@Minsky not sure why it doesn't work in the other chat but I wouldn't run a personal mathjax extension on an interactive site like this as it means you do not see the text everyone else sees, which confuses the conversation.
@DavidCarlisle Since everyone writes $formulas$ in the chat in the Physics SE forum, I came to think that they all run some extension. Of course, maybe I should just ask. Also, it's strange this isn't enabled by default, it's just a js script.
@DavidCarlisle I see, but in PSE (chat) I don't think it's just a setup, in the chat room itself, there is a link to this post: meta.stackexchange.com/questions/97938/…, which is where i found the extension.
but i think you meant in the Q&A site itself, which makes sense.
just read a comment there lol > amazing. let me install mathjx so i can type formulas as well
@Minsky sorry yes I only use chat here. But the basic comment I made above still applies, it doesn't help to run a personal mathjax in chat (unless everyone does it)
@tush I think the math font is more or less the same font as for the text. Of course, the two integral signs are quite special and were probably taken from another set of glyphs. I am really not sure which math font would have glyphs similar to these.
@DavidCarlisle -- Yeah, I noticed that, but sadly, I'm left answerless. Whoever came up with that formulation deserves to be subdivided. It looks intractable in any sensible system. Wait and see what Asmus has to say.
@DavidCarlisle -- Except I am pretty sure that formulation wasn't physically present in the published TR. (In other words, I'm trying to weasel out of responsibility.)
@DavidCarlisle should I make him jealous too? In the proposed table the meaning of X and ? is imho a bit unclear, does that mark things that are there (but what about digits?) or that should be there (but greek should be filled up).
@UlrikeFischer X means there ? means not there (in the current definition for the first table and in the proposal for the second) I thought about using checkmark and cross but wanted to keep it ascii (using Unicode for a unicode proposal seemed too adventurous)
@Skillmon this here (and you do not need to worry imho, the ho-tex test are questionable here, I often have to remove some invalid settings when the code changes.
@UlrikeFischer well, yes, that's something on which I will not act. etl defines a faster but slightly less stable version of \tl_if_head_is_group:nTF that is usable because I have total control over the expansion on those points.
(and I don't think that the ho-tex approach of setting up every catcode at the start of a package is reasonable)
@Skillmon I quite agree. But it does it only in some test files (to check if some catcode changed) but it is a pain if you add a char that wasn't previously in the code ...
@Nasser most likely there were already replies to this, but: Ford will tell you: You bought a truck which is designed to carry heavy loads, not to run fast. If you want a fast car, buy a Mustang.
