Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
I didn't do these (the merge tool won't let you do a merge if the target tag doesn't already exist).
I just did those merges. The use of the merge tool is pretty weird when the target tag doesn't exist:
If the Remove Tag does not exist, the Master Tag will be renamed to the Remove Tag.
Merged [random-processes] into [stochastic-processes]. I'll leave the markov tags alone for now, but a third opinion will sway me.
Although it is not much used, I think that [character-sheaves] should stay as a tag; this is an important concept, it is possible to imagine additional questions about it, and it is a more refined concept than is conveyed by the blanket tag of perverse sheaves. (There was an agreed upon principle at some point that proliferation of tags is okay, as long as they are not redundant or causing confusion.)
I would think that [invertible-sheaves] is okay too; such questions might be tagged with [coherent-sheaves] as well, but I don't see why the first should be completely subsumed by the second.
It might still make sense to combine [sheaves], [sheaf] and [sheaf-theory]. I think the usual rule would be that all of these should be [sheaves], although in this case people seem to have already got used to [sheaf-theory].
I was just looking through recently created tags (viewable to 10k+ rep users) and doing some cleanup. I merged a bunch of no-brainers (eg [metric-geometry] into [mg.metric-geometry]). Here are some others that I intend to do, but I could imagine somebody having some objection, so I'm posting them here first.
To make this a bit easier to process, I've put a number to the left of each tag. This is the number of questions with the given tag.
1 [tvs] -> 0 [topological-vector-spaces] (name change)
4 [counterexamples] -> 177 [examples] (maybe debatable)
7 [lie] -> 182 [lie-groups]
12 [algebraic] -> [tag-removed] (probably mistakenly created; now available for misuse)
3 [theory] -> [tag-removed] (probably mistakenly created; now available for misuse)
2 [geometric] -> [tag-removed] (probably mistakenly created; now available for misuse)
2 [faq] -> [tag-removed] (only used before the existence of meta)
1 [transitive], 1 [isoperimetry], 1 [determinant], 1 [amenability], 1 [square-free]
1 [flops] and 1 [flips-flops] should be at most one tag. They only exist together on a single question.
Should 1 [etcs] have a better name?
The proposed merges/removals look good to me. "Isoperimetry" and "amenability" should maybe stay. "Transitive" means too many different things.
Dear Anton,
I think amenability should stay. There have been many questions on amenable groups, although it seems (based on your enumeration of 1) that most haven't used this tag.
Best wishes,
Matt
I also have my doubts about the value of the tag [abstract-nonsense] that Harry Gindi introduced with his last question.
I just did the merges/removals I proposed above. For some reason, I didn't quite feel like killing [square-free]. I merged [opinion] into [soft-question], [flops] into [flips-flops], and [determinant] into [determinants].
@pavlov: I think people are using [tensor] differently from [tensor-products]. Perhaps renaming it [tensors] would make this clearer.
Here are a few more proposals:
1 [sport] -> [tag-removed]
1 [community-wiki] -> [tag-removed]
1 [lo.models-of-pa] -> 0 [models-of-pa] or 524 [lo.logic] (shouldn't look like an arXiv tag since it's not)
2 [linear] -> 501 [linear-algebra]
(Edit: whoops, my links were broken. Fixed.)
@Andrey: [applications] is being used in a somewhat inconsistent way. Sometimes it has the same meaning as [applied] and sometimes it is being used to denote that a question is asking "what are the applications of X?" (not necessarily outside mathematics). Personally I think the second use is redundant since the word "application" is already in the question.
I suggest to remove [algebra]×1. It's too general to be meaningful.
I suggest to merge [calculus]×147 with [ca.classical-analysis]×246. Any meaningful usage of [calculus] seems to be covered by [ca.classical-analysis], and the former tag is systematically misused.
I also suggest to get rid of [analysis]×303. This tag is also systematically misused. Whoever wants to follow all questions on analysis can just follow [fa.functional-analysis]×456, [ca.classical-analysis]×246, and [ap.analysis-of-pdes]×124. I can see no other legitimate use for this very general tag.
(Edit by Anton: linkified tags)
The [algebra] tag seems to have just been created. Since it's only on one question, I'll just merge it into the most appropriate tag for that question, [ra.rings-and-algebras].
[calculus] and [analysis] are pretty frequent tags, so I'd like to get a lot of feedback about those suggestions before merging/removing them. It seems like [calculus] is often used on pedagogical questions, so it would be inappropriate to merge [calculus] with [ca.classical-analysis]. In non-pedagogical questions, I feel like [calculus] is meant to communicate a completely elementary point of view. I also have some reservations about removing the [analysis] tag, but I can't articulate them right now (perhaps because they're not any good).
I can see a use for [calculus] that is strictly a subset of [ca.classical-analysis]. While MO does not really encourage elementary calculus questions per se, on questions about pedagogy and history there is something nice about being about to refer to calculus instead all of classical analysis. [Edit: I see that Anton expressed pretty much the same sentiments on this right before I saved my post.]
I'm all for removing the [analysis] tag. But someone (pavlov perhaps?) has to volunteer to sort through all 303 questions and figure out where each of those questions really belong, and in the process also bump each individual one of them up by re-tagging. I think at this point it is somewhat of a lost cause.
@Andrey: under the text entry field, there are a couple of radio buttons labelled "Text" and "Markdown". You need to select "Markdown", after which you can linkify by typing [this is a link](http://this.com/is/the/url). For this specific thread, what I do is just write out what I want to say, then linkify the tags by piping the text through this sed command:
sed -e 's|\[\([a-z\.\+\#\-]*\)\]|[[\1]](http://mathoverflow.net/questions/tagged/\1)|g'
It looks like two of the instances of [literature] should go into [reference-request] and one into [writing].
I just noticed that the tag [presentation] is used in at least two completely different senses.
That's kinda funny. Any ideas on how to split them up (i.e. what names should we use for the new tags?)
[number] seems to come up periodically for some reason ... perhaps killing it should be on my list of monthly maintenance tasks.
It happens whenever someone tries to use a tag with the word "number" in it without using hyphens, e.g. number theory.
Speaking of which, why do we have a [number-fields] tag as distinct from [algebraic-number-theory]?
there are duplicate tags for reverse mathematics: [reverse-mathematics] and [reverse-math]
There is currently precisely one post tagged [operad]. On the other hand, there are 33 posts tagged [operads].
Should [matrices] be renamed to [matrix theory]? As of now, there is no tag [matrix theory], though there is a question named What is the difference between matrix theory and linear algebra, (tagged of course [linear-algebra]).
[tag-removed] is a tag used to remove tags from other questions without bumping them (by merging). It cannot itself be removed in this way, so removing [tag-removed] would require bumping a lot of questions.
@gilkalai,
the problem is that each of these 100 questions will rise to the top of the front page as the [tag-removed] tag is removed. It's possible that we could ask the Stack Exchange people to run a special script against the database for us, removing [tag-removed] from any questions which also has another tag, but I'm not certain they'd be eager to do this for us.
I've found an old unregistered account that should be merged with VA.
@Zev: Done. (This really belongs in the "merge two user IDs" thread.)
I'm thinking of changing [ca.classical-analysis] to [ca.analysis-and-odes]. The real name should be [ca.classical-analysis-and-odes] but that's too long for the software. The main reason is that odes currently have no place to go except [differential-equations], which is a mess of odes and pdes. Ideally, [differential-equations] would disappear, but that doesn't seem to be feasible right now. If anyone objects to this name change, they should do so before Monday.
The change is done.
Gjergji, it takes some time for the system to update, and probably more with big edits like this one. The page http://mathoverflow.net/questions/tagged/ca.analysis-and-odes currently shows nearly 800 posts. This count will eventually appear on the tag board. Similarly, the old tag name will eventually disappear completely (provided it is unused).
@Francois: Thanks for the rename. It'd be better if it were more loudly announced (since it being a "major tag" that users use for feeds and filters). I only found out that it got renamed when I noticed that the ca.classical-analysis RSS feed all of a sudden became empty.
Also, I now challenge you to also fit "harmonic analysis" in that tag name (it technically also belongs, you know). :-p
The following (incomplete list of) tags have incomplete names due to length. Any suggestions for shortening them so that they fit?
I'm not sure if it appears here before, but it is worth re-requesting if it did appear:
[gn.general-topology] and [topology] and [general-topology] should all be merged into one tag, the rest should be made synonyms.
Is it clear that [topology] is always used in the sense of [gn.general-topology]?
I think plain 'topology' is more general than 'general-topology'.
@Dmitri; also as an argument for switching to 2.0, and joining the Stack Exchange network. They handle tag creation rather nicely.