In a recent comment, one occurrence of stars-to-italics ('particular') doesn't render, but one does ('abstract').

Screenshot of Markdown rendering issue

This seems to be a bug. Is this the right place to report it?

EDIT: Now also posted on Meta.SE.

EDIT: Per @MartinSleziak's investigations below, it seems to be something about a MathJax + Markdown combination like "$\cong$—spacer italics spacer—$\mapsto$" ($\cong$—spacer *italics* spacer—$\mapsto$) (as comment). It renders properly in the body of the post, so perhaps it's just a comment thing (as MartinSleziak helpfully indicated in the tags!).

  • 2
    I don't think that 'centreless' is a consequence of the abstract isomorphism $G \cong \mathrm{Aut}(G)$—at least, not obviously so—but rather of the *particular* isomorphism $g \mapsto \mathrm{Inn}(g)$. (Of course your statement about complete groups remains true, at least according to the Wikipedia definition, which currently builds 'centreless' into the definition ….) I bring this up because it seemed that @JustinBenfield was asking about abstract isomorphism. – LSpice Jul 17 at 21:29
  • (@JustinBenfield, I don't know if non-participants receive at-notifications, but sorry about the spurious one you'll have got from that.) – LSpice Jul 17 at 21:30

This answer if for copy-and-paste in case somebody wants to try whether the comment behaves differently if they change something:

I don't think that 'centreless' is a consequence of the abstract isomorphism $G \cong \mathrm{Aut}(G)$—at least, not obviously so—but rather of the particular isomorphism $g \mapsto \mathrm{Inn}(g)$. (Of course your statement about complete groups remains true, at least according to the Wikipedia definition, which currently builds 'centreless' into the definition ….) I bring this up because it seemed that @JustinBenfield was asking about abstract isomorphism.

I don't think that 'centreless' is a consequence of the abstract isomorphism $G \cong \mathrm{Aut}(G)$—at least, not obviously so—but rather of the *particular* isomorphism $g \mapsto \mathrm{Inn}(g)$. (Of course your statement about complete groups remains true, at least according to the Wikipedia definition, which currently builds 'centreless' into the definition ….) I bring this up because it seemed that @JustinBenfield was asking about *abstract* isomorphism.

  • I don't think that 'centreless' is a consequence of the abstract isomorphism $G \cong \mathrm{Aut}(G)$—at least, not obviously so—but rather of the *particular* isomorphism $g \mapsto \mathrm{Inn}(g)$. (It seems that the first part of the comment suffices to reproduce the problem.) – Martin Sleziak Jul 18 at 5:37
  • I don't think that 'centreless' is a consequence of the abstract isomorphism $G \cong \mathrm{Aut}(G)$—at least, not obviously so—but rather of the _particular_ isomorphism $g \mapsto \mathrm{Inn}(g)$. (The same problem is there if we use underscore instead of star for italics.) – Martin Sleziak Jul 18 at 5:38
  • 1
    I don't think that 'centreless' is a consequence of the abstract isomorphism —at least, not obviously so—but rather of the particular isomorphism. (What happens if I omit the MathJax from the comment?) – Martin Sleziak Jul 18 at 5:40
  • One more try: is it the dash? —*particular*—. It seems to be. – LSpice Jul 18 at 15:33
  • 1
    @LSpice Regarding the last comment with " —*particular*—" as compared with "— particular —" AFAIK formatting like this works in comments only if surrounded by some characters (spaces, comma, full stop...). You can find a few posts about this on meta, like; Italics markup not working in comments. – Martin Sleziak Jul 18 at 17:13
  • 3
    (I deleted several intermediate comments exploring various possibilities, and have re-posted a less mangled version of this comment.) Good point; I was just trying to edit down to an MWE. "$\cong$—spacer *particular* spacer—$\mapsto$" seems to exhibit the same effect, but "—spacer particular spacer—" doesn't, so I guess it's the MathJax + em-dash. – LSpice Jul 18 at 19:05
  • 1
    Since we seem to have a reproducible bug, does this post constitute reporting it, or is there some more central repository to which I should report it? – LSpice Jul 19 at 19:20
  • 1
    @LSpice I think both possibilities are ok, leaving just this post or making a bug report on Meta Stack Exchange. SE employees check also local metas - it was mentioned recently in this comment - so they should notice your post eventually. However, on the main meta the post would get more exposure. (Moreover, this is not specific just for MO, there are also several other sites that use MathJax). – Martin Sleziak Jul 19 at 19:31
  • 1
    @LSpice Here's a slightly more minimal example: $\cong$— *italics* $\cong$: code $\cong$— *italics* $\cong$. The lack of a space between the MathJax and em-dash is necessary ($\cong$ — italics $\cong$: code $\cong$ — *italics* $\cong$) , and there also must be MathJax on both sides (($\cong$— italics: code $\cong$ — *italics*). – j.c. Jul 20 at 19:19
  • 4
    @LSpice I'd guess it is worth linking the post on the main meta also here: MathJax + Markdown rendering in comments. – Martin Sleziak Jul 21 at 5:56

You must log in to answer this question.

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