances where I have seen some larger tag
being removed on MathOveflow were the tags merged by moderators into another tag and the tags removed manually by YCor. (I was involved in
the removal of (proofs), but that was a tag only with 9 questions.)
Although I suppose that some more examples can be found before my time on MO (when Ricardo Andrade and quid were active in editing and tag-related issues).
Again, I'll just reiterate what I mentioned above: I do not really see the point of removing the deprecated tags from new questions when it can be easily automated.
Basically the only upside I could see in doing this manually is that when such an edit is made, other users who see the edit might be reminded that there are deprecated tags. (Although
this post on meta shows that even if the edit summary explains why edit was made, users might still miss what the reason for the edit was.)
A very minor disadvantage is that somebody might think: "What? Questions about abstract algebra are not suitable here?" But this can be resolved by the text given in the warning. (And this problem is the same for all blacklisted tags - whether they are empty or they exist on some questions.)
@MartinSleziak The tag-excerpt option only. Blacklisting is mostly useful for repeat offenders: deprecated tags that get created over and over again. That problem is very rare on MO. —
François G. Dorais ♦ Feb 25 at 6:04
BTW depending on the interpretation of the phrase repeat offenders from your comment, deprecated tag might qualify. (Although I will admit that it is a bit of a stretch and you meant it a bit differently.)
Certainly they have been used many times
after they were deprecated -
June 2017 for abstract-algebra,
February 2016 for discrete-mathematics and
June 2013 for geometry.
I think it might make sense to make polytopes a synonym for convex-polytopes. Although non-convex polytopes are sometimes studied, convex ones are much more prevalent. Also there are only 36 questions tagged "polytopes" while there are hundreds tagged "convex polytopes."
MO editors' lounge
