TeX, LaTeX and Friends

12:21 AM

@TheSubstitute -- it may not be all that difficult to add a box at the end, even if it isn't within the scope of a proof environment. what theorem package are you using, or, if none, what document class? and does your proof end with text, or with (i hope not!) a display?

CarLaTeX

6:14 AM

@PauloCereda should answer this: tex.stackexchange.com/q/376417/101651. I think he's an expert :)

Paulo Cereda

7:09 AM

@yo' Really? That's very interesting!

@CarLaTeX oi

4:12 in the morning, getting ready for SP!

CarLaTeX

@PauloCereda Arghhhh

Paulo Cereda

@CarLaTeX I am an early duck. :)

CarLaTeX

7:25 AM

@PauloCereda too early for me :)

yo'

7:47 AM

@PauloCereda I mean, why spend almost 3 weeks in Rio when you can look around? :)

David Carlisle

@JosephWright remember what you were saying about me not using latex?

(/usr/local/texlive/2017/texmf-dist/tex/context/base/mkii/supp-pdf.mkii
[Loading MPS to PDF converter (version 2006.09.02).]
! Use of \@@mptopdf@@newabove doesn't match its definition.
l.136     \@@mptopdf@@newabove \csname n
                                        ewcount\endcsname \scratchcounter

Joseph Wright

@DavidCarlisle ?

David Carlisle

@JosephWright yesterday I got ~~tricked~~ invited into yet another github organisation and tried to regenerate the OpenMath specification....

Joseph Wright

@DavidCarlisle Ah: link?

@DavidCarlisle Speaking of GitHub, I was tempted to point out to Chris/Ross that the 'modern' way that discussion happens is GitHub issues/pull requests/branches/Trello/Gitter/... or similar (cf. GitLab), but I thought that might not go down so well ...

David Carlisle

@JosephWright ah found it..... because the openmath spec is generated from xml (see I didn't actually write it in latex syntax:-) it is set with "novel" catcodes to avoid having to quote everything and Hans' MPS reader apparently objects to # being catcode 12.....

@JosephWright the catcodes are normal when graphics is loaded but that gets read at begin document and the end of the preamble is

\let\cellsep&
\catcode`\&=12
\catcode`\_=12
\catcode`\^=12
\catcode`\#=12
\catcode`\$=12
\catcode`\£=1
\catcode`\ß=2

\begin{document}

@JosephWright and Heiko takes more care of me than we do....

@JosephWright still i suppose this input is weird enough that we don't need to support it, I'll just wrap the catcode setting in AtBeginDocument so it comes later.

CarLaTeX

8:21 AM

@UlrikeFischer I think you could trasform your comment into an answer here: tex.stackexchange.com/q/376434/101651

Jonas Stein

10:05 AM

Hi, how can I improve the following siunitx expression

        \SIrange[per-mode=fraction]{1e7}{10e7}{n \per \centi\meter \per \second }

in a way that I get 1 combined unit in the end with a fraction?

(sorry I am always confused with the formatting in this chat)
now it looks like

yo'

Electricity off in a large neighborhood :-(

egreg

10:26 AM

What's “n”?

Joseph Wright

@JonasStein You've used literal mode (by having the n), so siunitx does no reformatting (and per-mode = fraction does nothing)

Jonas Stein

10:43 AM

n is the number of neutrons

egreg

@JonasStein You should define a special unit for it.

Jonas Stein

I see. Thank you!

I would expect that I get an error, if siunitx can not execute my command... I think I would have searched the whole weekend without your hint.

if n is the number of neutrons, it must be italic in the unit, right?

egreg

@JonasStein It should be whatever it is in your field.

Jonas Stein

definition of \neutron worked, now I have to merge the units to one

Joseph Wright

@JonasStein range-units = single

Jonas Stein

10:53 AM

@JosephWright Thank you! ah on page 42... It is difficult to find, if one does not know the keyword.

would it make sense to add it as example to page 7, where the sirange is explained?

Joseph Wright

@JonasStein Not really: I can't have all of the options 'up front', I have to make some kind of decision on spreading them out

daleif

11:50 AM

@JosephWright @DavidCarlisle I'm playing a bit with the hiresbb option for graphicx: may I suggest that it should not fail if %%HiResBoundingBox is not found, maybe just fall back to the standard BB

Paulo Cereda

12:10 PM

@yo' Agreed, although you might consider that Rio and São Paulo are not the typical European metropolis -- Rio with near 7 million people and São Paulo close to 20 million!

@JosephWright pitchforks and torches, here we go!

yo'

@PauloCereda 7 mio is smaller than Paris :-) and 20 mio is just about Paris + London :-)

Paulo Cereda

@yo' ooh math skills <3

@yo' we could talk with the other blokes and try to make a collective adventure pack!

yo'

@PauloCereda may be :) However, I don't think many people will stay for ICM

Paulo Cereda

@yo' Let us poke people, I love doing that! :D

@DavidCarlisle /poke

Joseph Wright

12:25 PM

@daleif For 2e? That's @DavidCarlisle's call

Paulo Cereda

@JosephWright s/call/fault/g

daleif

12:43 PM

It was just that this is the only thing it says in the manual: `hiresbb Look for size specifications in %%HiResBoundingBox lines rather than
standard %%BoundingBox lines.` it does not mention that LaTeX fails if no `HiResBounding` is found in an EPS.

David Carlisle

1:04 PM

@daleif it might be possible now but wasn't then, currently it is essentially a zero cost option as it just redefines the macro holding the search string from %%BoundingBox to %%HiResBoundingBox but to do what you suggest would be considerably more complicated as on every line you would have to search for both as you don't know there is no hiresbb until the end so you would need to save the boundingbox just in case. I don't think I had enough free token memory tomake that work in emtex:-)

Joseph Wright

@DavidCarlisle Indeed: look at Frank's worry on the team list that people might end up using a macro in packages for one job only, an my response: 'so?'

@daleif Can I ask the use case for HiResBoundingBox? I'm not really sure it's that useful ...

daleif

1:29 PM

@JosephWright see my problem yesterday regarding making eps files for InDesign. dvips -E is not really precise enough, it my be 0.5bp when we try to calculate the baseline skift needed. That is noticeable in the output. So far the InDeign people have not really been impressed with what we can do in comparison with for example MathType.

Which is really good at inserting this baseline information even just a whole number basis.

I generate the HiResBB using epstool

I still need to do this on a larger number of equations to get some better statistics, I'll have a look at that next week.

Alan Munn

2:58 PM

@JosephWright This question tex.stackexchange.com/q/263938/2693 just got bumped. It seems to have very little usefulness (very dated, and quite local.) Can it just be deleted?

Steven B. Segletes

4:04 PM

Did something change when I wasn't looking? When I invoke \includegraphics[width=1in,decodearray={1 0 }]{example-image-A} I now get an error: Package keyval Error: decodearray undefined. What's up with that?

David Carlisle

@StevenB.Segletes well... (real question) how did you know the key was ever there, it's not in the graphicx documentation for example @JosephWright

Steven B. Segletes

@DavidCarlisle ummmm, because numerous Q's and A's make use of it on this site???

I must be in the alternate universe. Cool!

David Carlisle

@StevenB.Segletes they do? wow:-)

egreg

@DavidCarlisle tex.stackexchange.com/search?q=decodearray

Joseph Wright

@StevenB.Segletes We've been trying to get things back into some kind of order

@StevenB.Segletes If it's used I can happily put it back

David Carlisle

4:10 PM

@StevenB.Segletes anyway we'll probably put it into graphicx if it's used, the back end files were never supposed to add undocumented keys so that keys worked with [pdftex] but not [dvips] so if it goes back, it won't go into pdftex.def but into graphicx and we'd document what it does.

what does it do? :-)

Steven B. Segletes

@DavidCarlisle It's like a DVD Easter Egg...it allows you to impress friends and influence people with arcane knowledge. Actually, I was just reconsidering Bruno's question, tex.stackexchange.com/questions/150225/…

David Carlisle

@egreg A bunch of very untrustworthy people show up in that list, as I'm sure @StevenB.Segletes would agree...

Steven B. Segletes

@DavidCarlisle ...of the worst sort, indeed!

David Carlisle

@JosephWright humph put it into graphics with a default definition of "this does nothing" and enable in drivers that can do it?

Steven B. Segletes

One might seven say...shady. (LOL)

David Carlisle

4:15 PM

@StevenB.Segletes get out:-)

@StevenB.Segletes Seriously if you were wanting this now, we could provide a patch to put it back....

Steven B. Segletes

@DavidCarlisle Well, here is a good example of where it was actually useful: tex.stackexchange.com/questions/194378/…

and maybe this: tex.stackexchange.com/questions/222849/…

Also for watermarks: tex.stackexchange.com/questions/332722/…

daleif

Could someone give a good answer to this one: tex.stackexchange.com/q/376485/3929

Jonas Stein

@JosephWright not all. Just write the options I need in the first 10 pages ;-)

David Carlisle

@daleif is "it does" a sufficient answer?

Alan Munn

\documentclass[twocolumn]{article}
\usepackage{pgffor}
\makeatletter
\newcommand\showsize{\f@size pt\par}
\begin{document}
\foreach \x in {1,...,100}{
\fontsize{\x}{1.2\x}\selectfont\showsize
}
\end{document}

@daleif I'm inclined to post this: ^^^^

@daleif But I think the question confuses this issue with the fact that document classes have restricted size options. (Which is a more reasonable question.)

David Carlisle

4:34 PM

@daleif I gave an answer, not sure if it is good:-)

@AlanMunn oi I was still typing:-)

Alan Munn

@DavidCarlisle :) are you referring to my edit? Or the comment that nobody had answered?

daleif

I think the comments and answer is fine. It was also clearer from Alans comments that a clarification was needed.

David Carlisle

@AlanMunn the edit as it complained that I couldn't save a p as it was less substantial than someone else's edit:-)

Alan Munn

@DavidCarlisle :) well try to keep up.

@DavidCarlisle The part you just added is really helpful, btw.

Alan Munn

5:21 PM

I've just encountered a weird problem: pdflatex is loading a .bbl file from the wrong folder. So my file is samples/foo.tex and when I compile it it finds TestDocuments/foo.bbl which throws an error because that file was created by biblatex while foo.tex in samples/ uses natbib. This shouldn't happen, right? Both samples/ and TestDocuments/ are in the same enclosing folder.

egreg

5:33 PM

@AlanMunn Cannot reproduce. Do you have strange settings to TEXINPUTS?

Alan Munn

@egreg No, nothing strange at all.

David Carlisle

@AlanMunn are you running latex from the common parent?

Alan Munn

@DavidCarlisle I'm running from TeXShop or from the samples/ folder directly with the command line. No difference in behaviour.

David Carlisle

@AlanMunn can't happen then:-)

Alan Munn

@DavidCarlisle And just in case it was a TeXShop issue, I restarted TeXShop too with no change. (But since I can reproduce with the terminal it's likely nothing to do with TeXShop.)

@DavidCarlisle Humour my delusions.

egreg

5:38 PM

@AlanMunn kpsewhich -var-value TEXINPUTS says?

David Carlisle

@egreg ..// ?

Alan Munn

@egreg

.:{{}/Users/alan/Library/texlive/2017/texmf-config,/Users/alan/Library/texlive/‌2017/texmf-var,/Users/alan/Library/texmf,!!/usr/local/texlive/texmf-local,!!/usr/‌local/texlive/2017/texmf-config,!!/usr/local/texlive/2017/texmf-var,!!/usr/local/‌texlive/2017/texmf-dist}/tex/{kpsewhich,generic,}//

egreg

@AlanMunn Identical to mine, except for alan, of course.

Alan Munn

@egreg Not surprising, since I have never messed with it.

egreg

@AlanMunn Relevant part of the log?

Alan Munn

5:45 PM

(/Users/alan/Library/texmf/tex/latex/msu-thesis/TestDocuments/MSU-thesis-testfi
le.aux
/Users/alan/Library/texmf/tex/latex/msu-thesis/TestDocuments/MSU-thesis-testfil
e.aux:4: Undefined control sequence.
l.4 \abx@aux@sortscheme
                       {nyt}

(This is where is loads the wrong file).

Actual file is: **MSU-thesis-testfile.tex
(./MSU-thesis-testfile.tex
in the `samples/` folder not `TestDocuments` (although there's a copy there.)

@egreg ^^^ anything more useful needed?

@egreg I just figured it out. It's loading from the local texmf folder, that's why, because on the first run it doesn't find anything in the current folder.

@egreg This is an artefact of the fact that my development folder is symbolically linked with my local texmf folder.

Alan Munn

7:07 PM

@JosephWright This question tex.stackexchange.com/q/376061/2693 is closed as a duplicate, but the answer there doesn't even address the question. The question itself has a self answer which is correct, so can we reverse the duplicate direction?

egreg

7:24 PM

@AlanMunn Ah! That could explain the issue!

Alan Munn

@egreg Yes, that's exactly what was happening. I've never run into the problem before because I only had one test document. But as the testing became more extensive I created a folder of different versions. Disaster! :)

@egreg BTW any advice on how to win an \AtBeginDocument fight? (the apacite package does all sorts of things using it, which is a pain. I'd like mine to win.)

egreg

@AlanMunn Psychologists are known to be crazy; so using APA is something I consider very low in my list.

Alan Munn

@egreg Flagging this as "Not an answer". :)

Joseph Wright

7:59 PM

@DavidCarlisle I'm going to look at whether decodearray will work with (x)dvipdfmx: I think it should

@AlanMunn Done

egreg

8:31 PM

@AlanMunn Ask Chuck Norris, that I had to mention in my last answer: there was Jean-Claude van Damme; I also added Steven Seagal, of course.

Alan Munn

@egreg And you think psychologists are crazy...

egreg

8:51 PM

@AlanMunn I'm a mathematician, you know…

David Carlisle

@JosephWright if we put that back in graphicx is it worth doing the others at same time

Alan Munn

3 hours ago, by Alan Munn

@DavidCarlisle Humour my delusions.

David Carlisle

@AlanMunn we count symbolic linking as implicit messing around with the input path, so my comment remains true

Alan Munn

@DavidCarlisle These are different delusions now.

@DavidCarlisle But yes, you were correct in the supposition. :)

David Carlisle

@JosephWright although searching the internet for anyone ever using that drew blank but did turn up \pgfimage[mask=mymask,height=3cm,width=3cm,interpolate=false]{1}

12

Q: (How) Can I create picture effects with TeX and Friends?

I'm looking for a way to create subtle picture effects in beamer. I've got a few illustrations with a background that contrast offensively against a white background, and look completely out-of-place in my otherwise tamed template. Something like the following: For the reference, this particul...

graphics

Joseph Wright

9:15 PM

@DavidCarlisle If we put it in graphics then I'll work out as many drivers as I can (also for L3)

@DavidCarlisle Doing what?

@DavidCarlisle I did check on pgf before removing stuff: if Till doesn't use it, it's likely not that handy

David Carlisle

@JosephWright including an image with masking

Joseph Wright

@DavidCarlisle Till's version of masking is more sophisticated than the one we removed from pdftex.def

David Carlisle

@JosephWright yes

Joseph Wright

@DavidCarlisle Long-term L3 plan is to cover all of this, but I'm not sure at the moment that any 'real' users know how to use this stuff

David Carlisle

@JosephWright but that (possibly) means that the simple feature might be useful, or perhaps it doesn't.

@JosephWright no but I'm slightly worried we find some publisher workflow is using them and we only find out down the line when they update

Joseph Wright

9:21 PM

@DavidCarlisle I'm suspicious

@DavidCarlisle Possibly, but most of them are likely to be using DVI and to want the images 'as is'

@DavidCarlisle Perhaps one for the team meeting

David Carlisle

@JosephWright perhaps I should ping Heiko, at least with masking one could imagine a use, but what the .. is ocobjnum about? could anyone use it given the comments in pdftex.def?

Joseph Wright

@DavidCarlisle ocobjnum is extremely low-level

David Carlisle

@JosephWright yes but exposed as a user key for includegraphics...

Joseph Wright

@DavidCarlisle I know

@DavidCarlisle Undocumented

@DavidCarlisle I looked over the masking one: I think Till's more complete implementation is the way you'd want to use it, but not in graphics (you want a more complex framework)

David Carlisle

9:36 PM

@JosephWright well documented as comments in the old pdftex.def so same status as interpolate and decodearray

Joseph Wright

@DavidCarlisle Yes, yes; I know

David Carlisle

@JosephWright meanwhile I'm wondering abut travis ci, know anything about that...

Joseph Wright

@DavidCarlisle Might do

@DavidCarlisle One for Skype?

David Carlisle

@JosephWright probably not for this weekend at all but ive been gently persuading a 1999 workflow of xml-tex-pdf+html to work with current tools and currently the source and final result are on github so I figure if I had travis ci with java and tex it should be able to automate the whole lot

Joseph Wright

@DavidCarlisle Yup

Joseph Wright

10:04 PM

@DavidCarlisle Key is to work out what you need installed: I'd be tempted to set up a VM to get that right

David Carlisle

10:16 PM

@JosephWright I'd skip tex initially so probably just need java (+ a jar file I could check in to the repo or install from the net, whichever is more convenient) so just needs a vm running bash and java if I get that setup working making html I could add the tex to pdf variant later. I might try it on a throw away repo of my own rather than experimenting in an "organisation" repo where I'm supposed to know what I'm doing:-)

Robert Marshall Murphy

10:26 PM

I have a huge book I'm working on in LaTeX, using the The Legrand Orange Book as a starting point. I cannot construct as MWE, but I cannot make makeidx work. How can I track down my "main.ind:1: Missing number treated as zero" error?

David Carlisle

@RobertMarshallMurphy error from tex or makeindex?, what does the error look like exactly in the log?

Robert Marshall Murphy

There is no error as long a I don't makeindex. Once I do, the main.ind produces a fleet of errors, and I have to delete it.

Which log?

If I try to plow through (i.e. enter) it only gets hairier

David Carlisle

@RobertMarshallMurphy oh you are using the -file-line-error option that's why I didn't recognise the format. so the error is on line 1 of main.ind, what is line one of that file?

Robert Marshall Murphy

David Carlisle

OK so your definition of theindex is bad, stick \tracingall before \begin{theindex} and get a more informative error log...

Robert Marshall Murphy

10:40 PM

But I didn't make this file. It's automatically generated. What did I do?

David Carlisle

@RobertMarshallMurphy yes sure but the file looks OK but theindex is defined by your class and probably redefined by some package that is loaded so makeindex generated \begin{theindex} but that isn't starting an index, if you stick \tracingall in front the log will show what it is doing

Robert Marshall Murphy

The log file is much bigger now, but I still get the same error

David Carlisle

@RobertMarshallMurphy yes

Robert Marshall Murphy

Thank you so much for helping me. I only ever wanted to teach calculus, not become a typesetter! ;-)

David Carlisle

@RobertMarshallMurphy image is no good paste it as code into here, just from where it starts the \tracing,,,,

Alan Munn

10:45 PM

@RobertMarshallMurphy What you're calling the log is really the console output. When you get the first error, type X in the input area there and then use Show Log File in the File menu to look at the whole log.

@RobertMarshallMurphy Although it will be huge.

David Carlisle

@AlanMunn tis true althoigh after tracingall there is not so much difference as that turns on tracingonlne

@AlanMunn from \begin{theindex} to where it died shouldn't be that big, it's not like tracing beamer:-)

Alan Munn

@RobertMarshallMurphy You can copy from the console output as long as the run is complete. (Type X in the input area, then you should be able to copy stuff.)

Robert Marshall Murphy

texshop keeps crashing. Hold on a minute.

.2 \begin{theindex}

? x

Here is how much of TeX's memory you used:
60121 strings out of 492995
1294152 string characters out of 6132704
1952460 words of memory out of 5000000
60702 multiletter control sequences out of 15000+600000
92327 words of font info for 279 fonts, out of 8000000 for 9000
1141 hyphenation exceptions out of 8191
64i,14n,103p,10488b,2737s stack positions out of 5000i,500n,10000p,200000b,80000s
{/usr/local/texlive/2017/texmf-dist/fonts/enc/dvips/cm-super/cm-super-t1.enc}{/
usr/local/texlive/2017/texmf-dist/fonts/enc/dvips/newtxtt/txttDec.enc}{/usr/loc

David Carlisle

\columnsep 35\p@ \let\item\@idxitem is supposed to be setting col sep to 35pt and locally definiing item but something went wrong earlier so it is _typesetting _ 35 then trying to assign \p@ to be \let and failing.

@RobertMarshallMurphy that's not the interesting bit, copy the text from the log file from the first \tracing-> line to the line with the error.

Robert Marshall Murphy

There is an "\XC@tracing ->0" on line 19246 out of 297117....!

I cannot get TexShop to re-open now. I've never had this happen before

David Carlisle

10:59 PM

@RobertMarshallMurphy that's from \tracingall immediately before \begin{theindex} ? surprising amount of log from one line of input?

Robert Marshall Murphy

Yeah, it's a 10.6MB file. I deleted the PDF and I'm back in business.

David Carlisle

@RobertMarshallMurphy that's somewhat bizarre. It;s not unusual to get big traces from tracingall (I've had much bigger than that:-) but if you get that much from a single \begin{..} it usually means something has gone off in a loop somewhere, but anyway I'm not sure where you are but here it's just turned midnight so I'll give up, sorry:-)

Robert Marshall Murphy

No problem. Thank you for you time.

Paulo Cereda

11:44 PM

Home sweet home. :)

What did I miss today?

barbara beeton

@cfr -- i didn't see your comment -- where should i look?

Robert Marshall Murphy

Where can I learn about how to read an error log?

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends