@TorbjørnT. :38292189 Ah. Well such questions are difficult to answer under stress. My first reaction would be "how should I know and what sense does it make if I speculate? Did the poll try to find out?". There are so many possible reasons (but probably because it doesn't matter to them. It is more important for them to be able to locate their house and their work place then the US. To quote Sherlock Holmes: Why should it matter if the sun or the earth is in the center of the solar system?)
@CarLaTeX Hi Carla, I am really sorry for my awful late answer. Please forgive me.
I am really busy. The job is challenging, I am working on a bunch of projects in my spare time. Two weeks ago, my wife had an bad accident, when she was traveling in our car. Given the severity of the accident, we are lucky, that no one was harmed more. It was great luck. Nevertheless, it was a great shock for her and we are very lucky, that the garage was able to repair our car and not having an economic desaster.
My question yesterday arouse of one of the LaTeX-projects. So I am still busy TeXing :-)
Wednesday next week, I will have a speech of an hour about Typography, but I wasn't able, to prepare more than the title slide :-(
Fortunately, my vacation will start on Friday! Hurraayyyy!
General question: Do we have some documentation or similar on how to handle the following situation? Otherwise, maybe we should make one. Here is the situation: Given a text font, available in roman, roman bold, italics and italics bold. Including punctuation. How do we proceed to mix this with a suitable math font such that, letters (including support for \bm) are taken from the italic text font,
the punctuation from the text font is used, including bold, base sizes of parentheses, brackets etc are taken from the text font. That operator font is the text roman. that \mathrm is text roman. Etc.
The setup I have right now is really bad and not always working because I don't understand the font setup well enough (it is rather poorly described in the LaTeX companion, sorry Frank, at least I did not understand half of it).
So I was thinking of redoing the setup I had for this university text font I had and see if we could document the process for future use.
@CarLaTeX Thank you. My wife is already fine again. No severe injuries for all three persons in the accident, as far as I know. We can really be happy, that everything was so harmless!
Imagine: my wife drove with ca. 80 km/h (100 km/h was allowed, when suddenly the second car drove from the left perpendicular on the road and in front of her car. We are really lucky, that nothing more or severe happened.
@DavidCarlisle Hm. I'm not sure that I agree with @egreg. With ` \\ ` it works by looking for a space. ` \[1ex] ` is an optional argument and \\ [1ex] not. But you can't do this for gather. There are certainly documents around where there is space or a linebreak before the [t]. And if you reinsert an optional argument that is not t or p or c you would break documents where someone has a typo [h] or didn't want to decide yet: [c or t, check later].
@daleif -- unless i'm sorely mistaken, punctuation in mathis taken from the roman text font. and \mathrm does use the text font. okay, this is based on computer modern and other fonts we've used here at ams, but always from the "limited" layout used only up through pdftex. so this leaves the question of the italic letters, or if you're using xe- or luatex. (and whether you want upright punctuation in italic theorems.) is this a fair description of what you're looking for?
@JosephWright -- curious about your arguments, because i'm pretty sure that the check for the space was not in the original amsmath, but was added by michael on account of general frustration with questions and what was being sent in by authors. however, can't find a record of when it was added. there is discussion regarding empty first cells in structures with alignment.
@barbarabeeton There are two competing things going on. On the one had, something like \foo [bar] will always skip spaces as that's a fundamental TeX design decision, so we'd like \\ [bar] to be consistent (which is what the kernel does as-standard). On the other hand, as you say it's not at all obvious to users that \\ then [ on a new line should be treated as an optional argument. Probably in the end this one situation is a special case.
@JosephWright -- i'm in sympathy with the desire to have \\ [bar] to be "consistent", but i'm pretty sure the texbook states that control symbols do not require a space to be "complete", and thus a space following a control symbol will be interpreted as a space (while one following a control sequence will not). so this whole topic is a special case.
@JosephWright -- i think you're correct about control symbols and optional arguments, but i'll think hard about that. (in addition to the [1ex], there's also the possibility to use * on a double backslash, but that's only a variation on the theme.)
different topic. has anyone seen a representation of an algorithm where there are "second-level" instructions, numbered separately from the main list similar to a nested enumerate? if so, pointers to examples (preferably with input code) would be welcome.
Here is another interesting one for the users on the list. Have any of you ever dealt with MathType and InDesign? En Emeritus Professor at my department is writing his memoirs. A publisher is making the book in InDesign. They cannot handle LaTeX.
Normally their math books are written in Word (when they get them) using MathType. An interesting feature is that MathType can export the equations (line breaks are not allowed) as EPS, where in the EPS there is a comment of the placement of the baseline in the EPS (in relation to the bounding box).
So my question is: Can we do something similar with our toolboxs, LuaLaTeX are fine. I think it is an interesting challange. Just do not know how to get the baseline info out.
@PauloCereda Thanks. I was thinking about writing a script that can extract all the math (which will be converted into eps files) and make a new manuscript where all math is replaced by references to the filnames of the eps files. We just need a way to insert a comment into the EPS explaining where the baseline is in relation to the bounding box. Then the publisher has some javascript for InDesign that can adjust everything.
Thus we only need to decide on a math font and font size.
@daleif generators such as tex4ht are doing almost all that, extracting the math and making images and calculating the baseline offset, they just don't put that in a comment in the eps but that's surely just a detail....
@DavidCarlisle Might be interesting to look into that, since we can always just post process the baseline comment information into it. Is tex4ht using eps or PNG?
@PauloCereda might be difficult to convert that in relation to the bounding box generated for EPS.
@PauloCereda Can't we cheat? $formula$ -> EPS, BB gives us the totalheight in bp, then $\rule{1pt}{30bp}\quad formula$ -> EPS gives us a simular EPS, but the total hight is now 30bp + the depth we are looking for. The rest is arithmetic. Since we are extracting the stuff via a script anyways we can easily make two eps files.
@mickep It is not available for Linux, so not in my World. The publisher mentions that even using MathType everything is still a pain, as the BB for the EPS files are often too wide, so they have difficulties dealing with $formula$. the dot after the formula gets too far away
@mickep Their layout is fine. We can probably make it in LaTeX given time. But this is why he has a publisher, they take care of stuff like this. Math is just difficult for them
@michal.h21 can we get the baseline information out? Makeing the EPS' are triviel, I'm looking for figuring out where the baseline is in relation to the BB of the EPS.
I don't think I'd like to handle HTML, EPS is something they've done before, their tools just need to know the baseline in relation to the BB
@daleif they want only eps? I think I've done some experiments with putting math into boxes and measuring the baseline in the past, but it was quite hackish
It seems I can do it with $formula$ -> eps and $\rule[-30bp]{1pt}{30bp}formula$ -> eps and then analyse the BBs of those. Need to do a bit of statistics
@JosephWright @DavidCarlisle have you guys messed with floats such that labels before \caption now never resolves? Because thumbs up for that. I have a manus I'm cleaning up, the user is still on TL16, my TL17 notes unresolved on all his figure references because he generally uses \begin{figure}[...]\label{...}
@daleif I don't think this is a change in behaviour.
The following gives permanently unresolved references in TL15, TL16 and TL17.
\documentclass{article}
\usepackage{pgffor}
\begin{document}
\foreach \x in {1,...,5}{
\begin{figure}
\label{foo\x}
\caption{Foo\x}
\end{figure}
}
\foreach \x in {1,...,5}{
As we can see in Fig.~\ref{foo\x} there is something.
}
\end{document}
@AlanMunn Also if there is a \section{test} before it? Aka there is actually something in \@currentlabel. It is interesting that the user never noticed that they are all wrong. I guess that is why I'm here ;-)
@AlanMunn that is the point, I'm my doc they always fail even if there is, say, a numbered section earlier on. I haven't changed anything related in memoir...
@AlanMunn not an ordinary author ;-)
Slight OCD comes in handy for a manuscript like this
Ahh, no it is the caption package forgot all about it. Sorry @DavidCarlisle and @JosephWright too late in the day. That ought to go into the kernel BTW
@AlanMunn I've converted the the manuscript and his original did probably not use the caption package.
@yo' -- i tried it too, and got the same result you did -- failed the first time, connected the second. one thing i noticed was that the url that showed the first time ended with index.php or some such, but the url that's in the web page doesn't have that. so something is apparently being assumed in/by the browser.
however, re icm and tug, the icm satellite conference listing doesn't yet show a listing for tug. @PauloCereda -- alert!
@yo' To be honest, I wasn't planning on attending ICM because I am too dumb on math affairs, but I could extend my plans to visit the city and have a nice time around! :)
Hello, off topic question. In proving a theorem, I need to use several subsections. In this case, should I just say, "we devote the remainder of section 1 to proving this theorem" and not create a proof environment, or should I create a proof environment with several subsections inside of it?
@TheSubstitute The proof environment is used to make the proof stand out from the context, if the entire section is devoted to it there's no need to use more emphasis.
@FrankMittelbach (This is a reply about the \newpage fix.) First of all, let me thank you for the time you have spent reviewing my little suggestion. My idea was precisely that a solution thst cuts down the number of bytes written to the .dvi (or other output) file is more efficient. Knuth discusses a similar issue on p. 374, lines 10–15, of The TeXbook. See, in particular, the last sentence (“Input/output time takes longer than computation time…”).
