TeX, LaTeX and Friends

vlg

12:23 AM

got it

CarLaTeX

5:02 AM

@Werner Ccccoooonnnnggggrrrraaaattttssss!!!!

Joseph Wright

8:24 AM

@egreg Fixed

@egreg Does it cover everything? Should I write an extended version? (@vlg?)

GermanShepherd

10:13 AM

@DavidCarlisle Yes, xexetex.

JouleV

10:30 AM

Hi. What function should I use for this graph?

I'm not really good at trigonometry...

@Werner Amazing! Congratulations!

egreg

@JosephWright I'd also change ”Why choice one or other method?“ to “choose”.

David Carlisle

10:47 AM

@vlg no you misread.

user396517

@JouleV en.wikiversity.org/wiki/Damped_oscillation

Harald Hanche-Olsen

@JouleV Something on the form $-a e^{-bt} cos(ct)$ , I think, with $a=1.5$, $c=180$ if you're measuring t in degrees. Or $c=π$ if in radians. What about $b$? hang on …

@JouleV … it appears to decrease by a factor 0.75 per period. So $b=-\ln(0.75)/(2π)$ looks reasonable.

David Carlisle

11:11 AM

@GermanShepherd Does Kannada use hyphenation? You could use \sloppy to increase stretchability of white space. Otherwise, make a small example and ask a question on the main site

JouleV

11:35 AM

@user181777 @HaraldHanche-Olsen Thank you very much!

GermanShepherd

@DavidCarlisle I 'll try to make a small example of it. :)

Alan Munn

2:32 PM

@HaraldHanche-Olsen Apparently 5 seconds of 172 dB speech. :)

vlg

@JosephWright no, it's fine. I may have misread. As egreg pointed out, it's just that a \global is required, the wording confused me into thinking that direct setting via \mycount ... \relax "should be avoided"

Did a likely meaningless benchmark of a count getting set to sth, then getting set to 0; and a count getting set the same thing, but the getting advanced by negative itself. The latter showed up as slightly slower and with more ops.

vlg

3:50 PM

Is there a way to negate output? Like letting output go only to console, thus no errors for not being used in the proper environment, or something to the same effect? I want to benchmark a variant of a command, but it produces visible output, and is intended for used in a longtable, which may complicate the benchmark process

StrongBad

4:07 PM

@vlg what about sticking it in a savebox?

vlg

I'm not too familiar with savebox, will look up and try it. Testing on the whole document w/ output, ~170k instances of the two cmds in question, using the xparse \NewDocumentCommand adds and extra 6s on top of what I get with \def

Ulrike Fischer

4:23 PM

@MarcelKrüger as you just there. Is there an easy way to avoid that a luc/lua is created when using the harf mode?

Marcel Krüger

@UlrikeFischer I am not quite sure right now, I would not have expected this files to be created currently. B I am thinking about a bigger luaotfload rewrite to make it more independent from the fontloader: Basically with HarfBuzz I would want luaotfload to work without relying in the fontloader. Then this would no longer happens for non-fontloader fonts.

AlexG

@GermanShepherd : What about the flying airplane question? Still something that can be improved?

Ulrike Fischer

4:46 PM

@MarcelKrüger I think it doesn't happen in plain, so fontspec could be involved. But I need to make more tests.

barbara beeton

5:03 PM

This article -- What has WYSIWYG done to us? (published in the Seybold Report on 30 September 1996) -- was pointed out to me recently. It describes a number of the tools (structured templates vs. WYSIWYG, but not including Word) that were available at the time and points out the pros and cons of each with respect to a desired result that exhibits good typography and can be reformatted for various uses. I found it interesting; others may also.

Marcel Krüger

@UlrikeFischer If you only see this with fontspec, it is the same problem as in github.com/khaledhosny/harf/issues/30 We can not avoid this without changing fontspec because fontspec loads the font without harfbuzz before providing all the features and the mode.

Joseph Wright

@vlg That's not surpising: one wouldn't expect to use \NewDocumentCommand-defined stuff that much. That's a lot of document mark up!

Ulrike Fischer

5:24 PM

@MarcelKrüger Could be, I will test it, and if this is the problem, discuss it with Will. Some of the large fonts really strain the fontloader in this step ...

vlg

6:33 PM

I'm typesetting a logbook, and I made a command to input the data in a desired way, but it got out of hand and now it render the date as a rotated mrow along the resp. rows, but pagebreaking, and accurate counter stepping, resetting and a whole lot of necessaries- and you get a mess; but it automatically keeps track of date, collects and stores wanted vals in .dat values with the possiblity to graph them externally and add them later
Maybe overengineered, definitely suboptimalized.

Pippip19

Is there a policy on posting non-(La)TeX code?

I am asking in light of this question: tex.stackexchange.com/q/476008/172926

Joseph Wright

@Pippip19 That's not TeX at all, it's just the math mode inspired syntax, I suspect: do you use Knuth's alogorithm's?

David Carlisle

@vlg perhaps but that's not really a meaningfull test. NewDocumentCommand is not designed for defining programming constructs in a tight loop. As its name implies it is designed as a user command for document level constructs, so a document may only expect to define at most a few hundred of these not hundreds of thousands.

Pippip19

@JosephWright I was making a distinction between other programming languages.

Joseph Wright

6:48 PM

@vlg Sounds like programming, not document commands, as @DavidCarlisle says

David Carlisle

6:59 PM

@UlrikeFischer thanks for the luahb testing, I'm really short of time at present possibly can't look at that before the weekend (not helped by the fact that my car stopped at a roundabout this morning:-)

@JosephWright ^^ dll testing....

Ulrike Fischer

@DavidCarlisle it was busy day ;-). But quite interesting - and I want to keep it going.

vlg

@DavidCarlisle It works, I don't have anything to compare with sadly, and yeah, it's turned into bad tex programing project. I'm learning along the way. In any case, I'm not going to C++ for typesetting.

user0

7:17 PM

@StefanKottwitz I can wait until manual editing is possible. Now I think that creating user-0 was not a very good idea from me; however, I have not done anything through the new account. As of now, it is nothing more than yet another entry in texwelt's database.

David Carlisle

@vlg who said C++ ?? the idea of l3 is that you use the expl3 programming layer (with _ and : for programming) and then (if you want a latex2e-style interface) at the top you use a xparse and a \NewDocumentCommand to define a user interface.

Joseph Wright

@DavidCarlisle Very modern ;)

David Carlisle

@JosephWright that's me

Alan Munn

7:46 PM

My first ever GitHub bug report. I either write really good software or really useless software. :)

David Carlisle

@AlanMunn must be the latter?

Jun 29 '17 at 16:15, by Paulo Cereda

@DavidCarlisle you are not mean :)

Ulrike Fischer

8:02 PM

Oh god, and now swiglib ;-).

Harald Hanche-Olsen

8:20 PM

@UlrikeFischer Swiglib is some kind of successor to god?

manooooh

Hello everyone!!

Does someone know how to delete a blank space between \item (of enumerate) and \begin{theorem}?

It happens like MS Word when using nested enumerate lists

@marmot @CarLaTeX @samcarter prepare to see in a couple of hours a gallery of photos of a particular duck showcase... :) :) :)

vlg

8:39 PM

@DavidCarlisle Thought TeX was made for typesetting beautastic documents, not programming, I barely got into how some of the primitives or other things work, and then I learn that l3 has been here since 98 or somewhen there, and it expands a lot, but I still don't know understand much about the guts.

David Carlisle

8:53 PM

@vlg sure it is, but for example if you want to go \begin{tabular}{cc>{\raggedright}Xp{2{cm}} someone has to write a parser for that and make it into a primitive TeX \halign call, and that parser has to be written in tex, expl3 is designed to make writing such things easier.

@UlrikeFischer I was hoping to avoid that (and FFI)

Ulrike Fischer

@DavidCarlisle I didn't understand if it is a requirement or a side remark ;-( I do understand that if all dll/so need to be present, then one need a naming system. But I have only windows, so why can't the dll simply be in the bin folder like the other dll?

Alan Munn

@DavidCarlisle Undoubtedly.

Ulrike Fischer

9:12 PM

@DavidCarlisle I put the two dll in bin/mingw64/lib/lua/luaharfbuzz and it worked!

Harald Hanche-Olsen

@manooooh Why on earth are you doing that in the first place?

manooooh

@HaraldHanche-Olsen is that bad...? Anyway, then I figured out that I had to write something after \item and before \begin..., so now I do not need an answer

David Carlisle

@manooooh not necessarily bad but very weird. theorems are usually top level constructs like sections, you would not expect to find a theorem in an enumerate

Harald Hanche-Olsen

@manooooh Well, bad … I'd say it is staggeringly bad style, yes. Theorems are top level constructs, they were never meant to be stuffed into an enumeration. After all, they have their own numbering and all.

David Carlisle

@HaraldHanche-Olsen group think

Harald Hanche-Olsen

9:19 PM

@DavidCarlisle great minds …

Alan Munn

Does anyone here have experience making Applescript applets run on Mojave? I think it's now sandboxed them into oblivion.

manooooh

@DavidCarlisle thanks for your point of view. I am writing an answer of an exam, so they enumerated the exercises like 1,2,3 and (a),(b) inside each one, so I cannot do much. Then I realized that I had to add text in between, so there were no problems

Harald Hanche-Olsen

@AlanMunn Applescript works fine on Mojave as far as I know, but what is an Applescript applet? Never heard of them.

manooooh

@HaraldHanche-Olsen in an exam I have to mention a definition in a particular exercise inside enumerate because the exercises are enumerated (no surprise). But thank you for your suggestion, I will consider it!!

Alan Munn

@HaraldHanche-Olsen It's a way to make an applescript a double clickable app.

David Carlisle

9:23 PM

@manooooh but do you have a theorem in an exam answer??

manooooh

@DavidCarlisle yes, not in this particular exam (I have to mention two Definitions) but in another yes. So I say (...) To do that, consider the following theorem: \begin...

David Carlisle

@manooooh if you have consider the foll... \begin then the theorem isn't immediately after the \item which is the case you asked about

manooooh

@DavidCarlisle yes... at first I thought that I only needed \item \begin but then I realized that I had to mention something before \begin and after \item. Sorry for not mentioning it before

David Carlisle

@manooooh but then presumably the spacing issue doesn't arise. (It's still pretty rare to have a theorem construct inside a list but not quite as weird as having it immediately after the \item

manooooh

@DavidCarlisle it is not weird... it is just an "enumerated" examen which needs a theorem/definition!!

Anyway, I am aware thanks to you that it is not usual to write a math environment inside an enumerate

Harald Hanche-Olsen

9:36 PM

@manooooh Hmm? Math, including display math, is not uncommon within an enumerate. Theorems are a different matter.

manooooh

@HaraldHanche-Olsen I treat Theorems, Proofs, Definitions etc. as one

Harald Hanche-Olsen

@manooooh So do I. But when I hear “math environments”, I think of equation, multline, align and so forth. Theorems, proofs, definitions usually contain math, but that is a different matter.

manooooh

@HaraldHanche-Olsen oh yes you are right. My apologies. I should not say "math environments"

Harald Hanche-Olsen

Well no harm done; just some momentary confusion.

Alan Munn

@HaraldHanche-Olsen Would you mind testing one for me? I don't have Mojave installed on any of my machines.

Harald Hanche-Olsen

9:45 PM

@AlanMunn I suppose I could try.

vlg

As an aside, should would strive to use l3's function definitions such ascs_new:Npn, or the xparse ones, which are built using the former's? I ask, because xparse isn't included in the l3 bundle, which I thought of as strange

David Carlisle

@vlg as I said, they are working at different layers. If you are programming functionality use expl3, if you are defining a latex2e-style document syntax for some already written functions (which may be written in expl3 or llaetx2e code or whatever0 then use xparse.

Alan Munn

@HaraldHanche-Olsen Thanks. I might send you email but not immediately. It seems that the app won't run anymore even on High Sierra, so the problem might be more general than Mojave.

Harald Hanche-Olsen

@AlanMunn Okay.

manooooh

10:22 PM

Suppose we have the following set:

$A=\{(x,y)\in\Bbb R^2\mid0\leq x\leq1,0\leq y\leq2)\}$

Is it correct to separate each interval using only , or do we have to add a little space i.e. \, or any other after the comma?

P.S. I need the answer both for $ and \[

Does AMS say something about this? Thanks!

marmot

@manooooh Usually if you ask n people you will get 3n+2 different opinions. (I would add spaces.)

manooooh

@marmot oh, I always did with no space :-[ What kind of space would you use?

I would use , \;

marmot

@manooooh Yes, that works.

David Carlisle

@manooooh probably I'd add \, but I'd have to have a look at the typeset version , is really set up for coordinates or function argunments (x,y) or f(a,b) with tight spacing not for something separating clauses, which might want wider space.

manooooh

@DavidCarlisle how do I specify you the typeset version? It is in a normal article document class, no change of font size, no change of font

David Carlisle

10:37 PM

@manooooh oh I could make it myself,but I'm just chatting, what I mean is if i was doing that I'd decide after looking at it

manooooh

@DavidCarlisle \, > \; in this case? :P :P

David Carlisle

@manooooh well not actually you have defined \Bbb somewhere

manooooh

@DavidCarlisle ;) \mathbb

marmot

@manooooh I would do $A=\{(x,y)\in\Bbb R^2\;\;0\leq x\leq1 \wedge 0\leq y\leq2)\}$ .

manooooh

@DavidCarlisle ok no problem, it's enough for me to be told that a space must go. Thanks!

David Carlisle

10:39 PM

@manooooh \mathbb isnt defined by article either

manooooh

@marmot yes, I used to use \wedge, but then I realized I am not a mathematician, so sometimes saying \wedge is kind of... use a level that I do not have

@DavidCarlisle \usepackage{amsmath/amssymb}......

David Carlisle

@manooooh yes I know (I do maintain amsmath....:-)

marmot

@manooooh Well, IMHO the wedge is more precise. A comma may mean anything. Also the \mid can be mistaken for "divides".

manooooh

@marmot @CarLaTeX do you want to be spoiled right now?

I have a surprise... :)

Harald Hanche-Olsen

@marmot An often overlooked alternative is to use \text{and} (with suitable spacing). Especially if you are not doing mathematical logic, but just plain mathematics.

manooooh

11:08 PM

2

marmot

11:23 PM

@HaraldHanche-Olsen I agree.

manooooh

Hello! I am a duckling who lives in Argentina and loves to learn about LaTeX and its great utilities. I also like to see other animals. I am in love with the place and its people, so do not expect me to hate anyone; we are all a great community that tries to get ahead <3 (some contribute more and others less, but we all have the same goal).

Here you can see me for the "formal" photo:

Why not get a little more serious O.O:

Sometimes we should be happy, as I am every time I log in to TeX.SE:

When @DavidCarlisle corrects me I say:

If your explanations all I get is to have the ideas a little less clear:

A little more intimate:

And... that's all! I hope you may have known a little more about this duckling <3

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends