« first day (3613 days earlier)      last day (1303 days later) » 

cis
12:12 AM
Well, they're all wrong! :()
What makes the mass is mostly crap. :()
 
12:44 AM
@cis I'm not sure I would call Dijkstra mostly crap. :) cs.utexas.edu/users/EWD/transcriptions/EWD08xx/EWD831.html
3
 
 
6 hours later…
6:41 AM
@AlanMunn Which fonts do you use/prefer nowadays for IPA?
 
 
2 hours later…
8:27 AM
@AlanMunn ooh alas poor yorick
 
8:45 AM
user image
4
 
9:26 AM
instead of having to type things like \begin{tcolorbox}[rounded corners, top=0pt, left=5pt,right=5pt, colback=blue!5!white, text height=7.4cm]
is there some way to set the height to be "from here to the end of the slide"? This is for beamer
 
10:23 AM
@Anush you could create a tcolorbox environment.
 
@PauloCereda can you say more about that please?
 
@Anush you could create an environment with \newtcolorbox which will be a tcolorbox environment with the given options.
From the manual:
\newtcolorbox{mybox}{colback=red!5!white,
colframe=red!75!black}

\begin{mybox}
This is my own box.
\end{mybox}
^^ fixed, looked at the wrong code.
 
10:55 AM
oh ok but this won't help find the right height will it?
 
 
2 hours later…
1:20 PM
@UlrikeFischer I don't have too much use for IPA generally since I'm a syntactician. I've generally used the SIL phonetics fonts (Doulos or Charis), but now that more Open Type fonts have a phonetic series there is a lot more choice. I suspect though that these fonts might be missing the full range of combining marks that are sometimes used in phonetic transcription, but for most uses, they will be fine.
@UlrikeFischer IPA fonts are designed to be serifed, so I wouldn't use any sans font for IPA.
 
2:15 PM
@PauloCereda My kind of Fortran: written in lower case.
 
@AlanMunn :)
 
@AlanMunn bah print * rather than WRITE(6
 
2:34 PM
@AlanMunn PRINT*, "ALAN DOESN'T LIKE SHOUTY FORTRAN"
 
2:52 PM
@DavidCarlisle True.
 
cis
@ pgfplots

If in the pgfplots-manual is used:

`\addplot [very thick] file {plotdata/group-1.dat};`

he does not find `plotdata/group-1.dat` in my OS, should he?
 
How can I ignore "overfull hbox" errors in specific fragments of my document?
Like "#pragma warning off...." in GCC C code
 
I was asked to meet someone here but I have to teach a class so I'll return later.
 
@cis no, that's an example code which shows the syntax to load a file.
 
@LaTeXereXeTaL Hi! Thanks for coming here!
@LaTeXereXeTaL I hope my beamer/mandi answer did not come across too rude, I just think using it with beamer was quite odd.
 
3:09 PM
@cis By the way: "The data file (and all others referenced in this manual) are shipped with pgfplots; you can find them in the subfolder doc/latex/pgfplots/plotdata." (quote from the manual)
Though I don't have that folder in my installation, so perhaps it's missing with the current version.
 
About your question about how to make it compatible with beamer:

The very minimum to make it not incompatible with beamer (i.e. compiles but maybe the output does not make always sense) would be
- not reload xcolor/hyperref with conflicting options if the beamer class is loaded
- don't overwrite the `hyperref` settings if the beamer class is loaded
- not redefine qed and problem if the beamer class is loaded
- don't load enumitem if the beamer class is loaded

In the bigger picture, there might be many problems with actually using it in a beamer document, for example a beamer slide won't h
... and totally unrelated to the problem with beamer, maybe consider using `\centering` instead of `\begin{center}...\end{center}` in
\newcommandx*{\image}[4][1={scale=1},usedefault]{%
\begin{figure}[H]
\begin{center}%
\includegraphics[#1]{#2}%
\end{center}%
\caption{#3}%
\label{#4}%
\end{figure}}
to avoid the additional vertical spacing
 
cis
@TorbjørnT. I do not find such a folder :(
 
@cis As I said, I didn't have it either, and it's not on CTAN as far as I can see, but if you want the data you can get it from GitHub: github.com/pgf-tikz/pgfplots/tree/master/doc/latex/pgfplots/…
 
cis
@Skillmonlikestopanswers.xyz This is a strange way of giving manual examples. The pgfplotstable manual also uses sample tables with raw data, but these can be found.
@TorbjørnT. OK
 
3:27 PM
@cis It's possible that something went wrong on the latest update to CTAN, that the files are supposed to be there. I'll add an issue on GitHub.
 
cis
@TorbjørnT. I have no idea... I get:
c:\texlive>tlmgr update pgfplots
tlmgr.pl: package repository vesta.informatik.rwth-aachen.de/ftp/pub/mirr
or/ctan/systems/texlive/tlnet (not verified: gpg unavailable)
tlmgr.pl: saving backups to C:/texlive/2020/tlpkg/backups
tlmgr.pl: no updates available
 
@cis I didn't mean to imply that a new version was available.
 
cis
Ok, now I understand you.

Would be fine, if one could use pgfplots-manual-examples directly in the future.
 
4:00 PM
@ComFreek by not having overfull boxes? (you may prefer \hfuzz=\maxdimen :-)
 
4:12 PM
@AlanMunn thanks, there was a discussion about using the tipa fonts with lualatex, (I don't think that it is a good idea as copy&paste doesn't work correctly and so accessibility is not given), but it made me wonder if sensible font support is missing here.
 
@UlrikeFischer Yes, although the input method of TIPA is fast for people who are used to it, I strongly recommend against using TIPA, and using a proper font with xelatex or lualatex, for exactly that reason.
@UlrikeFischer But I think most of the TIPA input markup works anyway? (Never been too clear on that point.)
 
4:36 PM
@AlanMunn no the idea here was to use unicode input but map that to the type1 fonts. See ctan.org/pkg/unitipa
 
5:10 PM
@UlrikeFischer This is a deeply misguided idea.
 
@AlanMunn yes, I agree but I couldn't convince the author ;-(
 
5:26 PM
@samcarter_is_at_topanswers.xyz Thank you for the suggestions. I think I can incorporate most of these. mandi is currently too large and I am going to split it into two different packages, one containing the core functionality and the other containing stuff that isn't directly applicable to the core functionality. The new version will use isomath and Arev fonts. Do you know if that will present any problems with beamer?
 
cis
5:43 PM
@DavidCarlisle And why shall I use

this
\noexpand\nextgroupplot[]
\noexpand\addplot+[] table[x=X, y=Y1]{\noexpand\data};


and not this

\unexpanded{\nextgroupplot[]
\addplot+[] table[x=X, y=Y1]{\data};}

?
The 2nd one is way economic....

Or is this synonym?
Ok, 1st one does not work.... I understand.
 
@LaTeXereXeTaL definitly too large ;-). It seems to grab every available short command name it could find. Btw: For me it simply break when I load it with ! LaTeX Error: Command \C already defined. And why does it load filehook, suffix, letltxmacro, epstopdf? None seems to be used.
 
5:59 PM
@cis I have no idea, you are just posting fragments out of context (again)
 
cis
@DavidCarlisle Sry, you said it some time. Whatever.
 
@cis you are looking at this and the context is in your head, for the rest of us, even if we commented on similar code a few days ago, a fragment out of context is pretty much impossible to comment on
 
6:57 PM
@LaTeXereXeTaL I don't know how practical that would be, but did you already consider splitting by formatting stuff and other macros? All these formatting stuff like selecting fonts, setting the colour of links, defining how example look like etc. is certainly good to create this book with a certain design. Users with different projects might only want all your macros for units etc. and not the formatting stuff.
 
@UlrikeFischer Hello. The best answer I can give is that I absolutely did not just load packages for no reason. When I needed to accomplish something, I researched and used the package that let me do what I needed to accomplish. The target audience is introductory physics students, and that almost exclusively drove design.
@UlrikeFischer I do indeed try to weed out unnecessary packages, and I can say with almost 100% confidence that everything that is there is used at least once for one purpose or another.
@UlrikeFischer The \C command defines the symbol for the coulomb. At the time I didn't even know about XeLaTeX and surely can't expect students to know either. I can, however, probably change that to make the package compatible with other engines (I only use pdflatex...I have no need for other engines). but not without also making many other similar changes for semantic consistency.
@UlrikeFischer As I recall, suffix was needed to define a particular command I needed for one particular purpose. I seem to remember something needing filehook for something similar.
 
@LaTeXereXeTaL If with arev fonts you mean \usepackage{arev}, then beamer needs \usefonttheme{professionalfonts} for the font to work
 
@LaTeXereXeTaL I didn't see the start of the thread just ulrike's comment but for example epstopdf is hardly ever needed.
 
@LaTeXereXeTaL epstopdf is hardly ever needed as the functionality is built-in, letltxmacro defines exactly two commands and they are not used in the sty. suffix defines WithSuffix and this isn't used either. filehook has a bit more, but imho none is used.
 
@samcarter_is_at_topanswers.xyz mandi wasn't used to create the M&I textbook. I based many of the commands in mandi on the presentations in that book.
 
7:07 PM
@LaTeXereXeTaL Thanks for the clarification! My suggestion is nevertheless to split "formatting in the style of M&I" off from macros that could be used for all kinds of projects
 
@LaTeXereXeTaL I didn't compile with xelatex, but with pdflatex. It breaks if one use hyperref with the unicode option (which I do by default, and which will be the default at the next hyperref update):
\documentclass{article}
\PassOptionsToPackage{unicode}{hyperref}
\usepackage{mandi}

\begin{document}
abc
\end{document}
 
My understanding was that epstopdf is necessary to convert images on the fly to PDF. That's what I initially used it. I would have no way of knowing whether that has changed because I have an extraordinarily difficult time finding such documentation. I never see dev changes to the kernel.
 
@LaTeXereXeTaL No, not for the last 10 years or so
 
Well mandi IS my "all kinds of projects" which do need very consistent formatting. I expect students to use these constructs also.
@UlrikeFischer That explains part of it then, because I have no need (presently anyway) for unicode. Until very recently, I didn't even understand what unicode even is.
I'll just say that one HUGE problem is the lack of widely published documentation. Of course the experts here know all this stuff, and the developers know it because those two populations largely overlap. Unfortunately TeX SE is the "last resort" for most of my questions and even here I can't always find answers because of the restrictions on what can and cannot be asked. So I've done the best I could.
 
@LaTeXereXeTaL well more likely here it is just old not-updated 3rd party documentation on the web. That is people don't add \usepackage{epstopdf} by accident, they add it because some random blog on some website written in 2005 told them to. It isn't clear what the core documentation can do about that.
 
7:17 PM
@LaTeXereXeTaL (disclaimer: I never used it myself, only browsed the documentation to debug this beamer question) To me it looks not like "all kinds of projects" but very specific to create notes/reports/etc in this textbook style. Other documents types like presentations won't benefit from layout decisions driven by such documents, but maybe your users would still be able to use your macros for units etc.
 
@DavidCarlisle well we could add to the documentation of epstopdf a remark "in latex hardly ever needed nowadays" ;-)
 
@LaTeXereXeTaL I should say that loading the package doesn't do any harm in this case (the core code uses 90% of the same code, factored out as epstopdf-base) but the general issue is very common with lots of packages (epsfig for example)
@UlrikeFischer no one reads documentation (unless it is on twitter)
 
@DavidCarlisle And I try VERY hard to avoid cargo cult programming. But I would never just add something without first having a need for it. Now, packages are updated all the time, without breaking backwards compatibility or moving functionality to other places. I"m sure that happens.
I read documentation when I can find it! I may be in the minority but I do it.
 
@LaTeXereXeTaL as I say graphicx in texlive has automatically done eps to pdf for a long time not sure when exactly but certainly more than 10 years. As Ulrike says we should update the into for the epstopdf doc to say not to use it...
 
LOL I"m feeling rather attacked for doing things I could not have known not to do.
And while I'm at it, I am also genuinely VERY honored to even be in the virtual presence of the names I've seen in print and online for so many years. I'm very familiar with all of your work and efforts.
Physics education has undergone a much needed shift to computation over the past decade and LaTeX and friends simply MUST be a part of that movement. That's why I'm here. I plan to write the book.
 
7:26 PM
@LaTeXereXeTaL No not at all, but you said the problem was lack of doc and and as the graphics doc is written by me, the question is what could I say. The question isn't how you should have known not to load the epstopdf package (that's a bit hard to find out) the real question is what doc told you to load the package as that doc is almost certainly wrong. But I can not trawl the internet and fix every latex tutorial out there.
 
I'm having my afternoon office hour right now on Zoom but no students are present.
@DavidCarlisle Well one thing I DIDN'T do initially that I am doing NOW is documenting the presence and purpose of each and every package right there in the code.
@DavidCarlisle As I recall, when I was learning to use figures and floats, there was a time when, I believe, only certain formats were acceptable but using the graphics command allowed for on-the-fly conversion at compile time. It could be that once I got that working, I never had a reason to look at the document AGAIN because it never stopped working for me.
 
@DavidCarlisle sometimes you cannot avoid overfull hboxes and sometimes they are safe to ignore.
 
@LaTeXereXeTaL I get such templates quite often: people added over time this command and that package and never removed something not needed anymore, and then when it suddenly clashes with a change they have a very hard time to disentangle everything and debug it.
 
@ComFreek That's why it's user configurable when to complain about them:-)
115
A: Do I have to care about bad boxes?

David CarlisleIt is worth noting that TeX doesn't make the value judgements here. The user, or more likely, the class file on behalf of the user, has set constraints on the amount by which boxes may overflow, the amount of stretching allowed on short pages etc. TeX only warns if these user-set constraints are...

 
can I make this into an environment that only locally disables warnings?
I guess so, if I save the value of \hfuzz
How can I do that?
 
7:33 PM
@ComFreek \begin{setlength}{\hfuzz}{\maxdimen} no wanings here \end{setlength}
 
cool
I'll try that tomorrow, thank you!
 
@ComFreek although I wouldn't do that, normally you are already in some environment from a float or a list or whatever so just set hfuzz locally and it'll restore at the end of the normal envirnoment
@ComFreek also of course you might want to set it less than \maxdimen, if it is 5cm off the page you might want a warning. \hfuzz=2cm ...perhaps
 
@LaTeXereXeTaL but the main problem are not the package, the main problem are the large number of commands you are defining. Normally you shouldn't define one letter commands, they are reserved for accents. And you shouldn't use up a lots of two letter commands and other standard names. You were only able to define \kg and not clash with siunitx as siunitx very carefully defines this only locally.
 
@UlrikeFischer I only learned of the supposed issue with one letter commands this summer with someone mentioning, I believe XeLaTeX (which I've never used, so couldn't have known). I've never see the restriction on them being reserved for accents in docs. Again, I may have missed it but with no errors I couldn't have know (it shouldn't be allowed if it's not allowed seems to me).
@UlrikeFischer Where are these standard names documented?
And the only reason I loaded siunitx is I have always been under the impression that loading a package will flag any incompatibilities (e.g. previously defined commands, etc.). While I don't need or use siunitx I also didn't want to conflict with it.
@samcarter_is_at_topanswers.xyz If I want to set the math fonts for students so they will always be consistent (one of the goals), then I can do that but at the expense of not being compatible with beamer, correct?
 
7:57 PM
@LaTeXereXeTaL this is not about documented names. I mean if someone wants to write a command for a vector, or for kilogramm, or to insert an image the first command names they would think of would be \vec and \kg and \image, so there is a large danger to clash with other packages or local documents if you use and define lots of "natural" command names.
 
@LaTeXereXeTaL al the standard accent commands are single letter, \', \H, \c etc, however as it happens \C is not defined by default so you must have loaded two packages that define it.
 
@DavidCarlisle it is hyperref (puenc.def).
 
@LaTeXereXeTaL your package can check if it is running in beamer and set fonts appropriate for presentations different to the fonts it uses in documents.
@UlrikeFischer don't use pu encoding then?
 
@DavidCarlisle who?
 
@UlrikeFischer hyperref:-)
 
8:08 PM
@UlrikeFischer I have to maintain semantic connections with what students are thinking based on language in class and in textbooks. Those choices don't always overlap with the traditional TeX/LaTeX ways of thinking. My experience is that students take to LaTeX faster when they use more semantically appropriate commands.
 
@DavidCarlisle ah yeah, in general, i want to be as warnings-sensitive as possible of course :)
 
@DavidCarlisle well if one use the unicode option it does use puenc.def. I don't think one can avoid it (and with xelatex it is used since quite some time).
 
Looking at the suffix docs, I apparently loaded it to define a starred command (something which before LaTeX3 was essentially impossible for me because I didn't understand the logic). However, I can't find the command in question.
 
@UlrikeFischer yes I know..
 
I also rather arrogantly expected students to write in English, so accents used in other languages would never be used. That obviously needs to be fixed.
 
8:12 PM
@LaTeXereXeTaL you shouldn't be forcing T1 encoding either (makes your package incompatible with luatex and xetex)
 
Serious question: If single letter commands are to be reserved for accents, would it be possible then to have the compiler prevent them from being defined to begin with?
@DavidCarlisle Aha...well you have me there because fonts are nothing but a black box to me. The "encoding" concepts are, for some reason, extremely difficult for me to understand. I spent many hours this summer trying to learn about that.
@DavidCarlisle OTOH, I have no reason to expect students to use other engines.
 
@LaTeXereXeTaL hm, you mean your students will never have to cite some foreigners with some curious accents in their names?
 
@LaTeXereXeTaL no. It completely breaks latex if you do that Even if you write the character as ç if \c is redefined the letter will break.
 
@UlrikeFischer LOL Not in an introductory calculus-based physics course, the first of its kind they've ever taken.
 
@LaTeXereXeTaL that would be surprising. Many people are switching to lualatex as the default engine irrespective of font usage as they want to use Lua, or are using packages that require Lua, or just because they don't like using software that's older than they are.
 
8:17 PM
@DavidCarlisle So the concept of "reserved words" doesn't apply in LaTeX?
@DavidCarlisle I've heard of all these other engines for a long time, but again, I haven't come across any documentation telling me why I should use them so of course I haven't felt the need to experiment with them. I always thought LaTeX was THE gold standard and the other were created for niche purposes.
I realize I probably sound quite stupid.
 
@LaTeXereXeTaL it's a macro expansion language with lexical analysis configurable at run time, absolutely nothing is reserved:-) even basic syntax like \ or { or ` ` a being a letter are changeable at any point in the document. All you can do is document that people should play nice and not break things.
@LaTeXereXeTaL latex is written in tex. pdflatex is the latex code processed with pdftex, lualatex is the latex code processed with luatex. they are both latex.
 
@DavidCarlisle Yes I shouldn't have phrased it that way.
 
@LaTeXereXeTaL well you just asked about the accent commands: in lualatex there are not really needed anymore, one could disable them as you can use accented chars without them. So if you want to free the commands, switch to lualatex ;-)
 
@UlrikeFischer Is that because lualatex uses unicode by default?
 
@LaTeXereXeTaL more or less. It is because lualatex can handle utf8 natively. So "ä" is a simple letter like "a" and so no commands are needed to typeset it.
 
8:23 PM
@LaTeXereXeTaL but for example if you are a government department (or even a university department) with a legal requirement to have accessible PDF you may (soon, if not now) find it is much more feasible to use lualatex than pdflatex, as at its heart pdflatex is based around 8 bit fonts and that makes unicode based accessibilty harder than you might hope. So making the package incompatible with unicode engines is a severe limitation
 
@UlrikeFischer Does that mean that barring explicit definition conflicts with other packages, single letter commands would work with no problems?
@DavidCarlisle Oh yes. I saw the accessibility talks from Tug2020.
@UlrikeFischer Changing the compiler is trivial with editing % !TEX TS-program = blah.
 
@LaTeXereXeTaL well it's not so easy ;-). They are basically not needed, but nevertheless they are defined and as many packages write code that works both for pdflatex and the unicode engines they are often used with them too. So I would say, better avoid to define a one letter command ...
 
@DavidCarlisle I'm familiar with the \xspace controversy. LOL
 
@LaTeXereXeTaL no actually I misread it, as you have it it's just a bad idea,I thought you were using it where it did nothing useful
 
@DavidCarlisle who maintains the epstopdf docu? ctan says Karl, ah no we are "epstopdf-pkg".
 
8:36 PM
@UlrikeFischer you:-)
 
The documentation dated 2020-01-24 says Hans Oberdiek.
 
@LaTeXereXeTaL don't believe everything you read on the internet:-)
 
Ah well...
 
@LaTeXereXeTaL note the footnote
 
@DavidCarlisle Ah don't get me started on github...
Github's learning curve makes LaTeX look like a cakewalk...
 
8:40 PM
@LaTeXereXeTaL epstopdf (like a hundred or so of Heiko's packages) has been maintained for the last 4 or 5 years by the ho-tex github "Oberdiek Package Support Group" otherwise known as Ulrike and me.
 
@DavidCarlisle Oh dear...
 
@DavidCarlisle I added two sentences to the docu. Does it look right?
 
@DavidCarlisle I chip in too, mainly by 'move it to expl3'
 
@JosephWright yes sorry I was sparing you a share in the blame:-)
@UlrikeFischer TeX-system -> \TeX-systems I think (twice)
 
@DavidCarlisle Wait for l3color ;)
 
8:48 PM
@DavidCarlisle RIP myself, gone are all those pretty programs I use, bye Linux, bye VIM, bye TeX .... (yes, I'm comparatively young)
 
@Skillmonlikestopanswers.xyz your life without vim will be much improved
 
@DavidCarlisle yeah, without an editor I'm free to constructively use my time in Word.
@DavidCarlisle oh, wait, M$ Office is older than me as well ... :(
(or is it a sad smiley?)
 
@Skillmonlikestopanswers.xyz we shouldn't let children in to the chat
 
@DavidCarlisle if you don' teach children the ways, how would you want to ensure that future generations don't redo the same errors we did (and craft yet another Word)? Also, according to probably all definitions I'm an adult...
 
@Skillmonlikestopanswers.xyz I think adult in this room should mean "older than tex" seems reasonable to me.
 
8:55 PM
@DavidCarlisle enjoy your monologue :P
 
@Skillmonlikestopanswers.xyz worse than a monologue, I'd have to talk to egreg
6
 
@DavidCarlisle there are worse partners for a conversation, I'd say.
 
@Skillmonlikestopanswers.xyz am I nobody ??
 
@UlrikeFischer you're a woman, so I don't question your age.
@UlrikeFischer you're like what, 25?
 
@LaTeXereXeTaL yes and no. With \usepackage{arev} beamer is still compilable (in contrast to the current option clashes and double definition problems of mandi), but users won't get the font you want them to have unless they use the professionalfont theme.
 
9:08 PM
@samcarter_is_at_topanswers.xyz ah yes, the professionalfont theme, which means "I have no deeper knowledge of font handling in LaTeX, but my supervisors told me it should look like this: /shows image of Arial"
@samcarter_is_at_topanswers.xyz imho, this option got the wrong name... :)
 
@samcarter_is_at_topanswers.xyz Thank you.
 
@LaTeXereXeTaL you can use \@ifclassloaded{beamer}{...}{...}and skip your font settings in the beamer case. Beamer themes probably have a better handle on font setups suitable for a presentation
 
@Skillmonlikestopanswers.xyz I thought the name meant "I want to use the same font as in my thesis and which is designed for reading long printed texts and will torture my audience trying to decipher it from the projection"
@DavidCarlisle I wonder if stackoverflow.com/questions/63999870/… might be interested in whatever magic you use for the learnlatex examples.
 
9:27 PM
@DavidCarlisle imho only once, the other has a "a" before".
 
@UlrikeFischer ok but two \ before TeX :-)
 
@DavidCarlisle done
 
@UlrikeFischer thanks
 
@DavidCarlisle should we upload it?
 
@UlrikeFischer could do, although I don't think it needs to, could go out at the next update (bound to be one at some point_
 
9:39 PM
@DavidCarlisle do you think there will be one? Fine with me.
 
@UlrikeFischer not aware of anything specific, just general principle that code changes happen. perhaps it should go out now...
 
@DavidCarlisle it reminded me that there is a PR lurking around in ltxcmds regarding the iso date format.
 
10:03 PM
@UlrikeFischer oh sorry should I merge that now, just looked at the diff, looks plausible:-)
 
@DavidCarlisle I think it should work. Imho I had it in use for quite some time.
 
@samcarter_is_at_topanswers.xyz sadly, in my case the two are equivalent... :)
 
@UlrikeFischer done, do you want to push to ctan or me? (not this evening)
 
@DavidCarlisle I can do it, but only tomorrow in the evening.
 
@UlrikeFischer same here so I'll let you take the blame for that one:-)
 
10:21 PM
@Skillmonlikestopanswers.xyz Arial for text? Yuck!
 
10:37 PM
@DavidCarlisle I forgot a \begin{macro} ;-(.
 
11:35 PM
@DavidCarlisle @UlrikeFischer Regarding the dates in ISO format, the latest updates in ho-tex are written in the respective documentations (.dtx) as \begin{Version}{2020-05-10 v1.25}, but the old lines are in the form \begin{Version}{2020/05/10 v1.25}, this makes the style of the documentation look a bit strange...maybe you can change it massively using some regex (I think)
 
@PabloGonzálezL same is true in the latex format code, we didn't change the old \changes entries
 
@DavidCarlisle Haaa...I didn't know that :(
@DavidCarlisle Any particular reason for this?
@DavidCarlisle It seems strange to have dates separated by - and others by / , of course, it is my opinion as a user reads the documentations :)
 
@PabloGonzálezL it's a good test that the code matches the spec and accepts both forms?
 
11:51 PM
@PabloGonzálezL Probably because lots of code written by English speakers uses / (although with varying orders), so / was probably adopted as the standard? But ISO date style uses hyphens.
@DavidCarlisle Does the version checking code (when you load a package) require / or can it use either?
 
@AlanMunn yes but Pablo means when we changed latex to accept iso - format dates, why didn't we change all the existing dates to use that format
@AlanMunn either (in recent years)
 
@DavidCarlisle I guess the real reason is it wasn't worth the effort, and if / was required for earlier versions then changing it would be somewhat anachronistic.
 
@DavidCarlisle Okay, but in terms of style it's better if it's one or the other. I understand that in formal LaTeX code it is necessary to keep / because it has always been there, but, in the case of `ho-tex' I don't think it is so serious to modify them to have a unified style
 
@AlanMunn yes that was my thought at the time. but see ^^ @PabloGonzálezL doesn't like it:-)
 
@DavidCarlisle :)
 
11:56 PM
@DavidCarlisle Hehe...I'm sure that if my guide professor at the university had seen this "difference in style", he would have failed me :(
 
@PabloGonzálezL nah you just tell him he's wrong and he should respect history and give you top grade. What could go wrong?
 
@PabloGonzálezL Ralph Waldo Emerson wrote "a foolish consistency is the hobgoblin of little minds". Seems appropriate in this case. :)
 

« first day (3613 days earlier)      last day (1303 days later) »