TeX, LaTeX and Friends - 2015-05-28 (page 2 of 2)

Paulo Cereda

1:14 PM

@DavidCarlisle :)

David Carlisle

@PauloCereda @barbarabeeton is probably expert with TECO commands

barbara beeton

@DavidCarlisle -- nah, but i used to be pretty good at repairing badly punched cards and mangled paper tape.

Sean Allred

@DavidCarlisle she should feel right at home in emacs then :)

@barbarabeeton Oh please do: that's a very good example of why history is useful. It's been a few hours and a full night's sleep, but I think the point I was trying to make is that VCS exists because we make mistakes. Having the history is certainly invaluable, but having them burned into the history... that's a different story.

I don't want an audit log of all the ssh keys I've used :)

Paulo Cereda

@SeanAllred I use VCS because I wanted to look cool to other kids. :)

David Carlisle

@SeanAllred correcting mistakes is part of the reason but in a multi-user project, merging workflows is probably the main use. We almost never revert but use svn all the time with very small granularity (typically any change of more than a line or two, commit) possibly main use of the history as such is then svn blame (and svn log of the highlighted revision) to see who is responsible for any bit of code.

Sean Allred

1:25 PM

@PauloCereda same

David Carlisle

@PauloCereda failed

Sean Allred

@DavidCarlisle so mean

David Carlisle

@SeanAllred again you say that:(

Sean Allred

@DavidCarlisle out of love

David Carlisle

@SeanAllred so you say

Sean Allred

1:27 PM

@DavidCarlisle To be absolutely clear: I'm a huge fan of keeping detailed history for just those reasons you describe. It's invaluable to know who to talk to about a certain change and, if you have the commit message, that will be invaluable to the person who actually made that change. (Well, if the commit message is used well.)

I think my main point was against mercurial's difficult flow for correcting mistakes that should simply be wiped out of existence. mercurial sets you on a path and expects you to follow it -- it's far too restrictive for my tastes.

Paulo Cereda

3

David Carlisle

@SeanAllred svn doesn't allow history to be edited and can't say I've ever missed that feature

Joseph Wright

@DavidCarlisle :-)

David Carlisle

@PauloCereda has discovered lego, it seems.

Sean Allred

@DavidCarlisle @PauloCereda a legogogogogogogogogogogogo

@DavidCarlisle You would if you'd checked-in a password

(or any other sensitive information)

David Carlisle

1:31 PM

@SeanAllred but I don't make mistakes, that isn't why I need a VCS:-)

Sean Allred

@DavidCarlisle And we've come full circle :)

David Carlisle

@SeanAllred I know...

Sean Allred

@DavidCarlisle I know you know

Joseph Wright

@SeanAllred Too late: change the password :-)

Sean Allred

@JosephWright Sometimes, that isn't possible. :/ think more along the lines of 'I just checked-in the latlongs of all stations in the navy'.

David Carlisle

1:33 PM

@SeanAllred it used to be of course you could just edit the rcs or cvs ,v file and falsify history but svn got all modern and obfuscated things as a binary blob in a database which means it's harder to avoid using the documented interface.

Joseph Wright

@SeanAllred Still too late (the data is in a file without being hashed, so is at least available in potentia)

Sean Allred

@JosephWright the data must be in a file for the software to work, but must not be accessible via network

Joseph Wright

@SeanAllred Your password should never be written anywhere in a non-hashed form

Sean Allred

@DavidCarlisle keeping the common man in company chains

David Carlisle

@SeanAllred but in that case presumably you repository isn't publicly accessible either, so what's the problem?

Sean Allred

1:35 PM

@JosephWright token-based auth wouldn't work here

@DavidCarlisle In the U.S., there are many different levels of security clearance (as I'm sure -- actually, I know -- is the case in the British government). Just because the repository is private doesn't mean there aren't people working on the project who don't have clearance to see the data.

Granted, that's a complicated situation, but it's the one that is drowning out other examples in my head. I'd rather have it and not need it than need it and not have it. (Part of the reason I enjoy emacs.)

David Carlisle

@SeanAllred yes but having a repro with wider access than the code checked in sounds wrong, you shouldn't have to worry about checking code into the source control is leaking data. Just relying on humans to remember not to check in sensitive files sounds dangerous.

Sean Allred

@DavidCarlisle It is dangerous.

Joseph Wright

@SeanAllred :-)

David Carlisle

maybe you should be using OS/370, to keep people alert to such dangers: "OS/370 is a truly remarkable operating system. It's possible to destroy days of work with a single misplaced space, so alertness in the programming staff is encouraged."

@JosephWright did you see you'd broken eledmac with etoolbox change...

Joseph Wright

@DavidCarlisle I saw the message: as usual, it's a case of 'if package X requires package Y it should load it explicitly, not rely on package Z to do it'

David Carlisle

1:45 PM

@JosephWright eledmac is odd it loads all the packages it uses twice, and does \reserveinserts{32} twice, I'm sure no one really needs 64 independent streams of footnotes do they?

Joseph Wright

@DavidCarlisle Looks like a DocStrip mess-up

David Carlisle

@JosephWright that was my guess, but just looking at the dtx there are no fancy docstrip guards, it just looks like cut and paste error

1010011010

2:16 PM

@PauloCereda I'm surprised it didn't build a duck out of lego. The gif still delivered.

Paulo Cereda

@1010011010 A duck would be awesome. :)

4

@1010011010 <3

:-)

2:43 PM

@norbert -- please take a look at tex.stackexchange.com/q/247440 ; i don't know of any resource that would support what is requested, but maybe you could find out.

Arthur Reutenauer

3:06 PM

@barbarabeeton What you need is a dictionary, but to add the transcriptions in a meaningful would require to have real natural processing tools.

@barbarabeeton It’s likely to be rather complex.

barbara beeton

@ArthurReutenauer -- that much i know, or can figure out. (i can find things in a japanese or chinese dictionary ordered by radical, and am familiar with many of the conventions of japanese typography, but can't read it, and don't know what resources might have been developed for the purpose requested. hence my attempt to direct the query to "someone on the ground".)

@ArthurReutenauer -- "rather" may be an understatement.

Arthur Reutenauer

@barbarabeeton I would suspect so :-) That being said, such tools must already exist. You could maybe run them as external processes and send the text back to TeX.

@barbarabeeton That could be an interesting project.

barbara beeton

@ArthurReutenauer -- good point. development of such tools may have been undertaken within the japanese tex community, which i know is quite active, if not formally organized. nothing like this was mentioned at the tug meeting in tokyo (2013; tugboat proceedings issue), although there were some fascinating presentations on the traditional top-to-bottom, right-to-left format.

David Carlisle

3:22 PM

@ArthurReutenauer @barbarabeeton see furiganizer.com no idea how accurate it is but it adds some characters over some other characters

Arthur Reutenauer

@DavidCarlisle Thanks, I was actually already typing my answer. The bottom line is that it needs serious work anyway.

David Carlisle

@ArthurReutenauer yes I suppose it depends on the level, if you are trying to tokenize arbitrary Japanese text then it's presumably a lot harder than if you are just learning japanese and want help with picking up the vocabulary of a 5 year old native reader..

Arthur Reutenauer

@DavidCarlisle Yes, of course. If you have a good learner’s dictionary you can do stupid regexp-based matching on it, it would probably gives good results.

Faheem Mitha

@barbarabeeton I was specifically alluding to leading and trailing whitespace being treated as part of a label. And that still strikes me as odd.

In general, good programming practice leans heavily towards considering whitespace as not syntactically significant. Of course, TeX isn't the average programming language.

Joseph Wright

@FaheemMitha True for programming but not for typesetting, and that's rather the point

@FaheemMitha expl3 happily ignores whitespace :-)

yo'

3:35 PM

@JosephWright indeed.

Faheem Mitha

@JosephWright True, typesetting is different.

@JosephWright Good for expl3.

Joseph Wright

Open LaTeX3 question is whether we should add 'trim spaces' to a lot of interface cases: \label would be just one of many possibles

barbara beeton

@FaheemMitha -- i've run into problems coding in html, where spaces can suddenly "disappear" without warning, turning my carefully crafted prose into nonsense. working with actual text is different from coding in a highly structured programming language. just learn the necessary rules, and be careful when using them.

Arthur Reutenauer

Python, anyone? ;-)

yo'

@ArthurReutenauer yes, Stack Overflow :-)

barbara beeton

3:45 PM

@JosephWright -- probably not relevant, but the ams document classes define several control sequences with a space in the middle or at the end, using \csname ... \endcsname (haven't found any with a space at the beginning, and it wouldn't be easy to create or use one such); obviously, these are meant for internal use only, but please take this into consideration before deciding what to trim or collapse.

Faheem Mitha

@barbarabeeton I thought one writes in something more high level and then converts to html.

@JosephWright David said you shouldn't do it, because TeX doesn't. Link available on request.

To be clear, I was only talking about leading and trailing whitespace in labels.

Which technically aren't text, anyway.

David Carlisle

@FaheemMitha or rather I said that you shouldn't automatically expect white space to be insignificant as that isn't the default in TeX.

Faheem Mitha

I don't know if you saw the postings yesterday on this topic.

David Carlisle

@barbarabeeton latex defines hundreds of commands with spaces at the end of their names, and one command with dozens of spaces in its name.

Faheem Mitha

@DavidCarlisle Well, Ok. But I thought the take home message was - don't strip leading/trailing whiltespace from labels. Cos it's contrary to the Zen of TeX. Or something.

Arthur Reutenauer

3:50 PM

@yo' Didn’t get that one

Faheem Mitha

@DavidCarlisle On purpose?

David Carlisle

@FaheemMitha yes

Faheem Mitha

@DavidCarlisle One has got to wonder why.

Some alignment issues?

David Carlisle

@FaheemMitha in the specific case of \ref in latex2e I think you shouldn't change the behaviour as part of an unrelated change collecting the list of used references, but that isn't the same thing as saying a new system shouldn't do some things different;y

Faheem Mitha

@DavidCarlisle ok

The new system being LaTeX 3?

barbara beeton

3:54 PM

@FaheemMitha -- depends on what one is doing. if one is trying to present tex code on a web page to cut-and-paste, there isn't any decent, reliable and economical gui that i am aware of.

David Carlisle

@FaheemMitha the one with many spaces (60 something if I remember correctly) is used when outputting latex errror messages to trick tex into hiding some internals. Any command defined with \DeclareRobustCommand uses a command ending with a space as the internal command to which \protect is applied. This means that when written to a file it looks like the original.

@FaheemMitha perhaps, or some new set of commands defined by a latex2e package, but not an existing core command like \ref.

Faheem Mitha

@barbarabeeton Not gui. Higher level language. Like markdown. Maybe there are better options? Docbook?

@DavidCarlisle You mean the version without the trailing space is the "real" command?

@DavidCarlisle ok

David Carlisle

@FaheemMitha the version with the trailing space is the real command.

@FaheemMitha try \show\vspace you will see that command doesn't add any space it is just \protect\vspace_ where \vspace_ (with space not _) has the real definition.

Faheem Mitha

@DavidCarlisle oh

barbara beeton

@FaheemMitha -- one has to live within the "tools" prescribed by one's employer. and nothing higher level is provided. (and even the one provided corrupts the input code the first time through, and always has to be revised to get the correct result.)

Faheem Mitha

4:01 PM

@DavidCarlisle I feel my usual TeX-induced migraine coming on.

@barbarabeeton tell your employer to prescribe some additional tools?

barbara beeton

@FaheemMitha -- believe me, i've tried. nobody else needs them, so "the demand isn't sufficient to justify the time or expenditure."

Faheem Mitha

@barbarabeeton Oh. Sorry to hear that. Just curious - what would you use if you could?

Joseph Wright

@barbarabeeton As I said, would have to be discussed. There is a strong argument for leaving this alone, as if people are daft enough to write \begin{ foo } they get what they deserve!

barbara beeton

@FaheemMitha -- well, i learned html in the first place when download time was critical, and i have never found a higher-level interface that could produce code as tight as what is hand-tuned. for non-work tasks, i often have to edit out portions of automatically-produced html email, and i am appalled at the garbage spewed out needlessly; i can often cut a message down to 30% or less than the original, retaining exactly the same appearance. so i am not encouraged to search further.

Faheem Mitha

@barbarabeeton Ok, fair enough.

@JosephWright Sometimes people like to add whitespace for readability, in the perhaps naive expectation that it won't change anything. In virtually any programming language I can think of, it wouldn't.

Joseph Wright

4:11 PM

@FaheemMitha But it would inside a string, and that's more-or-less what TeX arguments are (as they are for typesetting)

barbara beeton

@FaheemMitha -- since i need to protect tex code in web postings against translation, i need to include specifiers like "class='notranslate' for code blocks. there may be an interface that does that, but again, that's unusual enough that the time spent looking would be more than the time spent writing out the html by hand.

Faheem Mitha

@JosephWright even leading/trailing whitespace?

@barbarabeeton oh. And all this is for the AMS?

barbara beeton

@JosephWright -- "daft" is a good description for many authors.

Joseph Wright

@FaheemMitha I'd suggest "foo" and " foo " would not be equal in most languages

Lua example

if "foo" == " foo " then
  print("TRUE")
else
  print("FALSE")
end

Faheem Mitha

@JosephWright True, but that was not the comparison I had in mind. I submit that a better comparison would be function arguments.

Joseph Wright

4:16 PM

@FaheemMitha It depends on the context, hence saying this is an open question for LaTeX3. One can certainly argue that e.g. environment names are 'programmatic' so should trim spaces, whereas say section names are 'text' so should presumably not.

Arthur Reutenauer

@FaheemMitha But what if these arguments are string with spaces inside them? ;-)

barbara beeton

@FaheemMitha -- yes. if you want examples, look at the "ams author faq".

Faheem Mitha

@barbarabeeton ok

@JosephWright True, some of this might be context-dependent.

Is AMS LaTeX development still active?

Joseph Wright

@FaheemMitha What is different in TeX from other languages is that we don't have 'strings' so there is no equivalent of the need to add say " around input that is 'text'. For example, in Lua we might have myfunc ( "foo" ) where the spaces are around the argument but not the text. The same is not true in TeX.

Faheem Mitha

@JosephWright yes, I see.

David Carlisle

4:23 PM

@FaheemMitha yes " a " has length three in any language I know. Also of course white space is significant in some languages, notably python, but also fortran 77 or some styles of F# or ....

Faheem Mitha

@DavidCarlisle True, in Python whitespace is syntactically significant. Unfortunately.

Just to be clear, I didn't intend to waste people's time with my shocking whitespace preferences.

barbara beeton

@FaheemMitha -- updates to amsmath and the ams document classes are "on hold" for lack of resources, and management decision that other projects are more urgent.

Faheem Mitha

@barbarabeeton I see. Thanks.

1010011010

4:56 PM

Hey, do I have any hope of showing the contents of a \box in the terminal or something?

Nasser

What is the trick to put node at end of line, on same path? if I use [right] it does not stay on same slope. This is MWE

\documentclass[12pt]{article}
\usepackage{tikz}
\begin{document}
\begin{tikzpicture}
\draw[->] (0,0) -- (30:2) node [right] {$v$};
\end{tikzpicture}
\end{document}

I want the label $v$ to put at end of the arrow, on same slope. Using [right] seems to not use same slope?

I need just to put small space, but keep the label at same slope.

Paul Gessler

5:10 PM

@Nasser quick way is node [pos=1.1] or adjust pos value to your liking. Unfortunately the position then depends on the total length of the arrow.

Nasser

@PaulGessler thanks,. this is much better:

1010011010

Alternatively, reboxing from lrbox to lower level commands... how feasible is it? Otherwise I'd somehow have to set a \vbox inside an environment, don't think that's easier ?

Sean Allred

@JosephWright @DavidCarlisle Major takeaways from the GraphicDesign question I posted: (1) use of color can be cost-prohibitive; (2) color can be used incorrectly (clashing colors); (3) the concept of a color scheme is prevalent; (4) color is used both in canvas and stroke; (5) color is most often a realization of branding (related to (3)). Of these, I think (1), (3), and (4) are particularly relevant to l3color.

Certainly something that should be implemented in the kernel is the idea of a color pallette -- likely implemented as a prop

4

A: How is color used in document design?

I do a lot of typesetting with LaTeX and use colour quite extensively in most of my documents. For my purposes, I use it in the following ways: Branding: This is probably the most common reason that I use colour, as frequently the documents I am writing are tied very specifically to an organis...

I swear it's pure chance that the answerer uses LaTeX; I'm not sure if that leaves open questions about his experience or if it makes the information even more relevant.

David Carlisle

5:27 PM

@1010011010 \showbox

Sean Allred

5:41 PM

@1010011010 Be careful though: the contents can be more detailed than you bargained for.

David Carlisle

@SeanAllred don't you like reading \showbox output? I guess I have spent longer doing that than reading pdf output from tex

Sean Allred

@DavidCarlisle it's certainly lovely and can be very engaging reading, perhaps I just need a breather every now and then :)

David Carlisle

@SeanAllred \showbox/\showputput is fairly straight forward, following \tracingparagraphs logging is more of a challenge.

egreg

but how can you do without miktex? what do you use? — user39158 10 mins ago

^^^^ :-)

David Carlisle

5:57 PM

@egreg Word?

egreg

@DavidCarlisle Yeah, my well kept secret!

3

1010011010

@SeanAllred \showbox\mp@topfigbox

! OK (see the transcript file).

Errrrrrrrrr... yeah

(The log file doesn't *show anything.)

David Carlisle

@1010011010 add \tracingonline=1 and \showboxdepth=\maxdimen and \showboxbreadth=\maxdimen

@1010011010 or more simply \showoutput which does all of that and a bit more

@egreg but if we all star it it won't be a secret anymore

1010011010

@DavidCarlisle Thanks... this does confuse me even more though....

Time for a question for the main site (due for tonight somewhere) since I couldn't find any information on the main site on this yet

egreg

@DavidCarlisle I might make a shocking revelation at Darmstadt.

David Carlisle

6:03 PM

@1010011010 what the output from \showbox ? (or the content of your particular box)

Sean Allred

> Hi, my name is Sean Allred, and I'm a chronic Word user.

egreg

@DavidCarlisle I could prepare my slides with PP.

David Carlisle

@egreg I tried that once (was a requirement of a workshop I attended). Never again:(

egreg

@DavidCarlisle I should say that I made a couple of presentations with Keynote (not about math or TeX, of course). Possibly easier than PP, but not something I'd use regularly.

Sean Allred

I don't think I've fully appreciated hyperref until this day: 728 page training companion; no PDF bookmarks.

This is hell.

Arthur Reutenauer

6:29 PM

@egreg ”How can you do without MiKTeX?” is a funny quote :-)

Paulo Cereda

6:46 PM

@egreg ooooh tell us moaaaaarrrrr!

1010011010

7:00 PM

@DavidCarlisle I just wanted to test it for a simple box across the document...

David Carlisle

@1010011010 ?

Paulo Cereda

@egreg Keynote? I used it as well a couple of times. :)

1010011010

@DavidCarlisle Yeah the problem really boils down to: how to capture the argument of an environment so it still works inside primitive box commands... lrbox won't cut it in that case

It's not typesetting anything or it's giving me errors, so I'm trying to trace what contents the box holds at different stages of the typesetting

David Carlisle

@1010011010 lrbox is just \setbox#1\hbox{...}`

1010011010

@DavidCarlisle So how would I go about coding that in an environment setting? I couldn't get \newenvironment{myenv}{\setbox#1\hbox\bgroup}{\egroup} to work

David Carlisle

7:05 PM

@1010011010 no, that's what lrbox is for

\newenvironment{myenv}[1]{\begin{lrbox}{#1}{\end{lrbox}....\usebox{#1}}

1010011010

@DavidCarlisle So the unhboxed contents of lrbox are still usable inside, say, a \vbox, so that it can still be vsplit?

Because precisely that wouldn't work

David Carlisle

@1010011010 lrbox is hmode (LRmode) like \mbox so if you want vertical mode material in there you need to nest a minipage, same as you would have to in mbox

1010011010

@DavidCarlisle I have a minipage inside the lrbox... that's what you mean right?

David Carlisle

@1010011010 or just use \setbox#1\vbox\bgroup directly, which is often simpler

@1010011010 yes

barbara beeton

@PauloCereda -- deadduckday.com

1010011010

7:11 PM

Okay basically I have two boxes, the basic idea is \newenvironment{myenv}{\begin{lrbox}{figurebox}\begin{minipage}}{\end{minipage}\‌end{lrbox}\begin{lrbox}{captionbox}\begin{minipage}<caption>\end{minipage}\end{lr‌box}\setbox\mybox\vbox{\unhbox\figurebox\par\unhbox\captionbox}} but apparently there's something wrong with the box contents :( since it tells me I'm doing illegal stuff

David Carlisle

@1010011010 you need minipage [t] or [b] not the default [c] as otherwise its in math mode and you can not remove the mathoff node.

Paulo Cereda

@barbarabeeton o.O

1010011010

@DavidCarlisle Math mode? Wow.... how does that even work

David Carlisle

@1010011010 the default vertical centering of minipage/parbox is done by $\vcenter{...}$

@1010011010 can't work out what your caption comes from, is that an argument to the enviornment?

1010011010

@DavidCarlisle Basically I redefine caption, store it in a macro and put it into an lrbox later on, so that I can just use \caption inside the environment

Just think of it as another box added to the list of elements inside mybox

I just wanted a random example which unhboxed two boxes

David Carlisle

7:16 PM

@1010011010 sorry there are too many missing pieces to guess what you are doing

1010011010

@DavidCarlisle As I said, question is in the making :-)

Will be much clearer without all the optional background info

yo'

@ArthurReutenauer Well, most people who say things like "Is there anybody using X" want to ask a question about X. I just wanted to point out that this chat is not a helpdesk :)

Arthur Reutenauer

@yo' Oh, you mean when I said “Python, anyone?” :-) Well, that’s not what I meant, but you probably guessed it.

David Carlisle

@1010011010 vsplitting the contents of lrbox....

\documentclass{article}

\newenvironment{foo}
{\centering\begin{lrbox}{0}\begin{minipage}[t]{\textwidth}}
{\end{minipage}\end{lrbox}%
\setbox0\hbox{\unhbox0\global\setbox1\lastbox}%
\setbox0\vsplit1 to \baselineskip
\fbox{\usebox0}\par\fbox{\usebox1}}


\begin{document}


\begin{foo}
aaa

bb

cc

dd
\end{foo}



\end{document}

1010011010

@DavidCarlisle Thanks for now, it looks promising but I have to look at this in an hour or so...

When I'm back home...

yo'

7:29 PM

@ArthurReutenauer yep I did guess it, since that's what many people ask here

Arthur Reutenauer

@yo' The conversation topic was the meaning of spaces in programming languages, and I wanted to point out that Python was an example of a language where spacing was very significant - it was mentioned again later in the discussion.

yo'

@ArthurReutenauer ah ok. sorry for my prejudice then

Arthur Reutenauer

@yo' It’s OK.

egreg

7:59 PM

@PauloCereda How could it be a shocking revelation if I told it now?

yo'

@egreg Tell only @Paulo. He'd never spread gossip. :)

egreg

8:19 PM

@yo' I know! ;-)

Paulo Cereda

@egreg Because you might need a backup. :)

@yo' yaaay I'm trustful. :)

egreg

@PauloCereda You're even a trusted user on TeX.SX!

Paulo Cereda

@egreg ooh it's truth! :)

yo'

@PauloCereda at least something in this conversation is :D

btw, moving again tomorrow, only to move again on Sunday.

1010011010

9:41 PM

@DavidCarlisle All right I'm here. So why the\global\setbox1\lastbox again?

Is it just about duplicating box 0?

David Carlisle

@1010011010 to get the vbox out of the hbox

1010011010

@DavidCarlisle because \lastbox is automatically a vbox?

David Carlisle

@1010011010 no \lastbox is whatever the last box in the list is (or void) but you know in advance that the hbox saved by lrbox just contains a single vbox (from the minipage)

@cfr one for you:-) tex.stackexchange.com/questions/247530/…

evinda

10:11 PM

Hello :)

David Carlisle

@evinda hi

evinda

I want to create a list so that the students write their name, number and their signature.
What kind of template should I use?

David Carlisle

@evinda just a table (tabular) with empty cells?

Nasser

10:27 PM

A Wikipedia article on writing math contradicts itself. See

Wikipedia:Manual of Style/Mathematics

This subpage of the Manual of Style contains guidelines for writing and editing clear, encyclopedic, attractive, and interesting articles on mathematics and for the use of mathematical notation in Wikipedia articles on other subjects. For matters of style not treated on this subpage, follow the main Manual of Style and its other subpages to achieve consistency of style throughout Wikipedia. == Suggested structure == Probably the hardest part of writing a mathematical article (actually, any article) is the difficulty of addressing the level of mathematical knowledge on the part of the reader. For...

I got these from the same page. At the top, it make definition that starts with symbol, but below it says not to start something with a symbol. Very confusing. It looks like who ever wrote the first part of the page is different person from the one who wrote the rest of the page

David Carlisle

@Nasser I can't see any contradiction in the two circled phrases

Nasser

@DavidCarlisle the definition starts with symbol (Let S and T be...) but below it says not to do that. ("sentences should not begin with a symbol") ?

David Carlisle

@Nasser "Let" is a word not a symbol.

evinda

@DavidCarlisle I wrote this:

\documentclass{article}
\usepackage[utf8]{inputenc}
\usepackage[margin = 3cm]{geometry}
\usepackage{booktabs}
\usepackage{siunitx}
\usepackage{mathdots}
\usepackage{polyglossia}
\setmainlanguage{greek}
\setotherlanguage{english}
\usepackage{libertine}

\begin{document}

\label{table:something}
\begin{tabular}{
>{$}c<{$}
*{2}{S[table-format = 5]}
*{2}{S[table-format = 2]}
*{2}{S[table-format = 3]}
}
\toprule
Όνομα & & {Αριθμός} & & {Τ.} & & {Υπ.} \\

@DavidCarlisle but I get an error...

David Carlisle

@evinda I get no error with xelatex on your example, just a (correct) warning that you should not use inputenc..

@Nasser It means literally not to start a sentence with a symbol. use Let $T$ be... rather than $T$ is ...

evinda

10:41 PM

@DavidCarlisle I get: Compile Error. Sorry, your LaTeX code couldn't compile for some reason. Please check the errors below for details, or view the raw log.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!
! Fatal fontspec error: "cannot-use-pdftex"
!
! The fontspec package requires either XeTeX or LuaTeX to function.
!
! You must change your typesetting engine to, e.g., "xelatex" or "lualatex"
! instead of plain "latex" or "pdflatex".
!
! See the fontspec documentation for further information.

@DavidCarlisle /usr/local/texlive/2014/texmf-dist/tex/latex/fontspec/fontspec.sty, line 43

@DavidCarlisle At my code there is no line 43

David Carlisle

Fatal fontspec error: "cannot-use-pdftex" that is not an errror with ypur document but with your command, as it says, you can not use pdflatex, you need xelatex or lualatex with fontspec

evinda

@DavidCarlisle What can I do? :/

@DavidCarlisle I use sharelatex

David Carlisle

@evinda use xelatex instead of pdflatex

evinda

@DavidCarlisle So do I have to use a specific usepackage?

David Carlisle

@evinda sharelatex includes xelatex and lualatex as far as I know (but Ive never used it)

@evinda no it is nothing to do with the document it is the tex engine processing the document that you need to change (or don't use fontspec)

evinda

10:45 PM

With which command do I use fontspec? @DavidCarlisle

Arthur Reutenauer

@barbarabeeton Great that you asked meta.tex.stackexchange.com/questions/6181/… Obviously I don’t have a good answer, but it’s worth thinking about these issues.

David Carlisle

@evinda look who just arrived:-) it's polyglossia (@ArthurReutenauer) this version works with pdftex, but you should be able to use a unicode tex with sharelatex I think

\documentclass{article}
\usepackage[LGR]{fontenc}
\usepackage[utf8]{inputenc}
\usepackage[margin = 3cm]{geometry}
\usepackage{booktabs}
\usepackage{siunitx}
\usepackage{mathdots}
%\usepackage{polyglossia}
%\setmainlanguage{greek}
%\setotherlanguage{english}
\usepackage{libertine}

\begin{document}

\label{table:something}
\begin{tabular}{
>{$}c<{$}
*{2}{S[table-format = 5]}
*{2}{S[table-format = 2]}
*{2}{S[table-format = 3]}
}
\toprule
Όνομα & & {Αριθμός} & & {Τ.} & & {Υπ.} \\
\midrule
& & & & & & \\[1ex]

evinda

Why does Όνομα appear with english letters? @DavidCarlisle

David Carlisle

@evinda I didn't look at the output:-), hang on....

@evinda because you put it in math mode (and textual greek doesn't work there) try this:

\documentclass{article}
\usepackage[greek]{babel}
\usepackage[utf8]{inputenc}
\usepackage[margin = 3cm]{geometry}
\usepackage{booktabs}
\usepackage{siunitx}
\usepackage{mathdots}
%\usepackage{polyglossia}
%\setmainlanguage{greek}
%\setotherlanguage{english}
\usepackage{libertine}

\begin{document}

\label{table:something}
\begin{tabular}{
c
*{2}{S[table-format = 5]}
*{2}{S[table-format = 2]}
*{2}{S[table-format = 3]}
}
\toprule
Όνομα & & {Αριθμός} & & {Τ.} & & {Υπ.} \\
\midrule
& & & & & & \\[1ex]

Arthur Reutenauer

@evinda You’re right, that’s odd (Όνομα rendered as 'Onoma).

David Carlisle

10:56 PM

@ArthurReutenauer LGR encoding used in math mode....

Arthur Reutenauer

@DavidCarlisle Oh, that’s fun :-)

David Carlisle

@ArthurReutenauer do you know how to use xelatex with sharelatex's online service?

Arthur Reutenauer

@DavidCarlisle No idea, sorry. I guess I should find out.

David Carlisle

@ArthurReutenauer me neither but I was just removing all traces of your work so the document ran with pdftex as you came in:-)

evinda

@DavidCarlisle Great!!! And something else... I want that there is more space between Όνομα Αριθμός and so on... Also it should cover the whole line of the page. How could I do this?

Arthur Reutenauer

10:59 PM

@DavidCarlisle I got that ;-) Less work for me!

David Carlisle

@evinda use \begin{tabular*}{\textwidth}{@{\extracolsep{\fill}c....

Arthur Reutenauer

@DavidCarlisle Reading sharelatex.com/learn/Choosing_a_LaTeX_Compiler between the lines it seems pretty clear that the answer is no.

evinda

@DavidCarlisle @ArthurReutenauer Now I get this error:
File ended while scanning use of \@@array.
<inserted text>
\par
<*> main.tex

I suspect you have forgotten a `}', causing me
to read past where you wanted me to stop.
I'll try to recover; but if the error is serious,
you'd better type `E' or `X' now and fix your file.

! Emergency stop.
<*> main.tex

*** (job aborted, no legal \end found)

Here is how much of TeX's memory you used:
12754 strings out of 493109
267416 string characters out of 6135010

What does this mean?

David Carlisle

@ArthurReutenauer. @evinda see sharelatex.com/learn/XeLaTeX

Arthur Reutenauer

@DavidCarlisle I’ll contact them, they seem pretty active and motivated.

David Carlisle

11:03 PM

@evinda it means your document has more { than it has } probably because you copied what I wrote above. @{\extracolsep{\fill} should be @{\extracolsep{\fill}}

Arthur Reutenauer

@DavidCarlisle Ah, great! Wonder why it’s not linked from the table of contents on the left.

David Carlisle

@ArthurReutenauer never try to follow the documentation:-) I just google'd for sharelatex xelatex and hit that:-)

Arthur Reutenauer

@DavidCarlisle Yes, I guessed that’s what you did :-)

Paulo Cereda

Hi @Arthur! Did you enjoy the capybara photos in the university? :)

Arthur Reutenauer

Hi @Paulo! Sorry I didn’t have time to look this morning, it was a pretty busy day.

Paulo Cereda

11:10 PM

@ArthurReutenauer Oh no worries. :) I just mentioned because they coexist in one of the biggest cities in Latin America, quite amusing!

Arthur Reutenauer

@PauloCereda Oh that’s really cute :-)

Paulo Cereda

@ArthurReutenauer :)

1010011010

11:22 PM

Can the order be controlled at which floats and footer are typeset?

Otherwise I'd need some weird construction which "fakes" the correct caption numbering... very awkward for future document accessibility

David Carlisle

11:53 PM

@1010011010 well the footer should come at the bottom (clue in the name) why have you got a caption in the footer?

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends