TeX, LaTeX and Friends

Stefan Kottwitz

00:00

a more canonical question about ! Package *** Error: Name clash, \*** is already defined. with possible solutions

it's a FAQ ... tex.ac.uk/cgi-bin/texfaq2html?label=alreadydef

Speravir

@StefanKottwitz On the other hand Heiko’s answer is a good one! (How surprising.)

Stephan Lehmke

@Speravir Though it doesn't look universal (only if the class has a good definition).

Stefan Kottwitz

@Speravir It will not go away ;-)

Well, no hurry, we always can reopen it, or even close

I just think, obvious duplicates should be closed asap before there's a xth answer clone

egreg

@StephanLehmke If the class hasn't a good definition, there's little one can do, aside changing the class. In the specific case the class author should simply say \RequirePackage{ifpdf}: easier and cleaner.

Stefan Kottwitz

NARQ or TL is less inconvinient

Should we close it as TL then, because of a specific class problem?

Speravir

00:05

@StefanKottwitz But NARQ is not right anymore (what was my intention for asking …)

Stephan Lehmke

@egreg Undefining \ifpdf after the class is loaded would activate the ifpdf definition.

egreg

@StephanLehmke This is a possible solution.

Speravir

@StefanKottwitz Oh, yes. Hopefully Vivi will not rant/vent about that.

Stefan Kottwitz

@Speravir Vivi may vote for reopening ;-)

Speravir

@StefanKottwitz I meant, she was very in rage, yesterday about closing, but not informing the OP, why. And about, in her feeling, too rude words, if I’ve got it right.

David Carlisle

00:10

@StephanLehmke tabu's too clever for it's own good:-)

Stefan Kottwitz

@Speravir searching for "name already defined" I got 5000+ matches on TeX.SE

percusse

@StefanKottwitz I've checked pgfplots and it's currently at 1.7 but texdoc shows 1.5.1 . But it might not be included in TL2012. Thanks anyway.

Stephan Lehmke

@DavidCarlisle :-) Though the column width calculation seems much less clever than tabulary at first sight (also I don't get the explanation in the documentation). So I think I'm staying with my patched tabulary for now ;-)

Gnintendo

is a less legible technique for blurring text than what is outlined here possible?:

31

Q: Blur the text so it's not readable

On a related note to How do I make my document look like it was written by an Cthulhu worshipping madman?, I want to create a LaTeX command that blurs the text so it's unreadable. Unfortunately, whatever I google to get ideas on how to do this, I find forum posts from everyone and his aunt havin...

fonts

I still find that legible ^

David Carlisle

@StephanLehmke I've never looked exactly what it does, it seems to be a variant of the tabularx algorithm but with a more sane syntax for varying the widths than just resetikng \hsize after the fact as suggested in TX

Stefan Kottwitz

00:17

@percusse Thanks, I updated that package

Stephan Lehmke

@DavidCarlisle That's correct for the positive arguments (like * notation in HTML). And indeed the negative arguments only seem an "emergency reduction" version of the positive ones, while tabulary really tries to derive the column width from the amount of material in the column.

@Gnintendo While retaining PDF format? Maybe by transforming the font into a bitmap and blurring that?

@Gnintendo Also you could set the text color to light yellow ;-)

Stefan Kottwitz

@Speravir I closed it as TL and wrote the reasons.

Speravir

@StefanKottwitz With "name clash" "already defined" site:tex.stackexchange.com I get only 13 results, and in fact some are doubled or too general: google.com/… Ooh, we came across.

Stefan Kottwitz

@Speravir The OP got help. So that part is done. Now it's about good content for the site. In the original and edited version, the question is too localized. It cold be edited and made very general, useful, thus reopened. Anybody could go ahead.

No need to deal with the name clash problem just for the JHEP3.cls class, prominently in the title.

Speravir

@StefanKottwitz Upvoted your comment. Please delete Vivi’s and Christian’s (cgnieder) comments, as they are obsolete now.

Stefan Kottwitz

00:27

@Speravir Perhaps a bit later, so that the comment posters don't feel bad.

It's kind of an open discussion still.

Could go to meta if desired.

Speravir

@StefanKottwitz Yes, you’re right! … Do you moderators have some ToDo- tool? Because some of my flags also were decided days later (and this was always useful).

Stefan Kottwitz

@Speravir Flags are like bookmarks and keep listed for mods. It can happen that it takes some time until somebody decides what to do with a flag, but all are dealt with.

Speravir

Speaking of another thing: In tex.stackexchange.com/questions/87171/… though it works, would you consider it good style to write commands with spaces inside like in \usepackage {lipsum}?

Stephan Lehmke

@Speravir This syntax is safer if { should be a letter ;-)

Speravir

00:42

@StephanLehmke What? :-) I think, I must also ask for an introduction into LaTeX …

Stephan Lehmke

@Speravir This is more or less a joke, but there are people who will put a space wherever TeX will read one...

Speravir

@StephanLehmke Aah!

Stephan Lehmke

@Speravir It's a bit like \let\foo\bar vs. \let\foo=\bar.

Speravir

@StephanLehmke Yes. But in this special case of the answer I find it (with the spaces) misleading, because the question is from a LaTeX newbie.

Stephan Lehmke

@Speravir OTOH it helps avoid some delusions about TeX syntax (in other languages it's a syntax error to put a space between function name and parenthesis).

David Carlisle

01:09

@Speravir spaces are perhaps uncommon but newlines I quite often use in that position if there is a long list of options or a long package list

@StephanLehmke you let @Bruno steal the points :(

Stephan Lehmke

@DavidCarlisle Sorry, but the explanation about \scantokens seemed convincing... You can fight back of course ;-)

Vivi

@Speravir I was not ranting about that question being closed, but rather about the lack of comments from those voting to close and about the comments made by cgnieder. I never said anything about being against that question being closed (even though I probably lean against not closing it, but I don't feel strongly about it. But leave a comment so the OP knows what's happening!).

@Speravir @StephanLehmke I don't mind if you delete the comments since I talked to cgnieder and I think we understand each other now.

Speravir

@DavidCarlisle YOU know, where you are allowed. As I saw the answer I thought of a situation, I sometime ago went into an error, because I left a space, where it would not be allowed.

cmhughes

anyone have an opinion on this one being a duplicate

11

Q: Span a tikz chain over multiple pages

I wanna create a flowchart as a chain. What I want to achieve is a chain spanned over multiple pages on top of every page (see picture): For the example picture I created it manually (copied the code from texample.net, made a standalone chaine, cropped it manually in half and used eso-pic to p...

tikz-chains

kan

01:25

@Vivi Thanks for that emacs answer once again. It helped me understand a couple of things.

Stephan Lehmke

@cmhughes I'm not decided about this. The included image in the linked answer is finite by definition while it is at least theoretically possible to generate an "infinite" TikZ picture by some recursive definition. It would be intriguing to ask whether the TikZ drawing process can be restriced to only actually draw a certain "window" of the whole. Though the OP isn't really giving a lot of concrete motivation.

percusse

@StephanLehmke You can clip it with any closed path

But still it looks like if one can do 2 parts, can also do n parts with the linked question.

cmhughes

@StephanLehmke good point. it seems he wants more of a header type approach, rather than a float. In which case, it seems that the image could be created in standalone and then cropped using appropriate packages (adjustbox?)

Stephan Lehmke

@percusse Yes but won't the resolution of TeXs registers get in the way of actually drawing on a "giant" canvas?

@cmhughes Not if it is, say, 150 pages wide.

cmhughes

@StephanLehmke is it possible to create an image that big?

percusse

01:31

@StephanLehmke That's the ambiguous part. If the drawing is drawn via TikZ, one can place n-many tikzpictures by clipping each time a different part (or saving it in a box etc.). However if the image is fixed then best is to provide a trimming rectangle depending on the page count.

cmhughes

@percusse yes, that's exactly what I was thinking (and Werner mentions in his comment too)

Stephan Lehmke

@cmhughes That's for the TikZ experts to say, or rather, it looks like the real question behind your "duplicate" to me. "multiple" at least makes no statement about how many pages are meant.

Vivi

@kan No problem. I still struggle with emacs and I happened to have looked into those commands that same day (yesterday?)

kan

@Vivi Oh, I see. Lucky coincidence. (And, yes, yesterday.)

David Carlisle

@StephanLehmke It sounds convincing but I think it's wrong actually (although the comment here moaning about losing the points was just moaning, before I read Bruno's answer:-)

percusse

01:35

@cmhughes Just like the spy library does, one can box the TikZ drawing. Then place it in a path picture argument of a node, then scale the whole thing such that it matches the paper width. But that's a bit of work and I'm watching Matlab crash and burn here.

Vivi

@kan Yes... I am having trouble exactly with my latex mode highlighting. I changed the default font to be monospaced so I could take full advantage of the rectangular yank, and now all my modes have a grey background :/ I don't know how to go back to my original choice of backgrounds without undoing my choice of default font....

leo

hello

:-)

Vivi

@kan every issue I have with emacs seems to take me a few hours to solve and I am not in the mood to do that now. Just yesterday I solved another issue, so I am happy with that for the moment.

@leo Hello :)

kan

@Vivi Yeah, darn emacs and its obscure terminology.

And, do you happen to know how to start emacs so that there is only one buffer visible (especially, while loading a file using $emacs <filename>.tex. I have to do C-x-1 all the time. :(

David Carlisle

@kan that's the usual case isn't it? what other buffer do you see (not that you should start emacs that often)

Vivi

01:40

I am pretty sure mine starts with only one screen. But I remember there is a way to have the default starting set up and perhaps yours was done so that it would start this particular way?

kan

@DavidCarlisle Welcome to GNU Emacs (appears as GNU emacs buffer in the Buffer menu) is one screen that comes up at the startup below the .tex file being loaded.

Vivi

@kan See if this will help: askubuntu.com/questions/4820/…

kan

@Vivi Well, I can't be really sure, but I don't think.

Stephan Lehmke

@DavidCarlisle Ah now we're getting a real fight ;-) Is it really the case that \verbatim@processline is executed when all the catcodes are redefined? Oh well, catcodes are a real mystery to me. I just tried to make labels based on the material in \the\verbatim@line ;-)

kan

@Vivi suggests that I just manipulate the variable named inhibit-startup-screen -- let me look around...

Stephan Lehmke

01:44

@DavidCarlisle So, where are we going with this now? Apart from the all-important tick, there also is the minor issue of which fix to suggest for tabu :-)

Vivi

@kan they say you might have to change that, but you should do the other things first and see if they work. I just checked and inhibit-startup-screen for me is set on non-nil

leo

@egreg can you help me

kan

@Vivi Yes, that's it! Fixed. I went: C-h-v-inhibit-startup-screen and set it to on (non-nil) and the help says the value of the variable is t.

leo

Vivi

Great! Another way to customize a variable is using M-x customize-variable RET variable-name . Also useful is M-x customize-apropos RET searchterm

leo

01:48

I don't kow what to put in there

kan

@Vivi So, M-x seems like a synonymous with modify, am I right?

Stephan Lehmke

@Vivi Emacs is like the internet. Wherever you go, there's several ways to go on and only one way to go back ;-)

Vivi

@kan I don't think so... I think M-x is a synonym to "execute command by calling variable name" instead of using a keybinding...

leo

for example is
TeX: medium
right?

user19161

@kan How is your Ubuntu experience so far?

leo

01:50

I don't know hot to interpret the XXX, YYY, ZZZ

Vivi

@kan at least that's how I see it.

kan

@Vivi Yeah, perhaps, you're right.

Vivi

@StephanLehmke xkcd.com/378 (you probably know this already... Interestingly, M-x butterfly works! @kan have you tried it?)

@kan sorry, I meant "execute command by name" instead of using keybinding

kan

@Vivi I think I have: for the font-lock-latex-section blahblah...

Vivi

@kan example: you can open file using C-x C-f, but the same command can be reached by using its name, M-x find-file

David Carlisle

01:54

@StephanLehmke mine certainly isn't a fix it just unilaterally disables the non working code without offering any replacement (as it isn't clear what the intention was) As Bruno commented the code uses \scantokens a lot, mostly in suspicous ways:-) it only really makes sense if it is accepting unnkown user onput, there is not much point using it on a fxed string. tabu seems to do both though, although the one that bit you was scanning the fixed string \def\:{|}

Stephan Lehmke

@Vivi Well it's the other way round really. You can reach every command via M-x and some of them are also bound to keys.

leo

got it :-)

Vivi

@StephanLehmke yes, that's what I mean...

David Carlisle

@Vivi well function name not variable name, but yes

Vivi

@DavidCarlisle yes, I corrected it later but said command instead of function :)

kan

01:56

@Jasper I like Ubuntu now; but I am afraid I cluttered it; so, I'll install a new version of ubuntu, sometime next year.

David Carlisle

@kan you must have something in your startup file, you are only supposed to see the welcome to gnu emacs buffer if you start emacs with no file specified

@Vivi ah yes:-)

Stephan Lehmke

@DavidCarlisle Ok this seems really incompatible with using it inside verbatim then. One of you or @bruno should say so explicitly in their answer, so I can accept that one ;-)

And I'll stay with tabulary.

kan

@DavidCarlisle No I swear, I have only auctexish stuff on the init file.

user19161

@kan OK, or you can always stick to the two-yearly LTS versions.

kan

Well, I am happy with all your help, and after just two futile attempts, I can now work with emacs...

@JasperLoy Or, I might even try some other flavour, you never know.

David Carlisle

02:00

@StephanLehmke no it'works in verbatim (as shown) it's just that scantokens isn't doing anything useful there. What I think he intended was : scan the tokens with the catodes used in the preamble, but if (as in your MWE) the preamble is carried in some macro, that catcode regime is unknown so what was omplemented was rescan with catcodes used in body (verbatim in your case)

user19161

@kan Though I recommend you try the GNOME shell immediately, just install gnome-shell and log out and log in to GNOME at the start screen instead of Ubuntu.

David Carlisle

@kan might be in your site-init file try emacs -q or emacs --no-site-file (or both those arguments)

kan

@JasperLoy I did that sometime ago, and I did not like it all that well.

@DavidCarlisle Well, but, now, I seem to have fixed it with setting inhibit-startup-screen variable. So, I am afraid, I won't be able to see any difference now?

user19161

@kan Ah, congrats! You can join the ranks of Mark Shuttleworth!

kan

@JasperLoy my parser fails but I am not particular about it either.

David Carlisle

02:06

meanwhile:

England 1st Innings
330 all out (145.5 overs)
India 1st Innings
297 for 8 (130.1 overs)

kan

I don't know but this test is supposed to be finger-biting, I am not sure.

toned down finger-biting.

rebooting...

David Carlisle

@kan yes well cricket tends to be slow motion suspense:-) I've not seem any of the play, but India looks to have had a good day

Speravir

@StephanLehmke What Bruno said: Florent Chervet is very active in comp.text.tex – strangely under the initials GL … and if you are not active in ctt, let @Bruno do the bug report (they can speak French together, if I’m not terribly wrong).

Werner

07:51

@PauloCereda At the moment, I'm not really concerned with the language. I though perl, since it has a grep/sed feel to it. I have never programmed in perl nor java, but my preference would be something that I can run from the command line in TeXnicCenter.

David's grep/sed script works after some modification, but only if I run it directly from the command line. As soon as I place it in a batch file and run it as part of a script within TeXnicCenter, variables don't seem to be stored anymore and things don't work...

I'm not sure whether it has to do with the fact that there are limitations to running a batch file from within a batch file; causing variables to not be stored anymore.

I tried to circumvent that somewhat by writing stuff to a file and reading it again, but TeXnicCenter (or probably DOS) complained when I tried to read something from the input stream using < file.txt.

While writing this, I almost feel like the typical LaTeX newbies, writing about something the "doesn't work" without being able to reproduce the problem here.

:-/

@DavidCarlisle Yes, I can give it command-line combinations using a batch script. I've noticed that pure DOS requires slightly different syntax to got things right, which might be the main problem. For example, I need to use double quotes instead of single quotes. Also, I'm checking for %%%%BoundingBox rather than the suggested %%BoundingBox - again, it seems like an "escaping" issue under Windows...

It's just annoying knowing exactly what I want and how I would think one should do it, but not knowing how to do it.

egreg

11:04

@JosephWright The "option clash for xcolor" question is a problem: while the linked answer tells something, in this case finding the culprit is harder.

Joseph Wright

@egreg In that case, probably not a dupe

egreg

@JosephWright However, it's quite localized (because of that strange package), so probably closing it is not a bad idea.

user19161

11:23

I saw the book lottery, so exciting!

Joseph Wright

@JasperLoy Did you enter

user19161

@JosephWright No, I saw that all the numbers I wanted were taken up already. =(

mhelvens

Hi all! I was wondering, how can I get one of those "this question lead to a package" things at the top of my question (tex.stackexchange.com/questions/78155/…)?

user19161

@mhelvens Well, two upvotes from me to get you started. =)

mhelvens

@JasperLoy Thanks. ;-)

Anyway, I can see it's not an 'official' thing. It's just some annotation at the top of the question. I found it on meta.

user19161

11:28

@JosephWright Too bad you can't win the book yourself. =)

mhelvens

@JasperLoy Tell me about that lottery. :-)

user19161

42

Q: Who Wants to Win a LaTeX Book?

I am organising a lottery and Springer have kindly agreed to donate a copy of LaTeX and Friends. Joseph Wright has kindly agreed to act a an independent witness/judge and he knows the winning number. <LEGAL STUFF> The lottery closes when the judges announce the winner. The judges decision...

discussion featured publicity

mhelvens

@JasperLoy Thanks.

Nuno Nunes

Hi all. I've left a comment on someone's answer to a question of mine. Could anyone please tell me if that person automatically receives a notification here on tex.stackexchange?

mhelvens

@NunoNunes They do.

Nuno Nunes

11:38

@mhelvens ah, great! thanks for that :- )

user19161

@NunoNunes You need to use @ plus the first three non-space characters followed by a space to notify someone who comments on a post, but you can just comment under the post without @ and the author of the post will be notified.

user19161

So for example, @nun pings you here in chat!

Nuno Nunes

That's exactly what I did, thanks. I was just wondering whether the author was notified about comments, too.

@JasperLoy

user19161

@NunoNunes Yes, you can use tab complete too, like you just did.

mhelvens

@JasperLoy You don't know. Maybe he just typed the whole thing. ;-)

user19161

11:44

@NunoNunes Also, in chat you can just click on the arrow on the right to reply to a line. You see the arrow if you hover your mouse.

Nuno Nunes

@JasperLoy Always learning! When I type @ followed by the first three letters a small window pops up with the full username. That's what I used/am using. Seems to be a new feature of this website.

user19161

@NunoNunes SE has new features and rules now and then, there are surprises every day!

Paulo Cereda

12:18

Dear reviewers (@AlanMunn, @lockstep, @egreg, @JosephWright, and others): the interview with Ulrike is ready for proofreading. :)

Joseph Wright

@PauloCereda Working on it now

Paulo Cereda

@JosephWright Thanks. :)

Joseph Wright

@PauloCereda A few typos and capitalisations sorted

Brent.Longborough

pbelmans.wordpress.com/2012/12/15/…

Interesting project - a 3000+ page book written in LaTeX. But some very curious coding conventions, including $$ for display maths, and {\it ...} for one specific type of name.

kan

@Brent.Longborough I should tell you, the math that they write about rather beautiful.

And, I am excited about xyjax!

Brent.Longborough

12:31

@kan Yes, it rather looks like that (to a complete amateur who got lost at paragraph 3 ):-> )

Paulo Cereda

@JosephWright Danke. :)

@Brent.Longborough: Hi Brent! :)

Long time no see. :)

kan

Yeah, I was about to say that and this whole Geometry thing just turned me on. :)

Brent.Longborough

@PauloCereda Hi, Paulo. I've been very busy coding, and travelling some.

Oh, and playing with L-systems in TikZ

Oh, and building my own Rainmeter clock

Oh, and learning Inkscape

Paulo Cereda

@Brent.Longborough Good times then. :)

Brent.Longborough

Yep

kan

12:45

BTW, that we are talking about Commutative diagrams, let me say that there is a program called Commu which is mac-only for drawing commutative diagrams quickly. Mac only. sigh :(

mhelvens

13:09

As I write packages, I keep needing constructs that don't exist yet, making me write yet another package. At this rate, I'll never break the recursion and finish my 'first' package.

Stephan Lehmke

@mhelvens On constructs, never forget the rule "first look whether it's already in the etoolbox package" ;-)

Paulo Cereda

@JosephWright: Can we publish the interview? :)

mhelvens

Although, perhaps this one already exists. I want to allow users to nest my new environment, but I'm afraid of messy name-clashes of utility macros. Tell me, is there a package that can generate unique csnames for me in a convenient way?

@StephanLehmke Not this time. See my previous message.

Stephan Lehmke

@mhelvens \global\advance\mycount\@ne\edef\newname{\expandafter\noexpand\csname name@\number\mycount\endcsname}

mhelvens

@StephanLehmke Yep. I keep using a similar construct in my own code. I use it so often, I feel I should extract it into a new package.

Stephan Lehmke

13:14

@mhelvens Ok. That's surely a way to write a lot of packages ;-)

mhelvens

@StephanLehmke Additionally, how can I be sure that I won't run into the same problem with \newname itself? I'd then have to add auxiliary macros and pass the csname as a parameter to solve that problem in a clean way.

@StephanLehmke I'd just like to package all of that stuff into a new package that gives you clean code.

Anyway, I keep coming across low-hanging fruit like that.

But it's boring work. Writing and documenting them.

Should be done, though.

Joseph Wright

@PauloCereda Fine with me

Paulo Cereda

Stack Exchange TeX Blog: TeXtalk: an interview with Ulrike Fischer

Paulo Cereda on December 16, 2012

Welcome to the TeXtalk! We have a very special guest for today: our friend Ulrike Fischer, 41k+ rep, 181+ badges, and our font expert. Get ready for this awesome interview!

Stay tuned for the next episode of the TeXtalk!

4

Great job, guys!

egreg

13:38

@mhelvens There's never a guarantee against name clashes; but if you use a common prefix, the chances are pretty low. Without an example of what you're trying to do it's difficult to say more.

mhelvens

@egreg Sure. I wasn't talking about name-clashes between different packages. In this case I was referring to nameclashes between auxiliary macros when people are nesting your construct. i.e. how do I cleanly organize my code so that my construct is nestable?

@egreg It requires some boilerplate. I understand how the boilerplate works, but I'd rather have a single easy-to-read construct that handles this stuff for me.

egreg

@mhelvens Environments open a group; if you don't use global assignments you can safely have many macros that have the same name, but different meanings in the various levels.

Marco Daniel

@PauloCereda: Did you test this small program: rodsbooks.com/refind

mhelvens

@egreg Indeed! And usually I like to open as few groups as possible. In this case, particularly, I want any local definitions inside my construct to survive.

egreg

@mhelvens This is too generic, sorry.

mhelvens

13:43

@egreg It's ok. I'm working on a simple solution as we speak.

@egreg While I have you here, are you interested in writing a paragraph about the limitations of your \minput (see tex.stackexchange.com/questions/86248/…). I can quote you by name in the package docs, if you'd like.

kan

@egreg How do you manage the .bib files in the way we have installed TeX? Is the following workflow right? 1. sudo jabref 2. mktexlsr ?

Paulo Cereda

@MarcoDaniel No. :(

Joseph Wright

@kan Huh?

egreg

@mhelvens I've added a comment about that.

Joseph Wright

@kan I keep my .bib files inside my local TeX tree, so they are found without hashing

Marco Daniel

13:47

@PauloCereda I think I will try it. ;-)

kan

@JosephWright Oh, I see. Sounds like a good idea.

Joseph Wright

(OK, more accurately I keep them in my documents area and link from my local tree, but the principal is the same)

mhelvens

@egreg Thanks!

kan

@JosephWright Oh, so only the link is in the local tree?

And, the actual file is at wherever your tex file is?

Joseph Wright

@kan Yes

@kan I have a master .bib file which has most of my refs in

kan

13:50

@JosephWright Oh, I see, makes sense.

egreg

@kan If you put your big .bib file in ~/texmf/bibtex/bib, all TeX programs can access it without hashing. And JabRef can easily use it (no hidden directories are involved)

kan

@egreg Oh, I see. I'll try it out this night...

and ask about the places I get stuck!

mhelvens

14:28

Hi all. Quick question. Is there a simple way to specify vertical whitespace that should 'overlap' with any existing vertical whitespace? I want to add some space above and below my code-blocks, but not 'double' the space when I have two consecutive code-blocks.

Feels like this should already exist, but a Google search didn't help.

Stephan Lehmke

@mhelvens \addvspace will sorta take the maximum of all consecutive spaces.

mhelvens

@StephanLehmke That's probably exactly what I'm looking for. Thanks!

Stephan Lehmke

@mhelvens Don't let any penalties interfere though.

mhelvens

@StephanLehmke Hm. I'm getting "! LaTeX Error: Something's wrong--perhaps a missing \item." where a simple \vspace just works. Does that ring a bell with you?

Stephan Lehmke

@mhelvens When using \addvspace you need to be in vertical mode, i.e. after \par. That \vspace "works" in horizontal mode is at the least a relative truth.

mhelvens

14:39

@StephanLehmke Well, a normal \vspace doesn't complain, at least. ;-)

@StephanLehmke I just added a \par. Indeed, that fixed it. Thanks!

Stephan Lehmke

@mhelvens Which is, from LaTeXs POV, a carelessness.

mhelvens

@StephanLehmke I can understand why you say that, if it violates the specs. However, it makes me wonder why the specs don't allow it when it seems to 'just work'.

Stephan Lehmke

@mhelvens No the specs allow it, but normally LaTeX is very picky wrt. what belongs to horizontal and vertical mode. For \vspace there even is a special "horizontal" case with \vadjust which I can only see as sloppiness...

mhelvens

@StephanLehmke Ok.

David Carlisle

15:05

@mhelvens \vspace goes to some considerable lebgths to "work" in horizontal mode, and the result is rarely what the user wants, so it isn't clear that the effort was worth it. (The space does not appear where the command is issued but at some other point not totally unrelated to the line after the point in the text where \vspace was issued.

mhelvens

@DavidCarlisle Yes, I've noticed this. ;-) Before I really got into (La)TeX the behavior of \vspace would often confuse me.

@DavidCarlisle However, why they don't treat \addvspace exactly the same is beyond me.

Well, I suppose I could come up with a reason. But it would only be speculation. I won't take the time to research this. ;-)

David Carlisle

@mhelvens Because if we treated it the same, it wouldn't work. \addvspace needs to see preceding spaces (otherwise there is no point) and if it is in horizontal mode and v-adjusted in to the vertical list teh preceding thing will be the box out of which it has just migrated

mhelvens

@DavidCarlisle Yep. I can see that.

David Carlisle

@mhelvens so what would you like it to do (other than give an error) of course it generates that particular missing item error in 2e just because it did in 2.09 and we didn't feel like changing it, even though it is almost never caused by a missing item in practice

mhelvens

@DavidCarlisle No, I understand now why it does what it does. I'm at peace with it. ;-)

David Carlisle

15:11

@mhelvens :-)

mhelvens

So I'm writing a super light-weight package (in package-writing-recursion-depth 4 or so). It introduces command \with which fully expands its 8 optional arguments and turns them into #1, #2, etc. in the mandatory 9th argument. Am I reinventing the wheel here?

I seem to be using it all over the place.

David Carlisle

@mhelvens 8 optional arguments? How is that usable in practice, since you can't omit one without omitting the following ones as they are positional? (of course fully expanding arguments and passing them to the underlying command is what xparse is all about)

mhelvens

@DavidCarlisle I simply extended the command to its limit, allowing the user to pass up to 8 arguments. They will be numbered from #1. I don't care that they won't be able to use the 4th without also using the 1st, 2nd and 3rd.

@DavidCarlisle And the idea behind this command is to be able to do it inline, without explicitly defining a new command.

@DavidCarlisle Example: \with[x][y][z] { Three letters: #1, #2 and #3 }

(Maybe I can still squeeze out a 9th argument. But that's not my biggest concern right now.)

David Carlisle

@mhelvens OK, well you could do it inline with exp_args but (over latex2e) loading all the l3 stuff is hardly lightweight so if that's all you need then perhaps just defining it directly makes sense.

@mhelvens if it's all inline why not just Three letters x, y, z ?

mhelvens

@DavidCarlisle The main purpose here is to make it easier to write recursive macros, and be able to reliably use auxiliary macros even after the recursive call.

@DavidCarlisle exp_args is something completely different, I believe.

David Carlisle

15:26

@mhelvens presumbly (since you use the #1 syntax` you define a macro internally

mhelvens

@DavidCarlisle I do. But that one's safe to redefine inside, since its only use will already have been expanded.

David Carlisle

@mhelvens well not completely different expargs would handle fully expanding teh optional arguments and making them mandatory hence going from \with[x][y][z] { Three letters: #1, #2 and #3 } to \withbase{x expanded}{y expanded }{z expanded}{...} the inline stuff after that would just be teh definition of withbase to define the helper macro and execute. But I've got to go...

mhelvens

@DavidCarlisle Ok, I suppose it's related. Thanks for your insights!

egreg

16:06

@PauloCereda Too easy, today. :) @DavidCarlisle How's the test match going?

Joseph Wright

@egreg England a shade ahead, I'd say, but could go either way

Paulo Cereda

16:19

@egreg :)

Alan Munn

@egreg Ahh, pragmatics. :)

Paulo Cereda

Why doesn't ESPN show cricket? :(

David Carlisle

@egreg where's Psmith when you need him

India: 326-9 (143.0 overs)
England: 330 & 161-3 (79.0 overs)
Close (England leading by 165)

egreg

16:42

@AlanMunn I didn't want to upset David by just saying that Juve won with an easy 3:0. :)

Paulo Cereda

@egreg Easy game. :P

David Carlisle

@egreg It's Ok you're allowed a bit of football discussion after cricket comments

egreg

17:37

@cmhughes Sorry, but your comment to my answer about uppercase Greek letter is incorrect: an upvote won't reward me with 10 rep. At least for today. :P

cmhughes

@egreg lol, I can't imagine you hitting rep cap

Alan Munn

17:59

@cmhughes You're missing a 'not'. ;)

Alan Munn

18:27

@JosephWright Do your mod powers let you edit comments? If so, could you correct the link text in my comment to pgfplotstable instead of pgfplots ? If it's impossible I can just delete the comment and add a revised one.)

David Carlisle

18:52

@egreg true, just merely upvoting doesn't really help you. Of course accepting would have given you 15 rep but unfortunately the OP had already accepted another answer.

egreg

@DavidCarlisle Yeah. :(

@DavidCarlisle I'm still asking whether it's simpler \mathit{\Gamma} or \varGamma. :P

David Carlisle

19:23

@egreg they'll probaby want \mathbf before long though,

Alan Munn

19:47

@JosephWright Thanks!

tohecz

20:34

Hello

How do I "disagree" with a flag? The review tools still show me some flags that I consider incorrect but there's no button to deal with that.

David Carlisle

@GarbageCollector spots that 250 > 200

2

egreg

21:44

@PauloCereda Inter and Napoli lost their matches. :) @DavidCarlisle I know, this is not sporting behaviour. :) (Note that I used the "u".)

Alan Munn

I've always liked the following line from `fontspec`:
`\aliasfontfeature{Color}{Colour}`

5

And I give up on markup in chat. It seems so haphazard.

egreg

@AlanMunn Didn't \aliasfontfeature{Color}{Colour} work?

Just one backquote fore and aft.

Alan Munn

@egreg Apparently not as is evident from the output.

egreg

@AlanMunn Quite bizarre.

Alan Munn

@egreg I've had this happen before, and I'm not quite sure why. Maybe it's something to do with the linefeed I put in to put the \aliasfontfeature on its own line.

tohecz

21:59

@AlanMunn you cannot use ` syntax in multiline answers

Alan Munn

(See, that one worked.)

@tohecz Ah, ok. But you can't also make just one line code and the other line not, right?

tohecz

@AlanMunn of course you cannot ;)

Alan Munn

@tohecz Didn't Kurt Vonnegut write a book about that situation?

tohecz

who? sorry, I'm working and want to have it done quickly, because this one is just fucked up

Alan Munn

@tohecz Forget it, it was an allusion to Catch 22.

egreg

22:08

@AlanMunn Kurt Vonnegut? He wrote Slaughterhouse-five. Joseph Heller, perhaps. Different numbers, you see?

Alan Munn

@egreg Oops. Fail. You're right. No wonder @tohecz didn't get it.

@egreg Notice how the starred version of the comment suddenly formatted the code right!

egreg

@AlanMunn Mysteries of markup.

WG-

Hello

egreg

@WG- Hi!

WG-

Ya I wanted to ask a question

But I think I will maybe better post a question on the board

Alan Munn

22:23

@WG- It's fine to ask here first if you're not sure if it's a good question for the site, or you need a really quick answer.

We might answer it or just tell you it's a good question for the site.

WG-

Well it is about how to typeset the current (electricity) derivative in shorthand notation.

Like m\ddot{x} + d\dot{x} + k but then for electricity. I got L\frac{\dd i}{\dd t} + Ri = v - v_{m}

but I don't want to write di/dt but just i with a dot above

so I was wondering what the best typesetting was.

egreg

@WG- i with a dot above seems rather ambiguous.

tohecz

@egreg yeah, non-capital i is a very bad notation for something you wan't to differentiate

WG-

I know thats why I am asking this question :-) because I don't either want to use di/dt

tohecz

@WG- I would go for the operator notation \partial_t i, but don't forget to definte it properly

WG-

22:30

Hmmm... yeah well that is either not very consistant in the way I typeset all the other derivative equations.

egreg

@WG- Isn't I the usual symbol for electric current?

WG-

the capital I is used when the current is constant, not time dependend. i is used when the current is time depended. Same counts for the voltage. V when voltage is constant, v when it is not.

tohecz

@WG- @egreg this.

Don't speak about confusion in $i=\Re I e^{\omega t}$ :p

egreg

@WG- Well, the usual answer is "when in Rome, do as Romans do". What's the common usage in your field?

If it's the dot notation, then \dot{i} will do.

WG-

Ya normally it is di/dt. But I want to be consitant and not change very thing back to dx/dt instead of \dot{x} etc. also di/dt consumes to much space. So I guess I will just go for \dot{i}

egreg

22:39

@WG- Probably the best is to define a macro for it: \newcommand{\deri}{\dot{i}}, so that you can easily change your mind and switch to the Leibniz notation by only changing the definition.

WG-

That is proberbly a good idea yes.

Thanks for the tip.

egreg

@WG- Referees are known for thinking differently from authors with respect to notation. :)

WG-

Hmmm I'm not sure what you mean with referees? You mean literally a referee, someone who in this cases checks my work ;-)?

egreg

@WG- Yes, those. :)

WG-

Harmonic oscillator

In classical mechanics, a harmonic oscillator is a system that, when displaced from its equilibrium position, experiences a restoring force, F, proportional to the displacement, x: : \vec F = -k \vec x \, where k is a positive constant. If F is the only force acting on the system, the system is called a simple harmonic oscillator, and it undergoes simple harmonic motion: sinusoidal oscillations about the equilibrium point, with a constant amplitude and a constant frequency (which does not depend on the amplitude). If a frictional force (damping) proportional to the velocity is also ...

here they use also \dot{i}

last column, last row of the table

:-)

egreg

22:50

@WG- And it's really awful. :)

tohecz

@egreg indeed

@WG- use \iota for the current from the very beginning.

or of course, \imath

egreg

@tohecz But \dot\imath will be very ambiguous.

tohecz

@egreg as long as you use the dotless one anywhere else, it should be ok

egreg

@tohecz I'm not so sure. People is used to the dotted i, and will have difficulties in spotting the difference. Anyway, the best is to follow usage.

tohecz

use a large dot :p

Paulo Cereda

23:11

I'm back. :)

egreg

@PauloCereda Welcome back!

Paulo Cereda

@egreg Thanks! :) Anything happened when I was gone?

Alan Munn

►

@PauloCereda In football news, Dallas 10, Pittsburgh 10 in the 3rd quarter.

Paulo Cereda

@AlanMunn Yay! And the Saints?

tohecz

@PauloCereda Yeah, I got completely pissed off by one article that I got in W@$%#$#@!$

egreg

23:20

@PauloCereda India-England is in pause.

Alan Munn

@PauloCereda No idea. I'm watching this game but I don't follow the scores generally. I just thought I'd throw in some new sport for a change. And fencing results are a bit too obscure.

@egreg The deleted post was saving you from Eminem. :)

Paulo Cereda

@AlanMunn I tried to follow [American] football, but the schedules are complicated for me to watch, and I still didn't get the rules (I usually watch the SuperBowl though). I tried rugby, which seems to make more sense to me, and it's funnier. :)

@AlanMunn: speaking of rugby, have you ever watched the Brazilian ads on this sport? :)

Alan Munn

@PauloCereda I know absolutely nothing about rugby. Just some random terms like 'scrum' and 'try' for which I don't even have anything close to a full meaning for.

@PauloCereda I'm not really a fan, but a good close game can be very entertaining to watch.

Paulo Cereda

@AlanMunn: youtube.com/watch?v=HFmpCX2IcB4 and youtube.com/watch?v=2gtQpNZGpEE :)

The last one is priceless. :P

Alan Munn

@PauloCereda LOL. I like the autograph one best.

Paulo Cereda

23:31

@AlanMunn I like the maria-chuteira one. :) "Eu adoro o rubgy." :)

egreg

@AlanMunn Great!

@PauloCereda The first rugby match I saw at the stadium was Italy-New Zealand. Still not classified as test match, at the time.

This is the Italian team (it had some foreign players, actually). Padova, 22 October 1977

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends