« first day (2886 days earlier)      last day (2046 days later) » 

12:00 AM
@manooooh do you use acrobat?
 
@DavidCarlisle yep
 
@manooooh that will be the reason
 
I have the same problem the last months (at first it was not a problem)
@DavidCarlisle oh. I even notice thelog file and the address where the program looks for the tex is correct
 
@manooooh just use a non-blocking viewer (I'd recommend MuPDF or Zathura, the latter having more features, but I just started using it recently and can't whole-heartedly recommend it yet)
 
@Skillmon does that program have certain characteristics, such as a rule that measures lengths and angles of documents? Because sometimes I need it
 
12:03 AM
@manooooh no, of course not. It has nothing GUI like cluttering your screen :)
 
@manooooh acrobat holds the file open for writing so you can not regenerate the pdf while it is running. Almost all other pdf viewers do not do that
 
@Skillmon oh :(
Thank you btw, I will consider that
 
@manooooh I think acroread is one of the few evil readers blocking the file. Almost every other non-Adobe made one should be fine.
 
@DavidCarlisle I don't understand. In the first compilation the .tex could not generate the PDF. No Acrobat program was opened during the compilation nor before (maybe in all the day yes)
 
@manooooh well from your description of the problem i guessed you used acrobat as your viewer, so that remains my guess:-)
 
12:07 AM
@Skillmon but most use Acrobat ... are we mostly doing it wrong?
AGAIN
It is happening now
 
@manooooh yes it is not good for any cycle where you are generating a pdf repeatedly as you have to close the file in the viewer every time
 
@manooooh yes. Don't ever trust any software published by Adobe :)
 
@manooooh check in task manager I bet acrobat (or some dll remnant of it) is still running
 
It forces me to have to create the files on the desktop ... When I want to accommodate them in other deeper directions it does not work, so my last compilation happens on the desktop before I have to do everything back if I want to modify something
 
12:10 AM
@manooooh qed. Don't use Adobe software.
 
:46879412no, I don't think I have ever generated a file on the desktop, (and i do use acrobat occasionally)
 
@DavidCarlisle I can't find it, I will post an image
 
@manooooh there is only one reason to use acrobat: If any funny person has created a PDF depending on the bleeding-edge, never-ever-used-by-sane-people features of the PDF standard, which happens to be only supported by Adobe's readers. On every other occasion there is something better available.
 
@DavidCarlisle ^^^^^^ there are more or less all the details*
@Skillmon okay... MuPDF or Zathtura are free?
 
@manooooh you have a process named "dasHost" :) Amazing. Reminds me of "Das Boot"
@manooooh both free as in beer and as in speech.
@manooooh but power user hating the mouse oriented. Maybe give Evince a shot if you want a feature rich reader that is free and free.
 
12:18 AM
@Skillmon lol I never saw it
 
@manooooh I don't mean the movie but youtube.com/watch?v=zuDtACzKGRs
 
@Skillmon O.O
@Skillmon https://wiki.gnome.org/Apps/Evince/Downloads check for Windows lol, :/
This is pdfTeX, Version 3.14159265-2.6-1.40.18 (MiKTeX 2.9.6500 64-bit) (preloaded format=latex 2017.11.26) 24 SEP 2018 21:22
entering extended mode
**./Primer_Parcial_Fisica_I_03_09_2018.tex
(Primer_Parcial_Fisica_I_03_09_2018.tex
LaTeX2e <2017-04-15>
Babel <3.15> and hyphenation patterns for 75 language(s) loaded.
("C:\Program Files\MiKTeX 2.9\tex\latex\base\article.cls"
Document Class: article 2014/09/29 v1.4h Standard LaTeX document class
("C:\Program Files\MiKTeX 2.9\tex\latex\base\size10.clo"
@DavidCarlisle ^^^^^^^^ That's the log file
With this simple code:
\documentclass[a4paper]{article}

%% Language and font encodings
\usepackage[english]{babel}
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}

%% Sets page size and margins
\usepackage[a4paper,top=3cm,bottom=2cm,left=3cm,right=3cm,marginparwidth=1.75cm]{geometry}

%% Useful packages
\usepackage{amsmath}

\title{Your Paper}
\author{You}

\begin{document}
\maketitle

gf
\end{document}
@DavidCarlisle are all the charges correct or are there a lot of garbage? In that case, what should I do to make it compile better?
 
@manooooh yes but that's latex not pdflatex so it's writing a dvi, the error is unrelated to any tex code in the document it is an operating system permissions level error.
@manooooh changes from what?
 
@DavidCarlisle charges* I mean if all that information is from this file or if it is from another. Maybe it is loading implicitly some packages that I don't want
@DavidCarlisle oh, and do you know what permissions should I have or how do I have to solve it?
 
@manooooh haven't thought of that. The situation isn't especially good on Windows then. Okular is also only available with a whole bunch of KDE software, as far as I know. There is also Sumatra PDF but I don't know which features it has. And there is also Foxit reader, but I don't know anything about it, I think it is free as in beer but not as in speech.
 
12:29 AM
@manooooh you won't need inputenc, once you update to current latex, but other than that it all looks normal enough, your settings for geometry look a bit suspect (you should use the option that makes it round to baselineskip)
@Skillmon I use xpdf, as i have a real window system in windows
 
@DavidCarlisle :) That would be great, but as I don't think @manooooh got an X server up and running in his Windows, I didn't propose xpdf.
 
@Skillmon haha okay, no problem. Maybe I will try searching another PDF output, I will check the others you pass me
 
@manooooh no as I say I guess it is application locking rather than user file permissions, but either way it is an operating system block on file access it isn't related to anything in the file
 
going to bed now. Good night everyone!
 
@DavidCarlisle hmm I don't know how to do it. But it doesn't matter, I was worried about loading 75 languages if I only use Spanish and English (maybe you can configure it to load less and compile faster?)
@Skillmon oh, bye! Thank you <3
@Skillmon @DavidCarlisle no I don't know what is that, I use the Windows that came me with my laptop
@DavidCarlisle ok, so the problem is Windows and their permissions. I still can not compile any .tex that is not hosted on the desktop
What the heck is ps2pdf? That name gives me the error of MikTeX GPL Ghostscript 9.19...
And the file hosted in Desktop works... My mind will explode
 
1:15 AM
@manooooh it's just ghostscript with options set to read ps and write pdf
 
@DavidCarlisle ok
 
 
5 hours later…
6:41 AM
@manooooh the X server is the graphical display server used on most Linux PCs, see wikipedia
 
7:18 AM
@Skillmon ok. Wow the last release was in 2012. Thanks 💪
 
 
1 hour later…
8:39 AM
@DavidCarlisle @JosephWright I added a small luatex-ja example to my testfiles and it fails on travis with
Custom whatsit box_boundary = 5[string "-- luatex-core security and io overloads ......."]:42: bad argument #1 to 'io_open' (string expected, got nil)
 
@UlrikeFischer luatex testing is fun isn't it
 
Locally luatex-ja is trying to load a cached font from TEXMFVAR: ....texlive2018/texmf-var/luatexja/ltj-cid-auto-adobe-japan1.luc and I think travis doesn't like this ...
 
@UlrikeFischer what happens locally if that font is not installed (or conversely is it installed in your travis vm) ?
 
@DavidCarlisle when I delete the texmf-var/luatexja folder it is recreated without problems. I'm not sure if I have everything installed on travis (how can one ...) but I did install luatexja and imho it is part of it.
 
@UlrikeFischer in the travis setup you can install texlive bits with tlmgr and if you need system fonts that are available in linux you could get them via apt-get <some-ubuntu-package-name>
 
8:57 AM
@DavidCarlisle I don't need system fonts, at least the document not. I'm not sure if luatex-ja tries to find some during the compilation.
 
 
2 hours later…
11:09 AM
@Skillmon, please continue our discussion here.
@Skillmon, to illustrate, what is my concern, in my question tex.stackexchange.com/questions/452380/… replace in nodes n2 and n3 shape ellip with block. when nodes are rectangles, they are positioned as i expected, but when are ellipses or circles, they not. border of shapes not obey distnce between nodes. with your solution (adding onchor) is solved position of this anchor, not the whole node.
 
11:48 AM
@Zarko I don't understand what you expect from TikZ in that case...
@Zarko TikZ draws an ellipsis around your text block, should it further stretch the ellipsis or deform it to achieve alignment? I'm confused.
 
12:07 PM
@Skillmon, i expect, that nodes with ellipses shape (the same case is for circle) has distance between their borders are equal to distance declared by node distance. in case of node distance=3mm and 7 mm the distance of ellipse positioned by below right=of <coordinate> is 3mm between top of ellipse (anchor=north) and coordinate and 7mm between given coordinate and left side of ellipse (anchor=west) .
 
@Zarko And you want TikZ to calculate the point on the ellipsis which has minimum distance to the given coordinate and set this distance to [7mm 3mm]^T? Don't you think this might be a huge overhead? Imho the behaviour of positioning is completely reasonable.
 
12:23 PM
@Skillmon, i'm aware that this is not simple ... however, than documentation for the positioning library is misleading. there should be clearly emphasized that library works only for rectangles ...
 
 
2 hours later…
2:28 PM
@JosephWright if I want to use the ctanreadme variable, in which list do I put the real file docfiles = {"CTANREADME.md",..} or textfiles?
@JosephWright sorry, just found it in the doc
 
2:40 PM
@UlrikeFischer textfiles
 
@JosephWright when running l2build install --dry-run --full the readme is not in the list, I hope it will be in the ctan build I'm just recreating ;-).
 
@UlrikeFischer Yeah ... renaming is currently being done after copying (I don't want to mess up files in the repo). I guess I could find a way to handle that in the dry run ... will think about it
 
@JosephWright it would be okay if it is mentioned with its original name.
 
@UlrikeFischer I'll ponder it!
 
@JosephWright but the result looks ok. I now pushed to the repository and when I'm back from chess will upload to ctan until someone cries ... @DavidCarlisle.
@JosephWright @DavidCarlisle if you want to test: as the fontloader changed its name (date) you need to run mktexlsr.
 
2:50 PM
@UlrikeFischer Great: so l3build is good-for-you at present? Next I guess I need to integrate CTAN upload (@DavidCarlisle)
@UlrikeFischer Right, noted
@UlrikeFischer Decisions about where to host ...
 
@JosephWright I was fighting a bit with two files which are in src in the repo but shouldn't go to tex. I managed to get them into source but doc would be better.
@JosephWright later ...
 
@UlrikeFischer :)
@UlrikeFischer Details? I'll take a look and see what needs to be done
 
@UlrikeFischer Thanks! I'll try to test this evening (but push to ctan whenever you are ready:-)
 
fontloader-reference-load-order.tex and fontloader-reference-load-order.lua, they are imported so I can't push them easily around, they are not functional but I would like to keep them for "info" around. (copies of the original plain-context fontloader).
 
@UlrikeFischer Right: I'll think about it and see what I can do
 
 
3 hours later…
5:50 PM
@boycott.se-yo' A jewel on Math.SE: $b\mathrel{{=}\llap{/}}0$
2
 
 
2 hours later…
7:30 PM
Hello guys!
How can we write double quotes in math mode but without curves? E.g. \[``No"\]. I mean with vertical double quotes, i.e. with "
If I write \[="Example"\] the output is not elegant: \[=" Example "\]
I mean I don't want “No” but "Yes"
 
@manooooh Why aren't you using text mode : \[ =\text{``No''}\] ?
 
@marmot I'm testing $$=\text{``No''}$$ in math.SE's MathJax but it doesn't work, the output is = ``No''
 
@manooooh you should have said you wanted a mathjax answer, asking here people would normally assume you were asking about tex.
 
@manooooh Seems like MatJax doesn't take no for an answer. ;-)
 
@DavidCarlisle ok. I thought MathJax was going to react in the same way...
@marmot sorry for the misunderstood
 
7:41 PM
@manooooh not really. (That's why it isn't enabled on this site, it would be confusing)
 
@DavidCarlisle what is not enabled on this site?
 
@manooooh mathjax
 
@DavidCarlisle oh ya. Forgot that. I want to see a processing of single quotes in titles :)
 
@manooooh but that's unrelated to mathjax that is just an issue of their markdown processing
 
@DavidCarlisle I still think it is the same problem when they had to implement MathJax in titles (by the way I mention it :))
 
7:46 PM
@manooooh you said that before but I don't know why, there is absolutely no connection.
 
@DavidCarlisle I take the opportunity to launch the anger haha. But never mind, it's totally unrelated
@DavidCarlisle and don't think that MathJax will be confusing here; I would not put it to replace images that show the problems, but for minor changes
 
@manooooh but anyway I don't think Unicode has a straight quote symbol other that ' and I'm not sure you can stop mathjax from interpreting that as a prime
 
@DavidCarlisle ok, so atm there's no solution for MathJax. Thanks!
 
@manooooh but people would, and when showing tex markup but not getting the backtick right instead of being readable but without the grey background you would see mathjax attempt to render it, it would be a disaster
 
@DavidCarlisle if we create a guide with a GIGANT DESCRIPTION that says "DON'T USE MATHJAX FOR REPLACING IMAGES" people will understand
 
7:51 PM
@manooooh no
 
@DavidCarlisle yes
Is my thought
 
@manooooh i won't argue with a duck in a football shirt
 
@DavidCarlisle hahaha good point! I'm just a duckling in the water
 
@UlrikeFischer ooh
oh no, I'm turning into Paulo:(
 
7:57 PM
@DavidCarlisle good team work ;-) I'm writing the bug reports and he is finding the solutions.
 
@UlrikeFischer meanwhile I've had to go back to 1.07 for a bit (trying to get some updates into latex and need a passing test suite:-) Hopefully I'll swich back to 1.09 later this evening
 
8:16 PM
@DavidCarlisle uploaded lualibs and luaotfload ...
 
@UlrikeFischer congratulations, and thanks again for picking them up!
@JosephWright
 Check failed with difference files
  - ../build/test-config-TU/tu-composites01.luatex.diff
  - ../build/test-config-TU/tu-tl2e7.luatex.diff

Failed tests for configuration build:

  Check failed with difference files
/usr/local/texlive/2018/bin/x86_64-cygwin/l3build:144: attempt to concatenate local 'failed' (a table value)
 
@DavidCarlisle I would say the time for the champagne is at the end of the week when there was no need for a fast update ;-) (But I already found a typo in the documentation ;-().
 
@UlrikeFischer As far as I understand, documentation is supposed to have typos
 
8:42 PM
@DavidCarlisle ;-).
 
8:53 PM
@DavidCarlisle Eerk: I'll sort that
 
@JosephWright of course it would be better if it hadn't failed, then it wouldn't have hit that:-)
@JosephWright down to last two or three failing if luatex is 1.07
 
@DavidCarlisle :)
 
@JosephWright @UlrikeFischer should .luc file loading be normalised away or should I try to pre-load the font in anOMIT\TIMO region within the lvt??
*** ../build/test-config-TU/tu-composites01.luatex.tlg	2018-09-25 21:15:33.071914000 +0100
--- ../build/test-config-TU/tu-composites01.luatex.log	2018-09-25 21:15:45.737877500 +0100
***************
*** 18,24 ****
  TESTING \^
  TESTING \~
  TESTING \textcommabelow
! (load luc: ../lmroman5-regular.luc) [1
  Missing character: There is no ̧ (U+0327) in font [lmroman10-regular]:+tlig;!
  Missing character: There is no ̧ (U+0327) in font [lmroman10-regular]:+tlig;!
  Missing character: There is no ̧ (U+0327) in font [lmroman10-regular]:+tlig;!
@JosephWright the other remaining diff is tu-tl2e7 which also has .luc diffs and (font)
***************
*** 26,32 ****
  .....\TU/lmr/bx/n/14.4 1
  .....\glue 14.4
  ....\TU/lmr/bx/n/14.4 T
! ....\kern-1.35359
  ....\TU/lmr/bx/n/14.4 e
  ....\TU/lmr/bx/n/14.4 s
  ....\TU/lmr/bx/n/14.4 t
--- 25,31 ----
  .....\TU/lmr/bx/n/14.4 1
  .....\glue 14.4
  ....\TU/lmr/bx/n/14.4 T
! ....\kern-1.35359 (font)
  ....\TU/lmr/bx/n/14.4 e
  ....\TU/lmr/bx/n/14.4 s
  ....\TU/lmr/bx/n/14.4 t
is the (font) an l3build change (no longer normalised away)
 
9:19 PM
@DavidCarlisle no, it is the new fontloader.
 
@UlrikeFischer ah, but the old one wrote (font) in some places didn't it?
@UlrikeFischer but in that case I can update as the new font loader will hit texlive and so travis in a day or so?
 
@DavidCarlisle I don't remember for sure, I had them and sometimes not. But on the whole it is so that the newer fontloader has more of them.
@DavidCarlisle I got the notice that it has been installed on CTAN.
 
@UlrikeFischer OK thanks.
 
@UlrikeFischer hmm
@UlrikeFischer meanwhile helpful message on the mailing list I see:-)
 
@DavidCarlisle that's why I made a context example ;-) (I also sent it Marcel, gmane doesn't let my messages to the list pass.)
 
@UlrikeFischer @DavidCarlisle Hans does not want to change the engine behaviour, so I reimplemented font.each in Lua avoiding this. I will create a PR in a moment, but do you prefer a new name to keep the old font.each or should we just overwrite it?
 
@MarcelKrüger I suspect that context has a such a overload for font.getfont and as soon Hans realize that font.each has the same problem he will add one for this function too. See mailman.ntg.nl/pipermail/ntg-context/2013/072166.html.
 
@MarcelKrüger I'd be wary of overloading core lua functions (or tex primitives) it makes reading the luatex manual while using latex tricky. If hans does it (and the luatex manual updates to match) then things are different...
@MarcelKrüger but overloading is better than reading the manual and having the system segfault:-)
 
10:12 PM
@MarcelKrüger font-def.lua contains
function font.getfont(id)
    return fontdata[id] -- otherwise issues
end
 
@DavidCarlisle That's the idea. The manual would not need updating, it just allows actually using the documented features. Anyway I agree that overloading core function should be avoided, let's see what happens in the fontloader. I uploaded an experimental implementation of font.each at gist.github.com/zauguin/925c12e156f2ab94f6660464e8efcb87 just in case someone needs a quick fix.
 
@MarcelKrüger will you sent another message to the list regarding the context example?
 
@UlrikeFischer I will do so. I wanted to follow up there anyway.
 
@MarcelKrüger actually I might need that if get the latex test suite to pass with 1.07, I 'll switch back to 1.09 when accessing the font.each seems pretty scary.. But you may yet get changes lower down, if you push a bit:-)
 
@DavidCarlisle if Hans adds it to the fontloader i should be able to make an update rather fast, now that I have sorted out all the pulls and imports (and I could correct the typos ;-)).
 
10:23 PM
@UlrikeFischer You can just use your normal mailer, you don;t have to go via gmane (non subscribers can post, I think)
 

« first day (2886 days earlier)      last day (2046 days later) »