On 10/07/2025 09:10, Jacob S. Gordon wrote:
Good to know! I tried with all the latest versions, and I can confirm
only `MathJaX` and `tex4ht` wrapped the nested subscript in an
`<mrow>` (excessively in the latter).
Has this been reported upstream anywhere?
I did not try to find existing bugs. I know about MathML almost nothing,
so it is hard for me to reason if it is really necessary and what issues
its absence may cause. My impression is that Eric was rather busy.
On 2025-07-09 07:23, Max Nikulin wrote:
When I tried latexml, for some reason it was excessively aggressive in>>
respect to removing files from /tmp.
Do you recall what it was doing with `tmp`? Removing logs? Files it
shouldn’t be touching?
I tried commands like
latexmlc --profile=math literal:"\$f' = df/dx\$" | tee /tmp/m2.xml
and I noticed that some files were removed, maybe even /tmp/m2.xml or
similar files with results of earlier commands. It was a VM with Debian
12 bookworm and latexml installed as a .deb package from system
repositories. Taking into account the expl3 issue I decided that it is
no go.
My initial goal was to prevent injections of arbitrary shell commands
through math snippets. I just tried exporting from Org using real MathML
tools besides stub commands that are enough to test escaping of special
characters.
Have you managed to run MathJax locally outside of browsers (using node)?
There is an issue with MathJax export: "<>&" characters are not are not
converted to "<" and similar entities. Some refactoring is required
to use the same function in ox-odt, ox-html, and for preview (so the
function should not be placed into ox.el)
Re: Inequalities in math blocks. Wed, 06 Oct 2021 09:39:23 +0200.
https://list.orgmode.org/m2bl42bo0k....@me.com
