« first day (3047 days earlier)      last day (1871 days later) » 

marmot
marmot
3:12 AM
user image
8
^^^^ I should really try to install @samcarter's magic...
 
 
6 hours later…
Faheem Mitha
Faheem Mitha
8:48 AM
@marmot Assuming you are a Linux user, what distribution do you favor? Just idle curiosity.
I've got KOMA 3.25. The latest version appears to be 3.26b. Do I really need to install the most recent version before reporting a bug? If so, dammit.
 
manooooh
manooooh
When we are writing a cases environment we use \text{if \(...\)}
First question. Is "if" a math operator? I think yes, because it is equivalent to "\wedge", right?
Second question. If the above has a "Yes", then if we are writing in another language instead of English then we should write "if", not translating it (I remember this because of a comment of @egreg), right?
(I mean, we should preserve the original, e.g. instead of (mód 5) we should write (mod 5))
However, there are a lot of spanish autors that write "si" (means "if" in english) (including me) but they also write "(mod 5)". Why is this happening? It seems a paradox, right?
If it is not a paradox, why the non-english authors do not preserve the original in this particular case?
 
Ulrike Fischer
Ulrike Fischer
9:04 AM
@FaheemMitha well if you have minimal example, you can ask someone with a newer version to make a short test. Or you report anyway but mention that it could already be resolved in a newer version but that you don't have the means to test.
 
Faheem Mitha
Faheem Mitha
@UlrikeFischer I see Debian has a newer version of TeX Live, but it's a bit of a chore to build.
Does anyone know if that is a prerelease or actually 2019?
The version number is 2018.20190227-2, so I'm guessing it's a prerelease.
Also is uploading a PDF to SE still not supported?
 
daleif
daleif
@manooooh in Danish we translate if to the Danish equivalent. I don't see it as an operator, it is just text. mod is often not translated as it comes from Latin (though it may vary)
@FaheemMitha I think this was discussed on the texlive list. Debian was using strange numbering. Norbert had an explanation. They should not take it before it is released. I think a Fedora did that once.
 
Ulrike Fischer
Ulrike Fischer
@FaheemMitha pretest has started: tug.org/texlive/pretest.html. We all are already using it starting from here: chat.stackexchange.com/transcript/message/49256645#49256645 ;-)
 
Faheem Mitha
Faheem Mitha
@daleif "They should not take it before it is released." You mean Debian should not package prereleases?
 
manooooh
manooooh
@daleif I think we are not handling the same definition of a "case". For example, \begin{cases}x&\text{if \(x\ge 2\)},\\0&\text{if \(x<2\)}\end{cases} is translated as the set \{x\in\Bbb R: (x\wedge x\ge 2)\vee(0\wedge x<2)\}. Do you agree with this?
 
Faheem Mitha
Faheem Mitha
9:19 AM
@DavidCarlisle And like Lake Wobegon
 
Ulrike Fischer
Ulrike Fischer
@FaheemMitha Are you meant the binaries from debian. Norbert explained here about the version numbers: tex.stackexchange.com/a/470815/2388
 
Faheem Mitha
Faheem Mitha
@UlrikeFischer Hmm. Read his answer. Doesn't make his numbering system clearer to me.
BTW, thank you for the answer. Replying with a comment now.
I think the leading 2018 indicates that it's a prerelease of 2019. If it is a 2019 release, it would change to a leading 2019.
If that is wrong, let me know.
 
daleif
daleif
@FaheemMitha yes, it may contain serious errors.
 
Paulo Cereda
Paulo Cereda
@manooooh IMHO it's your text and you are in charge of deciding how to make it clear to the reader. Something that I learned from experience (and thanks to @egreg for pointing me) is that one should not abuse heavy math notation when a simple text can clarifiy stuff with simple, precise words.
2
 
Faheem Mitha
Faheem Mitha
@daleif It does not sound like Norbert is listening, then.
 
daleif
daleif
9:32 AM
@manooooh perhaps logically, but not when written in spoken language. It would be similar to "$x$ if $x\ge2$ and $0$ otherwise".
 
manooooh
manooooh
@PauloCereda you are absolutely right. I sometimes think whether. Sometimes I think (it happened recently with this example) if it is better for the user to read a very formalized document or not, with the consequence of making the reading more or less understandable
 
egreg
egreg
@manooooh “If” is an English conjunction. Not a symbol
 
manooooh
manooooh
@daleif I agree with the "otherwise"
@egreg "if" is equivalent to "and" in set theory, and "and" is translated as "wedge" i.e. a logical math operator, hence "if" is also a logical math operator, hence it must be written as native language
However, you always make me fall to the ground, and that is unclear to the reader. It is always preferable to use "if" and its respective translation
 
daleif
daleif
@PauloCereda that is a good one. I edit a lot of our teaching materials and we had a discussion about when to use symbols and when not to (especially in displayed may), we ended up with if it comes last in a line, then write it in words, as in \qquad\text{for all $x \ge 2$} and not \qquad\forall x \ge 2. If it comes first then leave it as symbols, as in the well known definition for continuity.
 
Paulo Cereda
Paulo Cereda
@daleif that's a great approach!
 
David Carlisle
David Carlisle
9:47 AM
@manooooh that seems a strange translation, as your set is just {0} union {x | x>=2} but the cases construct is normally interpreted as a function, N->N
 
Paulo Cereda
Paulo Cereda
In the UK, you get a signed birthday from the Queen on your 100th birthday.
Tweeted by UberFacts on March 5, 2019 at 8:47 AM
@DavidCarlisle ^^ hmmmmmmmmm
 
David Carlisle
David Carlisle
@PauloCereda you also get such a card on your 60th wedding anniversary (Mum and Dad got one just before dad died)
 
egreg
egreg
@manooooh Your understanding about “if” is wrong.
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle oh
 
manooooh
manooooh
@DavidCarlisle if it does not adapt to what is being discussed, think of any function in pieces of real variable. One can easily translate this type of functions to a union of sets, with the particularity that in each union the use of the logical operator "and" is present
 
David Carlisle
David Carlisle
9:50 AM
@manooooh to model a function as sets you need to use some pair constructor not just the range.
 
manooooh
manooooh
@egreg if you have to translate a piecewise function into a set, you must translate the symbols to the set theory. Inevitably, the word "if" can not fit into a rigorous definition of a set, so how do you translate it? I have seen that it is translated as the logical operator "and", and "and" is equivalent to "wedge"/"land", etc.
@DavidCarlisle what do you mean?
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle exactly.
 
egreg
egreg
@manooooh No, you use words
 
David Carlisle
David Carlisle
@manooooh consider the function abs(x) written as cases x if x>=0, -x if x<0 end cases then that is a function but you are modelling it as a set which is going to end up as {x in R| x>=0} union {x in R | x>0} which is just { x in R | x >=0} and no information about the function.
 
manooooh
manooooh
@DavidCarlisle I have been told that abs(x)={x in R: (x and x>=0) or (-x and x<0)}. If you can reduce the inner expression through operations, well that's fine, but the rigorous definition of function by parts is the one that I just gave you
 
David Carlisle
David Carlisle
9:57 AM
@manooooh no abs(x) is not {x in R: (x and x>=0) or (-x and x<0)}.
 
manooooh
manooooh
@egreg well, I guess it depends on the level of depth an author wants to have
@DavidCarlisle a logician told me it was like that
 
egreg
egreg
@manooooh No, it depends on how clear they want to be. Avoid heavy usage of symbols when explaining things.
 
David Carlisle
David Carlisle
@manooooh if takes three argumenat and and two so I can not imagine what you mean by sayoimg they are the same. in propositional logic "if a then b" is the same as "not(a and not(b)"
@manooooh a logician may have told you that a cat is a dog but it's still wrong. One is a function from reals to reals the other is a subset of the real, these are different types of things so clearly not the same
 
manooooh
manooooh
@egreg ok, so that also can be appliead to things such that S={x in R: x>=2, x<4} instead of for example {x in R: x>=2 wedge x<4}
 
egreg
egreg
@manooooh That's personal preference. I usually avoid logical symbols as much as I can.
 
manooooh
manooooh
10:02 AM
@DavidCarlisle I think this is wrong, you cannot forgive the negative part! You are changing the initial domain (reals) to [0,infty)!
 
daleif
daleif
@manooooh you are not describing the domain in any of this. You are describing the result of applying a function (described as a case statement), thus you are describing the co-doimain, which for abs(x) is all reals >=0.
 
manooooh
manooooh
@daleif oh
Well I think that depends on what kind of math you are writing. In Models, you worry about co-domain; in Calculus you worry about domain (I guess)
Since I have never worked with mathematical models...
 
Shamina
Shamina
10:55 AM
I have a naive question, I have an equation inside an integral. Being long, I would like to put in the second line but none of the stuff seems to be working for e.g. dmath, align etc.
Is there a way out?
 
Paulo Cereda
Paulo Cereda
@Shamina Dante would suggest to abandon hope. :)
@Shamina Sorry, I couldn't resist. :) I've no idea, sorry...
 
manooooh
manooooh
@Shamina matrix?
 
David Carlisle
David Carlisle
@Shamina aligned
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle ooh that environment
Highly experimental PDF resource management
 
Shamina
Shamina
@manooooh No just an equation
 
Paulo Cereda
Paulo Cereda
10:59 AM
@DavidCarlisle ^^ @JosephWright is working on something very secret
 
David Carlisle
David Carlisle
@Shamina most of the ams environments have ...-ed forms for subterms of a larger expression, so align makes a whole display, but aligned makes a multi-line term that you can put in the integral.
@PauloCereda no, he's building a playground for @UlrikeFischer
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle ooh
And the Bär, hopefully.
 
Shamina
Shamina
@DavidCarlisle in the paper format of two columns. Equation is going out of the paper
 
David Carlisle
David Carlisle
@manooooh no.
 
Paulo Cereda
Paulo Cereda
Seeing the Bär sit near @UlrikeFischer in TUG 2018 was the cutest thing.
 
David Carlisle
David Carlisle
11:01 AM
@Shamina use so use aligned to make shorter lines.
 
Paulo Cereda
Paulo Cereda
@Shamina sorry for the naïve suggestion, but perhaps breaking the equation into subequations?
 
Shamina
Shamina
@DavidCarlisle Any special package will be required?
@PauloCereda I tried but I have braces and brackets, causing problems
 
David Carlisle
David Carlisle
@Shamina amsmath
 
Shamina
Shamina
@DavidCarlisle I'm getting this Package amsmath Error: \begin{aligned} allowed only in math mode. \begin{aligned}\label
 
Paulo Cereda
Paulo Cereda
@Shamina oh I see
 
Ulrike Fischer
Ulrike Fischer
11:07 AM
@Shamina well then put it in math mode $...$.
 
Shamina
Shamina
@UlrikeFischer Thanks
@DavidCarlisle I think it's still not working due to brackets
The whole expression is still in one line. Actually I have multi line expression
 
daleif
daleif
@Shamina perhaps you should post it as a question on the site, including a minimal example so we get a better idea of what you are doing and perhaps missing.
 
David Carlisle
David Carlisle
@Shamina if you are using \left and \right` then that is a faq there are dozens of questions on site about that, dont do that use \bigl( and \bigr)
@Shamina the aligned is supposed to be after your integral sign so already in math mode, but if you do not show what you are doing how can anyone tell you what you are doing wrong?
 
Shamina
Shamina
@DavidCarlisle Precisely, I was using them. Whenever I was using `\\` then the error of brackets missing...
@DavidCarlisle Sorry for this, I was in the process of posting a ques
 
David Carlisle
David Carlisle
@Shamina it will be a duplicate there are hundreds already eg
v
13
Q: Using \left( & \right) around amsmath's align delimiter ("&")

srunniI want to align equations (using the align environment from the amsmath package) inside matching \left( and \right) commands. For example: \documentclass[a4paper,12pt]{article} \usepackage{amsmath} \begin{document} \begin{align} a \left( b &= c \right)\\ d \left( e &= f \right) \end{a...

amsmath align delimiters
 
Shamina
Shamina
11:31 AM
@DavidCarlisle Thanks a lot! this resolved the problem
 
Faheem Mitha
Faheem Mitha
12:10 PM
Is there anything resembling best practice when including numbered PDFs inside a LaTeX document wrt to page numbering? Let's suppose that the PDFs already have numbers on them.
In some cases, the PDFs might be LaTeX documents themselves.
 
Paulo Cereda
Paulo Cereda
@FaheemMitha I really do not follow...
 
Faheem Mitha
Faheem Mitha
And then there might be direct page numbering collision.
@PauloCereda If one included PDFs inside a LaTeX document, then the pages of those PDFs would automatically be numbered by LaTeX.
Unless you excluded them.
 
Paulo Cereda
Paulo Cereda
@FaheemMitha Yes.
 
Faheem Mitha
Faheem Mitha
And there is something to be said for a consistent numbering scheme.
 
Paulo Cereda
Paulo Cereda
That part I got. :)
 
Faheem Mitha
Faheem Mitha
12:14 PM
But the included PDFs might already be numbered.
 
Paulo Cereda
Paulo Cereda
@FaheemMitha life is tough. :)
2
 
Faheem Mitha
Faheem Mitha
Which would mean two different sets of numbers on the same page.
@PauloCereda And then you die.
 
Paulo Cereda
Paulo Cereda
@FaheemMitha not without a fight. :)
What I usually do is crop the margins of the included PDF and let my master document handle the page numbering.
It's what I typically do when preparing proceedings.
 
Faheem Mitha
Faheem Mitha
@PauloCereda After a lifetime of struggling with TeX, exhaustion will probably have set in.
@PauloCereda I see. Crop as in reduce the margins in \includepdf or similar?
 
Paulo Cereda
Paulo Cereda
@FaheemMitha I usually use an external tool to crop the margins for me, generating a... well, cropped document. :) But I know there are some clipping (I think that's the word) parameters you can use to selectively display areas of the included PDF...
 
Faheem Mitha
Faheem Mitha
12:18 PM
@PauloCereda Ok.
Personally, I try to get \includepdf to work a bit harder.
Those LaTeX packages need to learn that life is stern, and life is earnest.
Assuming that it's possible, of course.
 
daleif
daleif
12:35 PM
@FaheemMitha Sometimes when we prepare PhD thesis' the student have published articles that need to be included in the thesis. In many cases we use tikz to cover the header and footer of the included PDF and then add the header and footer from the thesis. Then there is only one page number for the reader to worry about.
 
JouleV
JouleV
12:50 PM
0
Q: Minimal Makefile for pdflatex bibtex with output-directory

mcExchangeDoes someone have a minimal Makefile, that compiles a latex document using pdflatex bibtex pdflatex pdflatex while putting all ouput files in a special output-directory to avoid cluttering the root-directory?

bibtex pdftex makefile
@PauloCereda They are talking about arara, and I am expecting to see the author of arara to have some sentences there ;-)
 
Faheem Mitha
Faheem Mitha
@daleif Yes, actually I thought of that. You mean, like cover it with a white circle?
 
daleif
daleif
@FaheemMitha white rectangle yes. It makes use of the picturecommand and pagecommand options for \includepdf. Picture command is executed first, so we can add tikz through it, and then use page command to add the thesis header/footer (via \thispagestyle)
 
Paulo Cereda
Paulo Cereda
@JouleV Thanks, I might add something. :)
 
Faheem Mitha
Faheem Mitha
@daleif I suppose one can use one TikZ command for all pages.
 
daleif
daleif
1:08 PM
@FaheemMitha per included PDF, yes. I usually call them ninjas, so we may have \ninjaA , \ninjaB etc for the included articles. Note that it may be needed to split the article in two: the first page (that often looks differently) and the rest of the pages from the article.
 
Faheem Mitha
Faheem Mitha
@daleif ninja?
Why, because they're sneaked in?
 
daleif
daleif
@FaheemMitha sort of, or because it is using ninja tricks that the normal users might not be aware of.
 
Faheem Mitha
Faheem Mitha
Ah
 
 
1 hour later…
marmot
marmot
2:28 PM
@FaheemMitha I am a Mac user since about 7 years.
 
 
1 hour later…
Faheem Mitha
Faheem Mitha
3:39 PM
@marmot Which makes that Penguin logo a bit confusing.
 
marmot
marmot
4:10 PM
@FaheemMitha You may have an avatar with a star in it without driving a Mercedes;-)
 
Faheem Mitha
Faheem Mitha
@marmot In a computing context, penguins normally mean one thing.
Since the 1990s, at least.
Blame Linus.
 
Alan Munn
Alan Munn
@FaheemMitha Marmot went to DeveloperWeek and picked up a free T-shirt.
 
Faheem Mitha
Faheem Mitha
@AlanMunn I see.
And put it on his avatar.
 
marmot
marmot
4:33 PM
@FaheemMitha Well, I have been using Linux for decades but the requirement of coping with file formats of a certain type that is not to be named (which administration people ten to send you) and being able to access internet in some remote places made me switch. I am still mainly using the terminal, which is almost Linux.
 
Faheem Mitha
Faheem Mitha
4:45 PM
@marmot Personally I would have gone down fighting. But I'm ornery.
 
touhami
touhami
5:18 PM
Hello @DavidCarlisle and hello @all
@DavidCarlisle back to the question of yesterday, is the full truth what said in page 286 in description of \par primitive?
 
David Carlisle
David Carlisle
@touhami for some definition of "full" and "truth". the texbook, is never that clear about whether it is talking about the tokens or primitives. so unlike a blank line which inserts a \par token, the end of paragraph does the equivalent of the primitive \unskip\penalty10000\hskip\parfillskip even if you define those tokens to do something else. Also "exercises the page builder" is a magic phrase that implies a whole pile of processing happens at that point.
 
touhami
touhami
5:34 PM
@DavidCarlisle Ok thanks for that. I have a problem in understanding the timing. I want just to get things clear
 
David Carlisle
David Carlisle
@touhami it's the "exercises the page builder" bit that is the phrase that means the asynchronous output routine/page building code gets run, apart from that everything happens in a simple order
 
touhami
touhami
in the last paragraph of page 85 it seems like that lines and glus are add to the main VL in horizontal mode but in 286 lines are add to the main VL in vertical mode.
I found 286 more in order with the first description of page 85.
 
David Carlisle
David Carlisle
@touhami in horizontal mode things are being added to a horizontal list stacked side by side, there are no lines or vertical space. as hoiz mode is executed the horixontal list is broken in to lines and added in vertical mode to the mvl
 
touhami
touhami
@DavidCarlisle thnak you again
@DavidCarlisle what you said is what i know until reading paragraph of page 85 it confuse me some way
@DavidCarlisle but with page 286 things seem more clear.
@DavidCarlisle only one problem! in page 286 TeX backs to VM after lines breaking not before isn't?
 
marmot
marmot
5:52 PM
@FaheemMitha Well, I fought a long time, until a very special marmot convinced me. ;-)
 
David Carlisle
David Carlisle
6:31 PM
@touhami the actual line breaking isn't really in h or v mode. first, in h mode a single horizontal list is constructed for the paragraph, then that list is broken into lines, then vertical mode is entered and the resulting lines and baselineskip glue etc are added to the current vertical list.
 
touhami
touhami
@DavidCarlisle ok thank you
@DavidCarlisle hope that this will not confuse me even more :-)
@DavidCarlisle is the same true for page breaking and VM?
 
David Carlisle
David Carlisle
@touhami "mode" is all about stacking newly inserted boxes, that is whether \hbox{a}\hbox{b} makes ab or a over b. linebreaking isn't about inserting new content at all, it is an internal process that takes a horizontal list (that was constructed in h mode) and adds resulting lines in v mode
@touhami well similar except the timing is different as there isn't really a "page stacking" mode that is orthogonal to veryical mode in the way that vertical mode is orthogonal to h mode.
 
barbara beeton
barbara beeton
@manooooh -- I consider "if" to be text, not an operator.
 
David Carlisle
David Carlisle
Pancake day (traditional English ones with lemon and sugar, not the crispy duck variety....)
 
Ulrike Fischer
Ulrike Fischer
7:09 PM
@DavidCarlisle is that remark a complain or a sigh of relief?
@JosephWright @AlexG wants to patch hyperref links too: tex.stackexchange.com/a/477884/2388 ;-)
 
yo'
yo'
@barbarabeeton @manooooh well, you would never see a literal "if" or "and" in a Czech math text, but you would see literal "sin" or "log". Definitely "if" is a textual thingy.
Btw, Happy Carnival!
3
 
touhami
touhami
7:33 PM
@DavidCarlisle again thnak you sir :-)
 
David Carlisle
David Carlisle
@UlrikeFischer ping him towards the new repository...
 
whatisit
whatisit
Hi All, I am wondering if I can pick everyone's brain for some help. I updated MiKTeX yesterday, including some packages. My thesis requires 'morewrites' for several reasons (but mainly pythontex). However, after the update, I am getting error messages in the (application log) miktex-xelatex.log: "FATAL miktex-xelatex - Too many open files: path="(local path)_build_files\thesis.mw"" Morewrites package was not updated - but the l3build, l3experimental, and l3kernel packages were.
Document compiled immediately prior to the update - no changes were made following the update. However, I cannot get a MWE because when I comment out enough \include files, the problem disappears.
I was wondering if anyone can help me figure out how to narrow down this issue? thanks in advance!
 
Ulrike Fischer
Ulrike Fischer
8:08 PM
@whatisit well without code it is rather difficult to debug things like that.
 
David Carlisle
David Carlisle
@whatisit too many open files isn't a tex error (of the sort addressed by morewrites) it is (I would guess) a lower level system file handle error causing miktex to abort. I don't use miktex though so can't really help.
 
whatisit
whatisit
@UlrikeFischer yes, the problem I have is I cannot produce anything near a MWE...so I haven't been able to post the question on SX.
Commenting out random include files removes the problem (but also content that I need). There is no rhyme or reason why one file or another causes the issue - that I can identify yet.
@DavidCarlisle the document compiles on a first run. On the second, when getting references linked, it fails. Might this help focus my search somehow?
 
David Carlisle
David Carlisle
@whatisit this is windows miktex?
 
whatisit
whatisit
@DavidCarlisle yes
 
David Carlisle
David Carlisle
@whatisit somebody called @UlrikeFischer seemed to comment on a similar error here sourceforge.net/p/miktex/bugs/2499 a long while ago
@UlrikeFischer it seems to be tagging season:
Thank you very mutch for your reply,but i am not understoode,why i can't use callback.register,because in document (luatex.org/svn/trunk/manual/luatex.pdf) it only exists function callback.register to add callback. Why i couldn't nothing print in this callback. It's very important for me print something. For example when i want to do tagged pdf and use your package which name tagpdf,close all opened tags at the end of the page and open it at the start of the next page. Where i can find newest documentation about all tables,attributes and functions,which use in lua only for lualatex? — Aleksandr Kozlovskiy 1 hour ago
 
Ulrike Fischer
Ulrike Fischer
8:20 PM
@DavidCarlisle oh god, that all in one comment.
@whatisit sorry but debugging a morewrites problem without any clue what is in your document, which files it opens, and without test document is quite impossible.
 
whatisit
whatisit
@DavidCarlisle that does not seem to be the issue. I don't see errors in initexmf.log since January. The error message is similar, but I'm only receiving the message during compilations, not using the application itself.
@UlrikeFischer Yes, I completely understand. I'm not sure how I am supposed to accomplish this, unfortunately. Even uploading it to, say github, it's got a lot of files. Main file includes about 10 files, many of those input many more files. That's why I'm having difficulty
 
Ulrike Fischer
Ulrike Fischer
8:37 PM
@DavidCarlisle but reading more carefully, I would say he has a rather naive idea about "closing tags". And why do he thinks, that he needs to close tags at all? tagpdf does it.
 
 
2 hours later…
Ulrike Fischer
Ulrike Fischer
10:30 PM
@AlanMunn can you invent for me a short title for a linguistic talk? The author will be Riley Sloth Ph.D.
 
Alan Munn
Alan Munn
10:40 PM
@UlrikeFischer How about Semantics and pragmatics of marmot and duck calls
2
 
Ulrike Fischer
Ulrike Fischer
@AlanMunn perfect ;-).
 

« first day (3047 days earlier)      last day (1871 days later) »