TeX, LaTeX and Friends

David Carlisle

00:15

@PauloCereda clearly a fake, there is writing on the page

G. Bay

@DavidCarlisle LOL, but he was fast on that

CarLaTeX

05:01

@DavidCarlisle @HaraldHanche-Olsen, don't listen to David, I was on holiday, too, and I didn't discuss ducks during the last two weeks :):):)

Will Robertson

05:12

@JosephWright — hmmm, there's an idea :) tex.stackexchange.com/questions/432766/…

Skillmon

06:59

Do I understand correctly that \peek_meaning:NTF is the l3 equivalent of \@ifnextchar?

Joseph Wright

07:12

@Skillmon Not quite: you want \peek_meaning_ignore_spaces:NTF - we are currently looking at some issues concerning spaces and what 'ignore' means ...

Skillmon

@JosephWright thank you very much!

Skillmon

07:41

@JosephWright Regarding argument naming convention: If I write a macro that reads two consecutive weird delimited arguments, should the argument part of the name be ww or w? E.g. should it be \cs_new:Npn \foo_bar:ww #1 - #2 \q_stop or \cs_new:Npn \foo_bar:w #1 - #2 \q_stop?

Joseph Wright

@Skillmon Just :w - once an argument is non-standard, there is no pattern one can impose

Skillmon

@JosephWright ok. I think it is more intuitive to write :ww because it has two arguments, not one, but I obey the conventions:)

Joseph Wright

@Skillmon We've had this discussion on the team :)

Skillmon

@JosephWright Imho, then the discussion had the wrong result and I should have joined the team just for that discussion :)

Joseph Wright

@Skillmon Frank is pretty clear: the point of w is that the argument grabbing is non-standard, so can be anything with no necessarily link to how many things are grabbed. A lot of TeX primitives, for example, are w.

@Skillmon Problem is how do you tell from the signature where the arguments start and stop? You can't ...

Skillmon

07:49

@JosephWright I know (the latter part), and as I said, I'm going to use the convention. Arguments for both sides are plausible to me, but the non-convention would have been more intuitive for me personally.

@JosephWright is \group_begin: exactly the same as \bgroup, so I could check whether there will be a group with \peek_meaning_ignore_spaces:NTF \group_begin: { <true> } { <false> }?

Joseph Wright

@Skillmon \group_begin: is \begingroup (we don't use braces for grouping in code). You are looking for \c_group_begin_token.

David Carlisle

08:16

@JosephWright 88 commits and counting:-) how's the markdown doing, I was half watching the commits but haven't tried it...

Joseph Wright

@DavidCarlisle More-or-less done

@DavidCarlisle I think most of the remaining oddities are best handled by hand

@DavidCarlisle A few things to decide on: do we keep &nbsp, do we use &mdash or swap to ---, how do we mark up \meta by default

@DavidCarlisle I'm keen to remove the stray empty lines, but I'm not finding that so easy ... may leave it for a second pass with a simple 'do cleanup' script

@DavidCarlisle I'm checking over by hand now: once a decision is made on \meta (which at present is probably go with what there is now), I'll commit the .md files and start cleaning up by hand. For example, there are things like \csx{`} that need fixing, but occur only once in the entire source.

David Carlisle

@Skillmon I think if everything had been a macro and then www told you you could use #1 #2 #3 that may have been the plan, but many things with :w aren't macros and don't have #1 eg \if_catcode:w

@JosephWright that is the downside of markdown, it is rubbish as a markup language compared to latex or html:-)

@JosephWright who makes the decision by the way is it still uktug committee or is it time to relieve them of that duty?

Skillmon

@DavidCarlisle it is faster to be input than LaTeX if you're equally skilled in both of them. You can do much more stuff in LaTeX or HTML, but if I protocol a discussion, I usually use markdown, as I don't need the advanced stuff of LaTeX or HTML but the speed.

David Carlisle

@Skillmon yes but since the document in question is averaging less than 1 edit per year speed of typing isn't a major consideration here:-)

Skillmon

@DavidCarlisle that's a valid point you made there :)

Joseph Wright

08:27

@DavidCarlisle Practically, it's down to people doing the work :)

egreg

9

Q: Some of them would say I'm half a duck

Some of them would say I'm half a duck I may be used if you want to try your luck I inspired your sweetest dreams, will say the most accurate But for only 15 of them, it was worth a rate Some of them tend to collect me Even if I'm usually just worth a few penny Ilia L found ...

David Carlisle

@JosephWright yes but here I wonder if it should be formally handed over to a "github tex-faq organisation" (or something) so control could include people who are not on the committee (me, Karl, @egreg, ...)

Joseph Wright

@DavidCarlisle Good question: the text is basically Robin's so the link to UK-TUG is pretty tenuous anyway

@DavidCarlisle I might check in the Markdown at this stage: I think the script does as much as is sensible. Any further tidy-up is likely as fast done by editing the Markdown as anything else.

@DavidCarlisle My plan then is to clean it all up before squashing back into the gh-pages branch and actually getting something working

@DavidCarlisle I'll email

David Carlisle

@JosephWright if you wanted to retain the styling you could generate inline html say <span class="meta">... and style it globally, although lots of inline html makes hand editing the markdown in future a bit more weird

Joseph Wright

@DavidCarlisle Quite: I'm thinking we want basically 'vanilla' Markdown

David Carlisle

08:41

@JosephWright yes so long as that isn't too restrictive (a lifetime of working on markup languages and in the end it comes to this: the most complicated markup we are allowed is *foo* :)

Joseph Wright

@DavidCarlisle Sure

@DavidCarlisle The content of the FAQ is pretty simple in formatting terms: I've kept things like   precisely because that is easy

@DavidCarlisle I've checked in the Markdown files: I'll do some cleanup later today, but hopefully we won't need to rerun the scripts

David Carlisle

@JosephWright OK, thanks

Joseph Wright

@DavidCarlisle There are a lot, but I hope I can get through them quite quickly to do fixes

@DavidCarlisle I'll probably do them in batches of 10-20, alphabetically. I've already spotted a few oddities, for example places where we probably want blockquotes. Shouldn't take too long.

David Carlisle

@JosephWright is there any markdown validator anywhere or is it just a case of sticking it on gh and seeing if it looks OK?

Joseph Wright

@DavidCarlisle That's my plan :)

@DavidCarlisle GitHub use there own 'flavour' of Markdown (with the three backtick and code highlighting keyword business)

Skillmon

08:55

@DavidCarlisle what does \CT@arc@ do in \@cline do?

David Carlisle

@JosephWright 1998 was the dawn of a new era, the world would be free of unregulated markup, everyone could use xml and get a helpful fatal error if they put a foot out of line. whatever happened?

2

@Skillmon CT@ = colortbl internal, arc = arrayrulecolor

Skillmon

@DavidCarlisle thx. Just found it in the documentation (it get's let to different things so many times that I first didn't find the definition inside \arrayrulecolor->\CT@arc).

David Carlisle

@Skillmon I'm upset you don't find my command naming system self documenting and intuitive, @JosephWright would have used something really obscure like \colortabl_sometimes_locally_holds_the_arraycolor_settings_when_its_not_relax:

3

Joseph Wright

@DavidCarlisle People

David Carlisle

@JosephWright I'll be back....

Skillmon

09:03

@DavidCarlisle since colortbl is not in l3 I guess that name should be \colortbl@sometimes@locally@holds@the@arrayrulecolor@settings@when@its@not@rela‌x

egreg

@DavidCarlisle You miss the fact it is a variable, rather than a function, so it should be \l__colortabl_sometimes_locally_holds_the_arraycolor_settings_when_its_not_relax‌_tl

Much more informative and with three more underscores!

David Carlisle

09:24

@egreg better perhaps would be \__keep_out_of_my_stuff:D

Matsmath

Hello! It looks like that the automatic EPS-TO-PDF converter, creating files such as file-eps-converted-to.pdf is buggy. I have a simple image with a few points and lines which gets converted incorrectly. Is there a standard repository (like imgur for images) where I could upload this sample eps file so that others may inspect it?

Paulo Cereda

It's very cold today!

David Carlisle

@Matsmath no if you think it's ghostscript bug then you could report it there otherwise you can put the file on any public server (github, dropbox, whatever) and ask on stackexchange (but stackoverflow rather than this site it would be off topic here I think)

@PauloCereda -40C?

Skillmon

@DavidCarlisle is that the SI definition of "very cold"?

Matsmath

Ah, I see, so it is a sort of 3rd party plugin. OK, I will look into it, thank you!

Skillmon

09:31

@DavidCarlisle Or is it something related to the imperial system and you converted it for us?

Paulo Cereda

@DavidCarlisle not that cold. :)

@DavidCarlisle, @Skillmon ^^ :)

David Carlisle

@Skillmon conversion for -40 is particularly easy :-)

@Matsmath it just runs the epstopdf script on the eps to make a pdf file (and epstopdf is a very thin wrapper around ghostscript) so latex isn't really involved at all other than including the generated pdf.

Matsmath

Sure thing, it is all clear now. Thank you!

Skillmon

10:14

Is there any possibility to find out whether the current item in a \clist_map is the last one?

I currently use \int_set:Nn \l_tmpa_int { \clist_count:n { #1 } } \clist_map_inline:nn { #1 } { \int_decr:N \l_tmpa_int ... }

egreg

@Skillmon Are you sure you don't want to do \clist_use:Nn?

Skillmon

@egreg You mean something like \int_set:Nn \l_tmpa_int { \clist_count:n { #1 } } \int_step_inline:nn { \l_tmpa_int } { \clist_use:nn { #1 } { ##1 } ...? Is that faster? (it doesn't need to be expandable)

egreg

@Skillmon I mean \clist_set:Nn \l_tmpa_clist { #1 } \clist_use:Nn \l_tmpa_clist { material between items }

@Skillmon There's also \clist_use:Nnnn for finer control.

Skillmon

@egreg not possible, because I need to extensively parse the contents of each clist item. That's why I map to it

egreg

10:31

@Skillmon I'd convert the clist to a seq: \set_seq_from_clist:Nn \l_tmpa_seq { #1 }, then pop the last element \seq_pop_right:NN \l_tmpa_seq \l_tmpa_tl, do \seq_map_function:NN \l_tmpa_seq \skillmon_process:n and finally do \skillmon_process_item:V \l_tmpa_tl

@Skillmon \skillmon_process_item:n should do the parsing of the item as you wish; you can define \skillmon_process:n to call the previous function and then add the material in between items.

@Skillmon Generally, sequences are more flexible than clists

Skillmon

@egreg the version I currently have uses my clist approach with a decreasing counter and does work pretty well. I can't use the map_function ones without further complicating stuff, because I need to give an argument to the parsing function. So in fact I use something like \clist_map_inline:nn { #2 } { \skillmon_parse:nw { #1 } ##1 \q_stop }

egreg

@Skillmon Not what I'd do, but it depends on the target.

Skillmon

10:47

@egreg perhaps I'll show the code later this day or tomorrow. I'm trying to fix some issues with \cline and \rowcolor while automatically format some rows as head rows (\hline above and below the block of head rows, each row with \rowcolor{gray}). There is one thing currently not working which I need to solve and at the moment no key=value interface, which I plan to implement.

egreg

11:01

@Skillmon I never use colored tables /harsh voice à la Bela Lugosi

Skillmon

@egreg I never use colored tables if I have the choice, but since my Prof. doesn't share my typographical taste (BAD TYPOGRAPHY EVERYWHERE (which I wrote completely in uppercase letters to clarify that)).

David Carlisle

4

Paulo Cereda

11:18

@DavidCarlisle ooh

David Carlisle

@PauloCereda I think he'll be especially proud of the font use.

Paulo Cereda

@DavidCarlisle fantastic

CarLaTeX

11:35

@barbarabeeton This link ^^^ gives this vvv

Paulo Cereda

12:06

@CarLaTeX tug.org/TUGboat/tb39-1 exists, so I think it's just a wrong link. :)

Joseph

12:20

Daily joke for @PauloCereda: tex.stackexchange.com/questions/432740/… "My crystal ball is on holiday".

Paulo Cereda

@Joseph LOL

Joseph

I've seen the results of using German-style sarcasm south of the Equator.

;-)

Paulo Cereda

@Joseph oopsie

I had an HP 10s+ with me yesterday...

Someone said, "that's a strange cellphone".

/facepalm

David Carlisle

@PauloCereda I am sure you can get an HP simulator app for your phone:-)

Paulo Cereda

@DavidCarlisle :)

@DavidCarlisle perhaps I could write a phone emulator in this calculator too? :)

mickep

12:39

@PauloCereda You should port TeXLive to it!

Paulo Cereda

@mickep ooh

@mickep funny enough, I once managed to install TL 2009 in a Debian install running on my Wii console!

mickep

@PauloCereda Cool, did you write your thesis, dancing using the Wii controls?

Paulo Cereda

@mickep it would be awesome, but no. :)

David Carlisle

@mickep it is true to say that all the pages of @PauloCereda's thesis that have been written were written on a Wii

CarLaTeX

13:05

@PauloCereda Thank you, however the broken link should be corrected, that's why I told Barbara :):):)

TeXnician

@PauloCereda I found yet another weird JDK at Gitlab CI :D

Joseph Wright

@DavidCarlisle Working :)

David Carlisle

13:23

@JosephWright so it is:-)

Paulo Cereda

@TeXnician ooh I have interesting news for you! In a couple of minutes... :D

Joseph Wright

13:52

@DavidCarlisle I think almost everything actually works!

@DavidCarlisle Looking over the questions, we need some cleanup ...

@DavidCarlisle I wonder if I should say that we'll not commit directly to master (after this initial set up): we can have a 'staging' branch and then give version numbers to the FAQ ...

@PauloCereda, @StefanKottwitz As you are interested too: tex-faq.github.io

@DavidCarlisle I get a feeling this is a lot easier to edit than the old scheme ...

Paulo Cereda

@JosephWright awesome

Joseph Wright

@PauloCereda Looks good, doesn't it?

@PauloCereda I plan to clean up a bit more, come up with some guidelines for changes, then look into getting tex.ac.uk redirected

Paulo Cereda

@JosephWright it does!

Joseph Wright

@PauloCereda :)

@PauloCereda The content needs a good overhaul: as I've noted, I think that might be best done using a formal second branch and a logged merge approach (@DavidCarlisle?)

David Carlisle

14:14

@JosephWright yes I saw some fixltx2e etc...

Joseph Wright

@DavidCarlisle Exactly, plus inputenc, stuff on engines, etc., and still has material on ANT!

@DavidCarlisle I'd like to do a 'purely mark-up' sweep first

David Carlisle

@JosephWright do you need a staging branch or just do it via prs from branches in forked copies?

Joseph Wright

@DavidCarlisle Still wondering about \meta: making it just <_foo_> or even _foo_ is easier

David Carlisle

@JosephWright yes

also the google search goes to my fork

Joseph Wright

@DavidCarlisle Don't think you need a staging branch, not sure how PRs go in terms of building (we just use master as the live branch)

@DavidCarlisle In other news: tex.stackexchange.com/questions/432833/…;. I think I'll have to strip back out ISO dates in expl3 ...

David Carlisle

14:18

@JosephWright why do people do that rather than doing a miktex or texlive update?

TeXnician

@DavidCarlisle Because TeX distros obviously do lack a big button with "Click here to update, every other way will destroy your system"…

David Carlisle

@JosephWright if you backed out the iso format dates wouldn't it just die somewhere else if you update expl3 by a couple of years without updating any of the packages using it?

@JosephWright I wonder if the title ought to be "The English Language TeX FAQ" rather than "The English TeX FAQ"

Joseph Wright

@DavidCarlisle Perhaps 'The (English Language) TeX Frequently Asked Questions'?

@DavidCarlisle At some stage yes, but I'm mindful of the Linux distro approach to TeX system update

Ulrike Fischer

6

@samcarter @PauloCereda ^^^^

\documentclass[parskip=full]{scrartcl}
\font\duckchessfamily = "file:luafont-duckchess.lua:mode=node;+liga;+tlig;original=latinmodernsans" at 20pt
\begin{document}

\sffamily
We proudly preset the duck chess font:

{\duckchessfamily
1. Nf3 nf6 2. e4 e5 3. Bb5 lb4 4. Qe2 qe7 5. Rg1 rg7 6. Kf1 kf7
}

\end{document}

Paulo Cereda

14:33

@UlrikeFischer oooh

samcarter

@UlrikeFischer ohhh

Ulrike Fischer

@JosephWright you could add a moment of surprise by writing "The (French Language) TeX Frequently Asked Questions" ;-)

David Carlisle

@UlrikeFischer Je pourrais répondre à toutes les questions françaises

Ulrike Fischer

@DavidCarlisle j'en suis sure ;-).

Steven B. Segletes

@DavidCarlisle As a question of historical development, do you know when the geometry package added support for \newgeometry or if it came with version 1.0? I couldn't find any development history in the documentation.

@DavidCarlisle I have a user's log file complaining about \newgeometry and I found that their version of the package is from 2002 (v3.2), if that helps narrow it down. Basically, I would love to hear, "it was added after 2002." That would provide me an easy course of action with the user. Otherwise, I have to figure out why a version that should support it, doesn't.

Ulrike Fischer

14:43

@samcarter neat isn't it? It is a pity that I can't use "b" for the black bishop, but I see no good way to avoid it, and probably one should anyway define a "useblack" "usewhite" option.

David Carlisle

@StevenB.Segletes I have no idea:-)

Steven B. Segletes

@DavidCarlisle Thanks.

David Carlisle

@StevenB.Segletes get @egreg to grep for newgeometry (he had texlive installed for years going back to around then)

Joseph Wright

@StevenB.Segletes Check the TL SVN? Goes back a reasonable way, though perhaps not to 2002

David Carlisle

@StevenB.Segletes vvv

samcarter

14:46

@UlrikeFischer Are the colours hard coded in the font?

David Carlisle

2010-02-12 Release 5.0
  * Changing page layout mid-document.
    The new commands \newgeometry{...} and \restoregeometry allow

@StevenB.Segletes the package has changes.txt ....

Steven B. Segletes

@DavidCarlisle Awesome! Thank you so much.

David Carlisle

@samcarter sorry wrong completion on @s :-)

Ulrike Fischer

@samcarter Yes, the chess figures are actually pdf-images.

Stefan Kottwitz

@JosephWright Don't forget about the new EU GDPR for this web service including data processing by Github, run by a society and not a private hobby person anymore ;-)

Joseph Wright

14:48

@StefanKottwitz Luckily, there is no personal data at all here (have done GDPR training for work)

Stefan Kottwitz

@JosephWright perhaps (logging of IP addresses? user personal data registered for the back end?)

@JosephWright can be github responsibility but they would refer to the clients using github

samcarter

@UlrikeFischer You should propose to the international chess organisation to switch to yellow vs. black - otherwise they may get confused if suddenly all chess players want to use your cute new font!

Ulrike Fischer

@samcarter oh chess player are quite relaxed regarding colors. One of the first things I explain to beginners is that the field are named black and white even if they are actually brown and beige ;-).

mickep

Do the "duckhorse" walk as the usual horse?

Joseph Wright

@StefanKottwitz You are the expert here, but I didn't thing IP addresses per se were covered. In any case, as a user of GitHub I can't access the data, so I don't think it is an issue.

@StefanKottwitz We've had a discussion on the team about IPs: if they are included, basically every website in the world is in trouble. The feeling was that they don't seem to be.

Stefan Kottwitz

14:56

@JosephWright Every website is in trouble until things are made clear, people don't feel safe for now and wait what happens.

@JosephWright Just thoughts. I did some GDPR preparations that I will put online and hope for the best. Luckily I don't have any ads etc. so can go as a private person and maybe commercial "competitors" could send complaints but who shall this be.

Joseph Wright

@StefanKottwitz Like I say, as a GitHub user one can't access the data at all, so I don't think it's a significant concern.

samcarter

@mickep Depends if the duck horse is on land or on water :)

mickep

@samcarter Oh, I love new rules!

Stefan Kottwitz

@JosephWright Sure! Just mirroring a bit what I read everywhere about people having fear of what may start in a few days. But this may affect online shops etc. that actually works with personal data.

samcarter

@UlrikeFischer Oh, good. One thing less to worry about :)

Joseph Wright

15:00

@StefanKottwitz Exactly, and even there the rules allow you to hold data where you are actually using with permission: you'd have fun running an online shop with no addresses to post to!

G. Bay

lol

Joseph Wright

@StefanKottwitz Like you, my personal blog has no ads or similar, so I think I'm also OK

Stefan Kottwitz

@JosephWright A statement about Privacy protection would be good (or needed?) even if it just claims that no data is stored or processed (could be a formally required information to visitors)

egreg

15:42

@StevenB.Segletes 2010/03/13 v5.3 Page Geometry

@StevenB.Segletes Actually it was added in version [2010/02/12 v5.0 Page Geometry]

David Carlisle

57 mins ago, by David Carlisle

2010-02-12 Release 5.0
  * Changing page layout mid-document.
    The new commands \newgeometry{...} and \restoregeometry allow

I believed the documentation:-)

egreg

@StevenB.Segletes @DavidCarlisle tug.org/svn/texlive/trunk/Master/texmf-dist/tex/latex/geometry/…

Ulrike Fischer

@DavidCarlisle Oh, did you read it?

David Carlisle

@UlrikeFischer only the changes.txt not the actual package doc, of course.

StrongBad

16:30

I feel like I read somewhere that once the text is finalized that it is sometimes necessary to do some final small tweaks to get line breaks and the such just right (sometimes going as far as rewording a difficult sentence/paragraph). Is there anything "authoritative" on the subject (maybe Knuth wrote something)?

David Carlisle

@StrongBad Knuth wrote something and also Frank wrote quite a bit at the back of the latex companion detailing how many pages he manually adjusted

@StrongBad texbook touches on that in quite a few places eg: Therefore if you are fussy about the
appearance of pages, you can expect to do some rewriting of the manuscript
until you achieve an appropriate balance, or you might need to fiddle
with the ^|\looseness| as described in Chapter~14; no automated system will
be able to do this as well as you.

StrongBad

@DavidCarlisle thanks. that will do. Time to go ask a question on academia ...

percusse

16:46

ello' @DavidCarlisle, @egreg Can we get a quick L3 answer to this one ?

3

Q: \pgfplotsforeachungrouped: how to escape & and \\

Disclaimer: I've barely started to learn LaTeX so bear with me. Problem: I use a lot of matrices like this Which I absolutely hate to code manually. Obvious solution would be a nested for loop, something like this: \begin{equation} A = \begin{bmatrix} \pgfplotsforeachungrouped \m in {1,...,3...

matrices loops foreach branching

I think pgfplots is an overkill there

StrongBad

17:05

For those of you who actually know how publishers work:

0

Q: How often do publishers change/tweak articles to achieve visually appealing results?

When reading this question it seems the theoretical answer is that LaTeX is not perfect and manual/human intervention is required. In the TeX Book Knuth says: But the problem of page make-up is considerably more difficult than the problem of line breaking that we considered in the previous ...

journals formatting

I am curious about how often they tweak the desired layout and/or the author's material to get good results

G. Bay

now thats a good question

Steven B. Segletes

@egreg Thank you. Now all that I need to find is how the user, who could compile code with \newgeometry until recently, all of a sudden ends up with a 2002 version of the package. The unexpected "update", I presume.

David Carlisle

@percusse I suspect it's a dup of this, to which @egreg has given an l3 answer

4

Q: Use array-meaningful ampersand to change column in foreach loop

I wish to loop over this code: \begin{tabular}{l*{8}c} first column & column a & column b & column c & column d & column e & column f & column g \end{tabular} But the following does not compile: \begin{tabular}{l*{8}c} first column \foreach \i in {a,b,c...

tables loops ampersand

egreg

17:30

@percusse Done

percusse

17:42

@egreg @DavidCarlisle Thanks to both of you.

@StrongBad In this age, it is all due to bad academician habits. The worse they write the more copy editors have to tweak

@yo' should know better

Joseph Wright

17:59

@DavidCarlisle FAQ sweep going OK: there are a lot of out-of-date answers

barbara beeton

18:17

@CarLaTeX -- reported to tug webmaster already, by claudio beccari. karl will probably get to it reasonably quickly.

@JosephWright -- i think i had better take a look at some of these. i don't trust myself to do actual updates, so to whom should i send the updated text?

Paulo Cereda

@barbarabeeton speedy Italians!

barbara beeton

@UlrikeFischer -- you might want to change "preset" to "present", to avoid somebody trying to "reset". (delightful creatures indeed!)

David Carlisle

@JosephWright yes what's the plan, standardise the markup then divy out updating the texts?

Joseph Wright

@barbarabeeton Give me a little while to establish an approach :)

@DavidCarlisle Exactly

@DavidCarlisle I'm hoping to sort out the formatting today

David Carlisle

@JosephWright is jekyll running? I thought you could have just gh-flavoured markup without jekyll if you didn't want the templating stuff running over it?

Joseph Wright

18:30

@DavidCarlisle Yes: GitHub pages runs automatically, as far as I can tell

@DavidCarlisle We will want footers and similar at some stage

David Carlisle

@JosephWright yes I just checked some other pages i have, i mis-remembered

@JosephWright yes

Joseph Wright

@DavidCarlisle I'm working out how much tidy up to do: quite a lot of the pages with issues probably need updating anyway

CarLaTeX

@barbarabeeton Thanks! Claudio liked prof. van Duck's first article very much, he asked me to write an Italian version of it for ArsTeXnica :):):)

barbara beeton

4

and got another palindrome.

Paulo Cereda

@barbarabeeton yay

@CarLaTeX Wacky Italians <3

CarLaTeX

18:36

@PauloCereda :):):)

barbara beeton

@CarLaTeX -- that's a real nice invitation. do take him up on it!

CarLaTeX

@barbarabeeton Surely!

Ulrike Fischer

@barbarabeeton Ups ;-)

David Carlisle

@CarLaTeX hai bisogno di aiuto con la traduzione?

CarLaTeX

@DavidCarlisle Sono sicura di poter far affidamento su di te per quello!

David Carlisle

18:49

@CarLaTeX Potete crederci? Ho solo un distintivo in bronzo per babel!

CarLaTeX

@DavidCarlisle Non ci posso credere, con la tua padronanaza delle lingue, meriti molto di più!

David Carlisle

@CarLaTeX It's true, I do!

barbara beeton

@UlrikeFischer -- hey, your english is soo much better than my german ... i figure that was just a finger glitch. (on another topic, i like the "cracked crystal ball" image. i actually do have a sphere of rutilated quartz. that's the perfect crystal ball for me -- very pretty, but totally unreliable regarding clarity of answers.)

CarLaTeX

@DavidCarlisle :):):)

Paulo Cereda

@DavidCarlisle Sono italiano di Manchester.

David Carlisle

18:53

@CarLaTeX having mastered Italian and Portuguese, now I only have to work on that strange language spoken by @barbarabeeton

CarLaTeX

@DavidCarlisle LOL

Paulo Cereda

@DavidCarlisle xii-bnb.tex?

CarLaTeX

@DavidCarlisle An American friend of mine married a lady from Manchester and she complained that her mother-in-law didn't understand her!

Joseph Wright

19:08

@DavidCarlisle, @barbarabeeton First step in revising the FAQ I think is to remove the various entirely out-of-date answers ...

David Carlisle

@JosephWright depends a bit how much you worry about incoming links from various tutorials and the visual faq etc. If it wasn't for that and you just deleted out of date stuff then a large chunk would go...

Joseph Wright

@DavidCarlisle Take a look over e.g. tex-faq.github.io/#current-tex-related-projects: there are a lot of really old things there

David Carlisle

@JosephWright yes, I would say more than half of the answers I have looked at over last couple of days have been out if date to the extent that they are basically just wrong.

Joseph Wright

@DavidCarlisle Exactly

barbara beeton

@JosephWright -- a lot of answers are out of date for sure, but are the topics still relevant? if the topics are still solid, possibly better to mark them as "under construction" and solicit assistance in updating.

Joseph Wright

19:14

@DavidCarlisle I think we do need to look over all of them, if we want them to be useful

@barbarabeeton I'm thinking of answers such as tex-faq.github.io/FAQ-omegaleph.html: it's hard to see that this should be in an FAQ, link stability or not

David Carlisle

@JosephWright yes, I wonder if you should put a visible "text last revised 2010" banner on them all (or whatever date makes sense) and then we can gradually update

@JosephWright could be replaced by a one liner that said omega was the first unicode extension, and work fed in to xetex and (especially) luatex. I'd keep the link though, if stuff needs to be dropped I think we could consider dropping it from the site navigation but leaving a page with minimal information and a pointer to the rest of the faq

Joseph Wright

@DavidCarlisle I was wondering whether that can be automated, but of course the files are all new so it wouldn't work

David Carlisle

@JosephWright well I would pick a date when work on the tex sources stopped and just put that on all of them.,

Joseph Wright

@DavidCarlisle Sure: there was a reason I'm thinking we should have a branch for changes, and merge them properly ...

barbara beeton

@JosephWright -- while that answer is not going to be of much current use, i know from current experience that it's pretty hard to locate good historical information. (writing a paper with karl berry and dave walden on how tex fits in with the history of desktop publishing; will appear in the ieee annals in two parts.) would suggest that for some topics, move them from the current organization into a section "topics of only historical interest".

Joseph Wright

19:19

@DavidCarlisle All doable

@barbarabeeton That would be reasonable

David Carlisle

@barbarabeeton that fits with what I just commented as well as "move to a historic section" can just be a matter of changing the site navigation without changing the page url

@JosephWright if we edit via a fork how would a staging branch work? A contributor fork from the staging branch and Pr to update that, then "you" (or some team admin) moves to master if it doesn't break stuff?

Joseph Wright

@barbarabeeton, @DavidCarlisle Sure, we can keep them 'out of the way': I'm mindful that 430+ questions is too many for 'FAQ', really

David Carlisle

@JosephWright and it would get more if you added any more current faq about unicode or opentype or tikz etc

Joseph Wright

@DavidCarlisle Something like that: I was mainly think about the need to have clear versions. We could though simply put the updated date in each page as we do them ...

barbara beeton

@DavidCarlisle -- that sounds like a good compromise. i'm willing to look at historical stuff. the answers were good at one time, so maybe putting a date showing period of relevance?

@JosephWright -- but since the information is curated, it would be more reliable than much of the latex wikibook. hence valuable, if not necessarily super-"frequent".

David Carlisle

19:24

@JosephWright yes that's what I meant date them all now with some old date (2000 is a round number:-) but then updat ethe date individually (or automatically via jekyll?) as they are edited for text (as opposed to markup) changes

Joseph Wright

@barbarabeeton Sure: that's the plan :)

David Carlisle

@barbarabeeton I think that argument is sadly spurious. Since the information is (or was) curated, it is wildly out of date and misleading or wrong in large part

Joseph Wright

@barbarabeeton I'm thinking a small number of 'well known' people can add most stuff

@barbarabeeton Of course

barbara beeton

@JosephWright -- how about a section "infrequent, but nonetheless useful"?

mickep

Infrequently Frequently Asked Questions

David Carlisle

19:27

@barbarabeeton "questions old people used to ask in their youth" ?

Joseph Wright

@DavidCarlisle I'll try github.com/gjtorikian/jekyll-last-modified-at: give me a few minutes

David Carlisle

@JosephWright my understaning was you couldn't use jekyll plugins other than the preconfigured github setup unless you run it locally

G. Bay

iFAQ sounds like another apple blowtware hahaha

Joseph Wright

@DavidCarlisle Could well be

barbara beeton

@DavidCarlisle -- not necessarily. i'm thinking that quite a few questions about tikz are about itsy-bitsy problems that don't come up often, but are real head-scratchers when they do, and aren't necessarily easy to find in the manual. hardly obsolete.

David Carlisle

19:30

@barbarabeeton oh you meant new stuff, I thought you meant a section to move some of the existing stuff, OK I see.

Joseph Wright

19:43

@DavidCarlisle, @barbarabeeton Date line added

@barbarabeeton, @DavidCarlisle I suspect I've done as much 'cleanup' of the sources as I feel like doing today. Might be worth starting on the 'real' issues: working through, identifying stale content, etc. There's also some work to do on layout: I suspect we need a header/footer ...

@DavidCarlisle, @barbarabeeton Oh yes, one 'editorial' thing: should I clean up single versus double quotes? It seems to lean toward the latter, but not consistently ...

David Carlisle

@JosephWright also I think I miss the breadcrumbs or some other indication of where you are, but to start on real issues do you want me to fork or get write access and edit directly or ... or experiment on a staging branch or... (meanwhile we have a patch level 5 to release....)

Joseph Wright

@DavidCarlisle It's all doable: probably we want to sort breadcrumbs by adding some variables and then modifying the Jekyll

@DavidCarlisle I'm not fussed by any particular workflow ... you should have write access so that's probably easiest. It's not like we want/need formal PR/review/merge cycles ...

@DavidCarlisle What is needed is some idea of who is doing what ... probably just a heads-up is what's needed at the moment

@DavidCarlisle, @barbarabeeton I guess I should sort a bit of header information for 'Historical questions and answers'

David Carlisle

@JosephWright yes but was holding off suggesting any text changes in case you wanted global edits to the markdown which will get less consistent once edited by hand, but but then I could experiment with updating some that are definitely in my sphere like that mathml one...

Joseph Wright

20:01

@DavidCarlisle One reason for going for Markdown is that it's pretty easy to be consistent: not a lot of power to do different stuff

barbara beeton

@DavidCarlisle -- well, i'm talking about both, but obviously not being clear about which is which. (having spent the night before last in the charlotte airport when thunderstorms disrupted our connection. my head still isn't on straight.)

David Carlisle

@JosephWright sure but currently alll idioms coming from the tex are exactly the same, so if you decided to change it you could easily do a regexp replace rather than re-run teh convertor, but once hand edits are allowed there will be differences, and if you allow wysiwyg editing in the github web interface then all kinds of invisible stuff gets added

barbara beeton

@JosephWright -- can you identify a couple of each (single vs. double quotes)? there might be reasons that some should stay single; the u.s. convention is usually double, but there are exceptions. not sure, actually, why robin would have gone with double. had a discussion once with sebastian on the subject; i think he went with double for tex live documentation, but i think he was thereby "recognizing" the u.s. origin of tex.

@DavidCarlisle -- ugh! please disallow wysiwyg editing!

Joseph Wright

@DavidCarlisle OK ...

@DavidCarlisle Do we want to go for a proper review cycle then? I can act as editor, with nothing going into master unless I've OK'd it ...

David Carlisle

@barbarabeeton you can't have a globally accessible website on a site that offers an in browser editor, and disallow that people use it.

Joseph Wright

20:13

@DavidCarlisle But we can restrict push access to the main repo

@barbarabeeton From what I could see, it's more-or-less random, possibly reflecting the original (pre-RF) editing

David Carlisle

@JosephWright not necessarily just musing. Although maybe you/we should start with a page of editing guidelines eg on whether or how to add a new page, any markdown idioms to use

barbara beeton

@DavidCarlisle -- sigh. (which is why i often bring a file down as source, and strip out all the "framework" manually, to get something i can edit reliably.)

Joseph Wright

@DavidCarlisle Did you see the CONTRIBUTING.md file? That was where I thought I'd collect anything up. I'm still unsure about the stuff that was \meta: the angle brackets are nice, but they don't work in code blocks, ...

David Carlisle

@barbarabeeton that's what I do on my village website (which is google pages rather than gh but same thing) I let people edit, but I pull the source in to emacs and have some emacs lisp functions that sanitise it and re-commit:-)

@JosephWright no, I'll look

barbara beeton

@JosephWright -- okay. if that's the reason, then going for consistency is best. if it remains "uk", then you should decide. (taking care to handle quotes within quotes by alternating the pairs to retain clarity.)

Joseph Wright

20:17

@barbarabeeton Like I say, the balance is toward US-style double quotes

barbara beeton

@JosephWright -- i volunteer to start with updating the pages dealing with ams stuff. also tug, if such exist.

David Carlisle

@JosephWright running the w3c link checker over it at moment, it's found a few things, I'll let it get to the end and gather up the results

Joseph Wright

@barbarabeeton Cool

@DavidCarlisle Good plan

Ulrike Fischer

@JosephWright sorry to interrupt but do you know if it is possible to set texmfhome in a travis build to one of the folders of the repository? I thought that `env:
- TEXMFHOME=${TRAVIS_BUILD_DIR}/texmf` in the .yml should work but it doesn't (it sets the variable but the files are not found) ;-(.

Joseph Wright

@UlrikeFischer No, I've never tried that

@DavidCarlisle A lot of the links are likely dead

David Carlisle

20:25

@JosephWright do I have write access? I tried to push "Please note " rather than "Please not " in contributing.md, just as a test but it failed authentication

Joseph Wright

@DavidCarlisle Give me a second

@DavidCarlisle Try now

David Carlisle

@JosephWright e pushed thanks:-)

Joseph Wright

@barbarabeeton Do you have a GitHub account?

David Carlisle

@JosephWright hmm back to forks and/or staging accounts. I don't think I want to edit live in master as the choices are to edit in the browser and see the preview or edit in emacs in my local checkout but then push to the site unseen, or some combination of the two involving cutting and pasting, to preview. Perhaps it would be better if i forked, edited on my branch and then did a PR. Not so important in this pre-release stage but i don't really trust my typing on a live site

barbara beeton

@JosephWright -- i allegedly do, but haven't tried it recently. (id is bbeeton15; whether or not the password in my "clues" list really works is an open question, but i can probably get that changed if i say i've forgotten it.)

Joseph Wright

20:38

@DavidCarlisle Fair enough

StrongBad

@barbarabeeton hey that is my new password manager. I randomly bash things on the keyboard and they next time I need to log in say I forgot the password.

barbara beeton

@DavidCarlisle -- not only would i not trust my typing on a live site, but the update procedure in such a situation would probably invoke a lot of bloatware.

@StrongBad -- yeah, but the last time i did that the site told me that the temporary password would expire in 15 minutes, but the message didn't reach me for more than half an hour. not really very effective. good thing it wasn't an emergency.

David Carlisle

@barbarabeeton not sure what you mean, by editing on a live site here I just mean checking out the faq site editing it in emacs and pushing it back. As unlike the tex version I have no local preview so I can't see it until it's pushed back live, If you fork you can push it to a temporary test site then move it to teh live site when its good.

@JosephWright @barbarabeeton although having just tried it, it's not so bad, if you edit the markdown locally to have a spurious foo.md edit open in github as you can drop in the whole file and "preview" which is probably good enough here.

barbara beeton

@DavidCarlisle -- ah. the ams web pages have a "live" edit, and it uses something like dreamweaver. but i prefer to edit clean xhtml code in emacs, send it through the w3c validator, fix if necessary, and then post. the garbage that's on ams pages (both bad/corrupted-by-the-system code and faulty links) is major discouraging. those of us who do validate are in the infinitesimal minority.

David Carlisle

@barbarabeeton as I was ~~moaning~~ discussing with Joseph earlier today in this chat, markdown seems not to have any effective validation other than "post to the site and see if it looks OK"

Joseph Wright

20:49

@DavidCarlisle Like I said, Markdown doesn't really do 'not valid', just 'no right': you could set up Jekyll ...

David Carlisle

@JosephWright no (as with latex-project.org) I think having to have a local jekyll misses the point. If I had to build it all locally I'd rather the input was tex or xml the gain in using gh-pages (or google sites or any of these free servers) is the fact taht you can realistically ask peopel to contribute without having to set anything up locally

Joseph Wright

@DavidCarlisle Sure

@DavidCarlisle We could have stuck with HTML, but again that's an issue

@DavidCarlisle I think if we stick to a small group of trusted people at the commit side, PRs from others are OK being reviewed

David Carlisle

@JosephWright I think markdown will be fine, just grousing from an xml geek:-)

@JosephWright so.. if we are going to start on the questions, editing in the main repo, do we need some way of avoiding edit conflicts, or just trust git to flag such things

Joseph Wright

@DavidCarlisle Ho ho

@DavidCarlisle Well Git will flag it up, but might be better to agree on some pattern ... I was thinking of picking headings and saying 'I'll deal with XXX'

Stefan Kottwitz

@DavidCarlisle @JosephWright a LOT of people would be able to work on the content. Only a few people are able to work on LaTeX development. Do you really choose to spent your time on practical documentation?

Joseph Wright

21:00

@StefanKottwitz Not exactly

@StefanKottwitz I'm trying to deal with the fact that the FAQ have drifted somewhat: I'm happy to take suggestions from other people

@StefanKottwitz Remember it's got a long UK-based history

David Carlisle

@JosephWright well I'll start with the mathml one (and that may be all today)

Joseph Wright

@StefanKottwitz Remember that most of it was done by one person

@StefanKottwitz Documentation is pretty important

Stefan Kottwitz

@JosephWright I just vote for motivating people to add work on it (multiply) instead of doing it yourself (add and substract)

Joseph Wright

@StefanKottwitz Sure: at present the plan is to get enough editing done that the usefulness is obvious to others

@StefanKottwitz Need to sort out the DNS, etc., so the new set up is the 'live' FAQ

David Carlisle

@JosephWright I need a plan b for link checking as if I give it the top level index it stops after a bit with Maximum number of documents (150) reached! I think I'll grep for absoluter URL locally and then link check a single test document with all the links

G. Bay

21:09

Is there any updated comparison of LaTeX vs MS Word/LibreOffice Word?

Stefan Kottwitz

@JosephWright I just made a thought on motivation (promoting) instead of pure doing. SE did never make TeX content themselves but caused 205.000 TeX answers to appear over time. ;-)

Joseph Wright

@StefanKottwitz A lot of the answers come from a small set of people ...

@G.Bay No real change ...

@DavidCarlisle Cool

G. Bay

@JosephWright If you compare word 2016 ... the space between words is a lot better then before

Stefan Kottwitz

@JosephWright A lot of the answers are from non-developers but experienced users who are glad to have this kind of way to contribute to the big vision.

Sebastiano

21:27

Good evening to all users into chat and the moderators.

-1

Q: Change the part of the appendix with letters of the alphabet

In relation to my previous question in this link Change part with arabic number, I ask how to obtain, using the Legrand orange book source and his structure.tex file, an appendix (in italian language) that has the letters of the alphabet. See this figure: I only managed to get the chapters, ...

appendices parts

Why when I ask a question I have always negative vote?

I would like to understand where I'm wrong and what I wrote about inaccurate. Almost absolutely I consider a question without having read it badly.

David Carlisle

@Sebastiano every time you get such a vote you come here and ask and we always say the same. You have again asked a question with no usable example.

Joseph Wright

@G.Bay Word is still a word processor, not a typesetting system: the two tasks are different

G. Bay

true

Joseph Wright

@StefanKottwitz Sure, but not many non-devs will provide Git commits: suggestions for edits are welcome, and getting things up to date would be very sensible

Stefan Kottwitz

@JosephWright yes, git is already a minimum qualification

Sebastiano

21:33

@DavidCarlisle It is impossible. In the previous question there is full code. I wonder if anyone reads my questions correctly. Mine is not a complaint but I always want to improve and understand. If I write it here, surely someone could tell me a reason or an explanation of it. We're all to help each other and do the pundits.

David Carlisle

@Sebastiano well I certainly didn't I have been working on something else, but i had a look when you asked, but you still just provide two disconnected fragments of code and expect anyone trying to reproduce your image to re-do your edits.

Sebastiano

21:53

@DavidCarlisle Thank you very much. But it takes a lot to add a comment rather than immediately use a negative vote even justified? The complete code is on my previous question.

David Carlisle

@Sebastiano well as i said last time, I don't negative vote (and the question while not that good probably doesn't deserve a negative vote but people are free to vote and to be honest while the site is very tolerant of new users, you are not a new user and if you have been asked to provide a MWE in your previous questions and then you post a question without one, you shouldn't be surprised if some people get annoyed

@JosephWright '' doesn't work to make a double quote does it, it still gets rendered as two single quotes it seems

David Carlisle

23:06

@JosephWright I'll sort out ^^ evenings this week:-)

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends