| 1 |
% \iffalse meta-comment |
| 2 |
% |
| 3 |
% Copyright 2024 |
| 4 |
% The LaTeX Project and any individual authors listed elsewhere |
| 5 |
% in this file. |
| 6 |
% |
| 7 |
% This file is part of the LaTeX base system. |
| 8 |
% -—————————————— |
| 9 |
% |
| 10 |
% It may be distributed and/or modified under the |
| 11 |
% conditions of the LaTeX Project Public License, either version 1.3c |
| 12 |
% of this license or (at your option) any later version. |
| 13 |
% The latest version of this license is in |
| 14 |
% https://www.latex-project.org/lppl.txt |
| 15 |
% and version 1.3c or later is part of all distributions of LaTeX |
| 16 |
% version 2008 or later. |
| 17 |
% |
| 18 |
% This file has the LPPL maintenance status "maintained". |
| 19 |
% |
| 20 |
% The list of all files belonging to the LaTeX base distribution is |
| 21 |
% given in the file `manifest.txt'. See also `legal.txt' for additional |
| 22 |
% information. |
| 23 |
% |
| 24 |
% The list of derived (unpacked) files belonging to the distribution |
| 25 |
% and covered by LPPL is defined by the unpacking scripts (with |
| 26 |
% extension .ins) which are part of the distribution. |
| 27 |
% |
| 28 |
% \fi |
| 29 |
% Filename: ltnews39.tex |
| 30 |
% |
| 31 |
% This is issue 39 of LaTeX News. |
| 32 |
|
| 33 |
\NeedsTeXFormat{LaTeX2e}[2020-02-02] |
| 34 |
|
| 35 |
\documentclass{ltnews} |
| 36 |
|
| 37 |
%% Maybe needed only for Chris' inadequate system: |
| 38 |
\providecommand\Dash {\unskip \textemdash} |
| 39 |
|
| 40 |
%% NOTE: Chris' preferred hyphens! |
| 41 |
%% \showhyphens{parameters} |
| 42 |
%% \hyphenation{because} |
| 43 |
|
| 44 |
\usepackage[T1]{fontenc} |
| 45 |
|
| 46 |
\usepackage{lmodern,url,hologo} |
| 47 |
|
| 48 |
\usepackage{csquotes} |
| 49 |
\usepackage{multicol} |
| 50 |
\usepackage{color} |
| 51 |
|
| 52 |
\providecommand\hook[1]{\texttt{#1}} |
| 53 |
|
| 54 |
\providecommand\meta[1]{$\langle$\textrm{\itshape#1}$\rangle$} |
| 55 |
\providecommand\option[1]{\texttt{#1}} |
| 56 |
\providecommand\env[1]{\texttt{#1}} |
| 57 |
\providecommand\Arg[1]{\texttt\{\meta{#1}\texttt\}} |
| 58 |
|
| 59 |
|
| 60 |
\providecommand\eTeX{\hologo{eTeX}} |
| 61 |
\providecommand\XeTeX{\hologo{XeTeX}} |
| 62 |
\providecommand\LuaTeX{\hologo{LuaTeX}} |
| 63 |
\providecommand\pdfTeX{\hologo{pdfTeX}} |
| 64 |
\providecommand\MiKTeX{\hologo{MiKTeX}} |
| 65 |
\providecommand\CTAN{\textsc{ctan}} |
| 66 |
\providecommand\TL{\TeX\,Live} |
| 67 |
|
| 68 |
|
| 69 |
\providecommand\githubissue[2][]{\ifhmode\unskip\fi |
| 70 |
\quad\penalty500\strut\nobreak\hfill |
| 71 |
\mbox{\small\slshape(% |
| 72 |
\href{https://github.com/latex3/latex2e/issues/\getfirstgithubissue#2 \relax}% |
| 73 |
{github issue#1 #2}% |
| 74 |
)}% |
| 75 |
\par\smallskip} |
| 76 |
%% But Chris has to mostly disable \href for his TEXPAD app: |
| 77 |
%% \def\href #1#2{#2} % Only For Chris' deficient TeX engine |
| 78 |
|
| 79 |
% simple solution right now (just link to the first issue if there are more) |
| 80 |
\def\getfirstgithubissue#1 #2\relax{#1} |
| 81 |
|
| 82 |
\providecommand\sxissue[1]{\ifhmode\unskip |
| 83 |
\else |
| 84 |
% githubissue preceding |
| 85 |
\vskip-\smallskipamount |
| 86 |
\vskip-\parskip |
| 87 |
\fi |
| 88 |
\quad\penalty500\strut\nobreak\hfill |
| 89 |
\mbox{\small\slshape(\url{https://tex.stackexchange.com/#1})}\par} |
| 90 |
|
| 91 |
\providecommand\gnatsissue[2]{\ifhmode\unskip\fi |
| 92 |
\quad\penalty500\strut\nobreak\hfill |
| 93 |
\mbox{\small\slshape(% |
| 94 |
\href{https://www.latex-project.org/cgi-bin/ltxbugs2html?pr=#1\%2F\getfirstgithubissue#2 \relax}% |
| 95 |
{gnats issue #1/#2}% |
| 96 |
)}% |
| 97 |
\par} |
| 98 |
|
| 99 |
\let\cls\pkg |
| 100 |
\providecommand\env[1]{\texttt{#1}} |
| 101 |
\providecommand\acro[1]{\textsc{#1}} |
| 102 |
|
| 103 |
\vbadness=1400 % accept slightly empty columns |
| 104 |
|
| 105 |
|
| 106 |
\let\finalpagebreak\pagebreak % for TUB (if they use it) |
| 107 |
\let\finalvspace\vspace % for document layout fixes |
| 108 |
|
| 109 |
\makeatletter |
| 110 |
% maybe not the greatest design but normally we wouldn't have subsubsections |
| 111 |
\renewcommand{\subsubsection}{% |
| 112 |
\@startsection {subsubsection}{2}{0pt}{1.5ex \@plus 1ex \@minus .2ex}% |
| 113 |
{-1em}{\@subheadingfont\colonize}% |
| 114 |
} |
| 115 |
\providecommand\colonize[1]{#1:} |
| 116 |
\makeatother |
| 117 |
|
| 118 |
|
| 119 |
% Undo ltnews's \verbatim@font with active < and > |
| 120 |
\makeatletter |
| 121 |
\def\verbatim@font{\normalsize\ttfamily} |
| 122 |
\makeatother |
| 123 |
|
| 124 |
|
| 125 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 126 |
\providecommand\tubcommand[1]{} |
| 127 |
\tubcommand{\input{tubltmac}} |
| 128 |
|
| 129 |
\publicationmonth{June} |
| 130 |
%\publicationyear{2024 --- DRAFT version for upcoming release} |
| 131 |
\publicationyear{2024} |
| 132 |
|
| 133 |
\publicationissue{39} |
| 134 |
|
| 135 |
\begin{document} |
| 136 |
|
| 137 |
\tubcommand{\addtolength\textheight{4.2pc}} % only for TUB |
| 138 |
|
| 139 |
\maketitle |
| 140 |
{\hyphenpenalty=10000 \exhyphenpenalty=10000 \spaceskip=3.33pt \hbadness=10000 |
| 141 |
\tableofcontents} |
| 142 |
|
| 143 |
\setlength\rightskip{0pt plus 3em} |
| 144 |
|
| 145 |
\medskip |
| 146 |
|
| 147 |
\section{Introduction} |
| 148 |
|
| 149 |
The \LaTeX{} Project team remains strongly focused on producing |
| 150 |
automatically tagged PDF output for accessibility and reuse. At the |
| 151 |
beginning of 2024 the ISO PDF/UA-2 and the WTPDF (well-tagged PDF) |
| 152 |
standards were released and we are glad to be able to report that it |
| 153 |
is now possible to use \LaTeX{} to automatically produce documents |
| 154 |
that conform to these new standards.\footnote{At the present time we |
| 155 |
are still in a trial/prototype phase in which only a limited set of |
| 156 |
document classes and packages are supported. Over the next releases |
| 157 |
we expect to gradually lift these restrictions and eventually provide |
| 158 |
the full functionality as part of the core distribution, rather than |
| 159 |
through \texttt{latex-lab} modules.} A sample collection of such |
| 160 |
documents ranging from classical texts, such as the Bible, to recent |
| 161 |
technical papers submitted to arXiv.org can be found at |
| 162 |
\url{https://github.com/latex3/tagging-project/discussions/72}. |
| 163 |
|
| 164 |
In February Ulrike and Frank presented the current project status |
| 165 |
during the 5th International Workshop on \enquote{Digitization and |
| 166 |
E-Inclusion in Mathematics and Science 2024} (DEIMS 2024) at Nihon |
| 167 |
University, Tokyo, Japan; see~\cite{39:deims}. |
| 168 |
|
| 169 |
|
| 170 |
\section{News from the \enquote{\LaTeX{} Tagged PDF} project} |
| 171 |
|
| 172 |
In the previous \LaTeX{} News~\cite{39:ltnews38} we announced some prototype |
| 173 |
support for tagged tabulars. Some of the necessary code has |
| 174 |
now been moved from \texttt{latex-lab} to the corresponding packages |
| 175 |
(using sockets and plugs) and to the \LaTeX{} kernel (for those parts |
| 176 |
that are also necessary for other aspects of tagging). |
| 177 |
|
| 178 |
The kernel code specific to tagging is implemented in the file |
| 179 |
\texttt{lttagging.dtx}. For now it contains \cs{UseTaggingSocket}, a |
| 180 |
special invocation command for sockets that are specific to |
| 181 |
tagging. This enables us to also provide \cs{SuspendTagging} and |
| 182 |
\cs{ResumeTagging}, i.e., a very efficient way to temporarily disable |
| 183 |
the whole tagging process. This is, for example, necessary if some |
| 184 |
code is doing trial typesetting. In that case the trials should not |
| 185 |
generate tagging structures\Dash only the finally-chosen version |
| 186 |
should. Thus, \pkg{tabularx}, for example, stops the tagging while |
| 187 |
doing its trials to figure out the correct column widths to use, and |
| 188 |
then re-enables tagging when the table is finally typeset. |
| 189 |
|
| 190 |
Over time, \texttt{lttagging.dtx} will hold more general tagging code |
| 191 |
as appropriate. For now it is only documented as part of |
| 192 |
\texttt{source2e.pdf} but long term we will provide a separate guide |
| 193 |
for tagging, which will then also include the information currently |
| 194 |
found in various other places, e.g., \texttt{tagpdf.pdf}. |
| 195 |
|
| 196 |
We also added support for a few missing commands |
| 197 |
described in Leslie Lamport's \emph{\LaTeX{} |
| 198 |
Manual}~\cite{39:Lamport}: If \texttt{phase-III} is used |
| 199 |
the \cs{marginpar} command will be properly tagged (depending on |
| 200 |
the PDF version) as an \texttt{Aside} or a \texttt{Note} structure. |
| 201 |
In the standard classes \cs{maketitle} will be tagged if the additional |
| 202 |
testphase module \texttt{title} is used. |
| 203 |
|
| 204 |
The \texttt{math} module has been extended and now includes |
| 205 |
options to attach MathML files to the structures. |
| 206 |
First tests with a PDF reader and screen reader that support |
| 207 |
associated files look very promising. Examples of PDF files tagged with the |
| 208 |
new method can be found at |
| 209 |
\url{https://github.com/latex3/tagging-project/discussions/72}. |
| 210 |
|
| 211 |
At last various small bugs and problems reported at |
| 212 |
\url{https://github.com/latex3/tagging-project} |
| 213 |
have been fixed. |
| 214 |
Such feedback is very valuable, |
| 215 |
so we hope to see you there and thank you for |
| 216 |
any contribution, whether it is an issue or a post on a discussion |
| 217 |
thread. |
| 218 |
|
| 219 |
|
| 220 |
|
| 221 |
\section{Enhancements to the new mark mechanism} |
| 222 |
|
| 223 |
In June 2022 we introduced a new mark mechanism~\cite[p.~76]{39:ltnews} that allows keeping track of multiple |
| 224 |
independent marks. It also properly supports top marks, something that wasn't |
| 225 |
reliably possible with \LaTeX{} before. |
| 226 |
|
| 227 |
There was, however, one limitation: to retrieve the marks from the |
| 228 |
page data it was necessary to \cs{vsplit} that data artificially so |
| 229 |
that \TeX{} would produce split marks that the mechanism could then |
| 230 |
use. Unfortunately, \TeX{} gets very upset if it finds infinite |
| 231 |
negative glue (e.g., from \cs{vss}) within this data. This is not |
| 232 |
totally surprising because such glue would allow splitting off any |
| 233 |
amount of material as such glue would hide its size. \TeX{} |
| 234 |
therefore responds with an error message if it find such glue while |
| 235 |
doing a \cs{vsplit} operation (and it does so even if a later glue |
| 236 |
item cancels the infinite glue). |
| 237 |
|
| 238 |
To account for this, the code in 2022 attempted to detect this |
| 239 |
situation beforehand and if so did not do any splitting but, of |
| 240 |
course, it would then also not extract any mark information. |
| 241 |
|
| 242 |
In this release the approach has been changed and we always do a |
| 243 |
\cs{vsplit} operation and thus always get the right mark data |
| 244 |
extracted. While it is not possible to avoid upsetting \TeX{} in case |
| 245 |
we have infinite negative glue present, it is possible to hide this |
| 246 |
(more or less) from the user.\footnote{A note to \pkg{l3build} users |
| 247 |
that make use of its testing capabilities: the new mechanism |
| 248 |
temporarily changes \cs{interactionmode} and, for implementation |
| 249 |
reasons in \TeX{}, that results in extra newlines in the \texttt{.log} |
| 250 |
file, so instead of seeing \texttt{[1] [2]} you will see each on |
| 251 |
separate lines. This means that test files might show differences of |
| 252 |
that nature, once the code is active, and must therefore be |
| 253 |
regenerated as necessary.} With the new code \TeX{} will neither stop |
| 254 |
nor show anything on the terminal. What we can't do, though, is |
| 255 |
avoid an error being written to the log file, but to make it clear |
| 256 |
that this error is harmless and should be ignored we have arranged the |
| 257 |
code so that the error message, if it is issued, takes the following |
| 258 |
format: |
| 259 |
\begingroup |
| 260 |
\makeatletter |
| 261 |
\def\verbatim@font{% |
| 262 |
\small\ttfamily} |
| 263 |
\makeatother |
| 264 |
\begin{verbatim} |
| 265 |
! Infinite glue shrinkage found in box being split. |
| 266 |
<argument> Infinite shrink error above ignored ! |
| 267 |
l. ... } |
| 268 |
\end{verbatim} |
| 269 |
Not perfect (especially the somewhat unmotivated \texttt{<argument>}), |
| 270 |
but you can only do so much when error messages and their texts are |
| 271 |
hard-wired in the engine. |
| 272 |
|
| 273 |
\endgroup |
| 274 |
|
| 275 |
So why all this? There are two reasons: we do not lose marks in edge |
| 276 |
cases any more, and perhaps more importantly we are now also reliably |
| 277 |
able to extract marks from arbitrarily boxed data, something that |
| 278 |
wasn't possible at all before. This is necessary, for example, to |
| 279 |
support extended marks in \env{multicols} environments or extract them |
| 280 |
from floats, marginpars, etc.\tubcommand{\looseness-1} |
| 281 |
|
| 282 |
Details about the implementation can be found in \texttt{texdoc |
| 283 |
ltmarks-code} or in the shorter \texttt{texdoc ltmarks-doc} (which |
| 284 |
only describes the general concepts and the command interfaces). |
| 285 |
|
| 286 |
|
| 287 |
\section{Providing \pkg{xtemplate} in the format} |
| 288 |
|
| 289 |
In \LaTeX{} News~32, we described the move of one long-term experimental idea |
| 290 |
into the kernel: the package \pkg{xparse}, which was integrated as \pkg{ltcmd}. |
| 291 |
With this edition, we move another long-term development idea to stable status: |
| 292 |
\emph{templates}. |
| 293 |
|
| 294 |
In this context, templates are a mechanism to abstract out various elements |
| 295 |
of a document (such as \enquote{sectioning}) in such a way that different |
| 296 |
implementations can be interchanged, |
| 297 |
%% CAR added text: --- FMi: but it is wrong |
| 298 |
%% variations of the document markup can be accommodated, |
| 299 |
and design decisions can be implemented efficiently and controllably. |
| 300 |
|
| 301 |
In contrast to \pkg{ltcmd}, which provides a mechanism that many document |
| 302 |
authors will exploit routinely, templates are a more specialised tool. We |
| 303 |
anticipate that they will be used by a small number of programmers, providing |
| 304 |
generic ideas that will then be used within document classes. Most document |
| 305 |
authors will therefore likely directly encounter templates only rarely. |
| 306 |
We anticipate though that they will be \emph{using} templates provided |
| 307 |
by the team or others. |
| 308 |
|
| 309 |
%% CAR added this paragraph: FMi: doesn't belong in ltnews |
| 310 |
%We shall continue to develop these generic ideas and interfaces in various ways: for example, to support the varied document markup possibilities |
| 311 |
%provided in \pkg{ltcmd} and elsewhere. |
| 312 |
|
| 313 |
The template system requires three separate ideas |
| 314 |
\begin{itemize} |
| 315 |
\item Template \emph{type}: the \enquote{thing} we are using templates |
| 316 |
for, such as \enquote{sectioning} or \enquote{enumerated-list} |
| 317 |
\item A template: a combination of code and keys that can be used |
| 318 |
to implement a type. Here for example we might have |
| 319 |
\enquote{standard-\LaTeX{}-sectioning} as a template for |
| 320 |
\enquote{sectioning} |
| 321 |
\item One or more \emph{instances}: a specific use case of a template |
| 322 |
where (some) keys are set to known values. We might for example see |
| 323 |
\enquote{\LaTeX{}-section}, \enquote{\LaTeX{}-subsection}, etc. |
| 324 |
\end{itemize} |
| 325 |
|
| 326 |
As part of the move from the experimental \pkg{xtemplate} to kernel integration, |
| 327 |
the team have revisited the commands provided. The stable set now comprises |
| 328 |
\begin{itemize} |
| 329 |
\item \cs{NewTemplateType} |
| 330 |
\item \cs{DeclareTemplateInterface} |
| 331 |
\item \cs{DeclareTemplateCode} |
| 332 |
\item \cs{DeclareTemplateCopy} |
| 333 |
\item \cs{EditTemplateDefault} |
| 334 |
\item \cs{UseTemplate} |
| 335 |
\item \cs{DeclareInstance} |
| 336 |
\item \cs{DeclareInstanceCopy} |
| 337 |
\item \cs{EditInstance} |
| 338 |
\item \cs{UseInstance} |
| 339 |
\item \cs{IfInstanceExistsTF} and variants |
| 340 |
\end{itemize} |
| 341 |
|
| 342 |
To support existing package authors, we have released an updated version of |
| 343 |
\pkg{xtemplate} which will work smoothly with the new kernel-level code. The |
| 344 |
existing commands provided in \pkg{xtemplate} will continue to work, but |
| 345 |
we encourage programmers to move to the set above. |
| 346 |
|
| 347 |
|
| 348 |
\section{New or improved commands} |
| 349 |
|
| 350 |
\subsection{\pkg{doc}:\ Provide \cs{ProvideDocElement}} |
| 351 |
|
| 352 |
In addition to |
| 353 |
\cs{NewDocElement} and \cs{RenewDocElement} we now also offer a |
| 354 |
\cs{ProvideDocElement} declaration that does nothing unless the doc |
| 355 |
element could be declared with \cs{NewDocElement}. This can be useful |
| 356 |
if documentation files are processed both individually and |
| 357 |
combined. |
| 358 |
|
| 359 |
|
| 360 |
\subsection{\pkg{doc}:\ Better support for \pkg{upquote}} |
| 361 |
|
| 362 |
In \LaTeX{} News~37~\cite{39:ltnews37} we wrote that support for the |
| 363 |
\pkg{upquote} package was added to the \pkg{doc} package, but back |
| 364 |
then this was added only for \cs{verb} and the \env{verbatim} |
| 365 |
environments. However, |
| 366 |
in a typical \texttt{.dtx} file, most of the code will be in the body |
| 367 |
of some \env{macrocode} or \env{macrocode*} environments, and neither |
| 368 |
of these was affected by adding \pkg{upquote}. |
| 369 |
We have now updated \pkg{doc} so that |
| 370 |
\pkg{upquote} alters the quote characters in these environments as |
| 371 |
well. |
| 372 |
% |
| 373 |
\githubissue{1230} |
| 374 |
|
| 375 |
\subsection[\pkg{ifthen}:\ Allow active characters in comparisons] |
| 376 |
{\pkg{ifthen}:\ Guard against active characters in comparisons} |
| 377 |
The \cs{ifthenelse} command now ensures that \verb|<|, \verb|=| and |
| 378 |
\verb|>| are safe |
| 379 |
in numeric tests, even if they have been made active |
| 380 |
(typically by \pkg{babel} language shorthands). |
| 381 |
% |
| 382 |
\githubissue{756} |
| 383 |
|
| 384 |
\subsection%[New conditionals: \cs{IfClassAtLeastT}, etc.] |
| 385 |
{New conditionals:\ \cs{IfClassAtLeastT} and friends} |
| 386 |
|
| 387 |
Around 2020 we added a number of conditionals with CamelCase names, i.e., |
| 388 |
% |
| 389 |
\cs{IfClassAtLeastTF}, |
| 390 |
\cs{IfClassLoadedTF}, |
| 391 |
\cs{IfClassLoadedWithOptionsTF}, |
| 392 |
\cs{IfFormatAtLeastTF}, |
| 393 |
\cs{IfPackageAtLeastTF}, |
| 394 |
\cs{IfPackageLoadedTF}, and |
| 395 |
\cs{IfPackageLoadedWithOptionsTF} |
| 396 |
% |
| 397 |
to help arranging conditional code that depends on the release of a |
| 398 |
particular class, package or format. However, we only provided the |
| 399 |
\texttt{TF} commands and not also the \texttt{T} and \texttt{F} |
| 400 |
variants. This has now been changed. |
| 401 |
|
| 402 |
In 2023 we introduced \cs{IfFileAtLeastTF} but we did not also provide |
| 403 |
\cs{IfFileLoadedTF} at the same time. This conditional and its |
| 404 |
\texttt{T} and \texttt{F} variants have now also been added. Remember |
| 405 |
that one can only test for files that contain a \cs{ProvidesFile} |
| 406 |
line. |
| 407 |
% |
| 408 |
We did the same for the conditionals \cs{IfLabelExistsTF} and |
| 409 |
\cs{IfPropertyExistsTF}, also introduced in 2023.\footnote{By mistake they |
| 410 |
were initially introduced under the names \cs{IfLabelExistTF} and |
| 411 |
\cs{IfPropertyExistTF}; we corrected that at the same time. This is a |
| 412 |
breaking change, but the commands have been used so far only in |
| 413 |
kernel code.} |
| 414 |
% |
| 415 |
\githubissue[s]{1222 1262} |
| 416 |
|
| 417 |
|
| 418 |
\section{Code improvements} |
| 419 |
|
| 420 |
\subsection{Load packages only at the top level} |
| 421 |
|
| 422 |
Classes and packages must be loaded only by using the commands |
| 423 |
\cs{documentclass} and \cs{usepackage} or the class interface commands |
| 424 |
such as \cs{LoadClass} or \cs{RequirePackageWithOptions}; moreover, |
| 425 |
all of these must always be used at the top level, and not inside a |
| 426 |
group of any type (for example, within an environment). |
| 427 |
Previously \LaTeX\ did not check this, |
| 428 |
which would often lead to low level errors later on if package declarations |
| 429 |
were reverted when a group ended. |
| 430 |
\LaTeX\ now checks the group level and an error is thrown |
| 431 |
if the class or package is loaded in a group. |
| 432 |
% |
| 433 |
\githubissue{1185} |
| 434 |
|
| 435 |
\subsection{Keep track of lost glyphs} |
| 436 |
|
| 437 |
A while ago we changed the \LaTeX{} default value for |
| 438 |
\cs{tracinglostchars} from \texttt{1} to \texttt{2} so that missing |
| 439 |
glyphs generate at least a warning, but we forgot to make the same |
| 440 |
change to \cs{tracingnone}. Thus, when issuing that command \LaTeX{} |
| 441 |
stopped generating warnings about missing glyphs. This has now been |
| 442 |
corrected. |
| 443 |
% |
| 444 |
\githubissue{549} |
| 445 |
|
| 446 |
\subsection{Improve \pkg{fontenc} error message} |
| 447 |
|
| 448 |
If the \pkg{fontenc} package is asked to load a font encoding for which it |
| 449 |
doesn't find a suitable \texttt{.def} file then it generates an error |
| 450 |
message indicating that the encoding name might be misspelled. That |
| 451 |
is, of course, one of the possible causes, but another one is that the |
| 452 |
installation is missing a necessary support package, e.g., that no |
| 453 |
support for Cyrillic fonts has been installed. The error message text |
| 454 |
has therefore been extended to explain the issue more generally. |
| 455 |
% |
| 456 |
\githubissue{1102} |
| 457 |
|
| 458 |
|
| 459 |
\subsection{Warn if counter names are problematic} |
| 460 |
|
| 461 |
In the past it was possible to declare, for example, |
| 462 |
\verb/\newcounter{index}/ with the side-effect that this defines |
| 463 |
\cs{theindex}, even though \LaTeX{} has a \env{theindex} environment |
| 464 |
that then got clobbered by the declaration. |
| 465 |
% |
| 466 |
This has now been changed: if \cs{the}\meta{counter} is already |
| 467 |
defined it is not altered, but instead a warning message is displayed. |
| 468 |
% |
| 469 |
\githubissue{823} |
| 470 |
|
| 471 |
\subsection{Extended information in \cs{listfiles}} |
| 472 |
|
| 473 |
The \cs{listfiles} command provides useful information when finding issues |
| 474 |
related to variation in package versions. However, this has to date relied on |
| 475 |
the information in the \cs{ProvidesPackage} line, or similar: that can be |
| 476 |
misleading if for example a file has been edited locally. We have now extended |
| 477 |
\cs{listfiles} to take an optional argument which can include the MD5 |
| 478 |
hash and size of each file in the \texttt{.log}. Thus for |
| 479 |
example you can use |
| 480 |
\begin{verbatim} |
| 481 |
\listfiles[hashes,sizes] |
| 482 |
\end{verbatim} |
| 483 |
to get both the file sizes and file hashes in the \texttt{.log} as well as |
| 484 |
the standard release information. |
| 485 |
% |
| 486 |
\githubissue{945} |
| 487 |
|
| 488 |
\subsection{Optimize creation of simple document commands} |
| 489 |
|
| 490 |
Creating document commands using declarations such as \cs{NewDocumentCommand}, etc., provides a very |
| 491 |
flexible way of grabbing arguments. When the document command only takes simple |
| 492 |
mandatory arguments, this has to-date added an overhead that could be avoided. |
| 493 |
We have now refined the internal code path such that \enquote{simple} document |
| 494 |
commands avoid almost any overhead at point-of-use, making the results |
| 495 |
essentially as efficient as using \cs{newcommand} for low-level \TeX{} |
| 496 |
constructs. Note that as \cs{NewDocumentCommand} makes engine-robust commands, |
| 497 |
the direct equivalent to \cs{newcommand} is \cs{NewExpandableDocumentCommand}. |
| 498 |
% |
| 499 |
\githubissue{1189} |
| 500 |
|
| 501 |
\subsection{\texorpdfstring{\raggedright}{}Handling of end-of-lines in \texttt{+v} arguments of \cs{NewDocumentCommand} and friends} |
| 502 |
|
| 503 |
The \texttt{+v} argument type provided by declarations such as \cs{NewDocumentCommand}, etc., allows |
| 504 |
grabbing of multiple lines of text in a verbatim-like argument. Almost always, |
| 505 |
the result of this grabbing will be used in a typesetting context. Previously, |
| 506 |
the end-of-line characters were stored literally as category code~12 |
| 507 |
(\enquote{other}) \verb|^^M| tokens. However, these are difficult to work with |
| 508 |
in general. We have now revised this behavior, such that end-of-line characters |
| 509 |
are converted to the \cs{obeyedline} command when parsed by \texttt{+v}-type |
| 510 |
arguments. |
| 511 |
This change may require adjustments to the source of some documents, but the |
| 512 |
enhanced ability of users and programmers to exploit the \texttt{+v}-type |
| 513 |
argument means we believe it is necessary. |
| 514 |
|
| 515 |
\subsection{Declaring appropriate sub-encodings for \texttt{TS1} symbol fonts} |
| 516 |
|
| 517 |
In 2020 we incorporated support for the \texttt{TS1} symbol encoding |
| 518 |
directly into the kernel and in this way removed the need to load the |
| 519 |
\pkg{textcomp} package~\cite{39:ltnews31} to make commands such as |
| 520 |
\cs{texteuro} available. |
| 521 |
|
| 522 |
There is, however, a big problem with this \texttt{TS1} symbol |
| 523 |
encoding: only very few fonts provide every glyph that is supposed |
| 524 |
to be part of \texttt{TS1}. This means that changing font families |
| 525 |
might result in certain symbols becoming unavailable. This can be a |
| 526 |
major disaster if, for example, |
| 527 |
the symbol \cs{texteuro} (\texteuro) or \cs{textohm} (\textohm) |
| 528 |
no longer gets printed in your document, just because you altered the |
| 529 |
text font family. |
| 530 |
|
| 531 |
To mitigate this problem, in 2020 we also introduced the declaration |
| 532 |
\cs{DeclareEncodingSubset}. This declaration is supposed to be used in |
| 533 |
font definition files for the \texttt{TS1} encoding to specify which |
| 534 |
subset (we have defined 10 common ones) a specific font implements. If |
| 535 |
such a declaration is used then missing symbols are automatically |
| 536 |
taken from a fallback font. |
| 537 |
|
| 538 |
%% CAR -- this paragraph maybe needs more substantial rewording ?? |
| 539 |
While this is not perfect, it is the best |
| 540 |
you can do other than painstakingly checking that your document |
| 541 |
%% CAR: was "painstrickenly" -- should be a word, but!! |
| 542 |
uses only glyphs that the font supports |
| 543 |
and, if necessary, switching to a different font or avoiding the |
| 544 |
missing symbols. See also the discussion in \cite{39:ltnews33}. |
| 545 |
|
| 546 |
To jumpstart the process we also added declarations to the \LaTeX{} |
| 547 |
kernel for most of the fonts found in \TL{} at the time\Dash with |
| 548 |
the assumption that such declarations would over time be superseded by |
| 549 |
declarations in the \texttt{.fd} files. Unfortunately, this hasn't |
| 550 |
happened yet (or not often) and so many of the initial declarations |
| 551 |
went stale: several fonts got new glyphs added to them (so their |
| 552 |
sub-encoding should have been changed but didn't); others (mainly due |
| 553 |
to license issues) changed the family name and thus our declarations |
| 554 |
became useless and the renamed fonts (now without a declaration) ended |
| 555 |
up in the default sub-encoding that offers only a few glyphs; |
| 556 |
yet others such as CharisSIL (which triggered the GitHub issue) were |
| 557 |
simply not around at the time. |
| 558 |
|
| 559 |
We have, therefore, again attempted to provide the (currently) correct |
| 560 |
declarations, but it is obvious that this is not a workable |
| 561 |
process. As we do not maintain the fonts we do not have the |
| 562 |
information that something has changed, and to regularly check the |
| 563 |
ever growing font support bundles is simply not possible. It is |
| 564 |
therefore very important that maintainers of font packages not only |
| 565 |
provide \texttt{.fd} files but also add such a declaration to every |
| 566 |
\texttt{TS1...fd} font definition file that they distribute. |
| 567 |
|
| 568 |
To simplify this process, we now provide a simple \LaTeX{} file |
| 569 |
(\texttt{checkencodingsubset.tex}) for determining the correct (safe) |
| 570 |
sub-encoding. If run, it asks for a font family and then outputs its |
| 571 |
findings, for example, for \texttt{AlgolRevived-TLF} you will get: |
| 572 |
\tubcommand{\begin{small}} |
| 573 |
\begin{verbatim} |
| 574 |
----------------------------------------- |
| 575 |
Testing font family AlgolRevived-TLF |
| 576 |
(currently TS1-sub-encoding 9) |
| 577 |
----------------------------------------- |
| 578 |
Some glyphs are missing from sub-encoding 8: |
| 579 |
==> \textcelsius (137) is missing |
| 580 |
==> \texttwosuperior (178) is missing |
| 581 |
==> \textthreesuperior (179) is missing |
| 582 |
==> \textonesuperior (185) is missing |
| 583 |
Some glyphs are missing from sub-encoding 7: |
| 584 |
==> \texteuro (191) is missing |
| 585 |
All glyphs between sub-encoding 6 and 7 exist |
| 586 |
All glyphs between sub-encoding 5 and 6 exist |
| 587 |
All glyphs between sub-encoding 4 and 5 exist |
| 588 |
Some glyphs are missing from sub-encoding 3: |
| 589 |
==> \textwon (142) is missing |
| 590 |
All glyphs between sub-encoding 2 and 3 exist |
| 591 |
Some glyphs are missing from sub-encoding 1: |
| 592 |
==> \textmho (77) is missing |
| 593 |
==> \textpertenthousand (152) is missing |
| 594 |
All glyphs between sub-encoding 0 and 1 exist |
| 595 |
All glyphs in core exist |
| 596 |
----------------------------------------- |
| 597 |
TS1 encoding subset for AlgolRevived-TLF (ok) |
| 598 |
Use sub-encoding 9 |
| 599 |
----------------------------------------- |
| 600 |
\end{verbatim} |
| 601 |
\tubcommand{\end{small}\noindent}% |
| 602 |
This output is meant for human consumption, e.g., you see which glyphs |
| 603 |
are missing and why a certain sub-encoding is suggested, but it is not |
| 604 |
that hard to use it in a script and extract the suggested sub-encoding |
| 605 |
by grepping for the line starting with \texttt{Use sub-encoding}. |
| 606 |
|
| 607 |
Of course, this check will only work if the missing glyphs are really |
| 608 |
missing: some fonts placed \enquote{tofu}\footnote{Little squares to |
| 609 |
indicate a missing symbol.} into such slots and in this case it looks |
| 610 |
to \TeX{} as if the glyph is provided. For example, for the old |
| 611 |
Palatino fonts (family \texttt{ppl}) it would report |
| 612 |
\begin{verbatim} |
| 613 |
----------------------------------------- |
| 614 |
TS1 encoding subset for ppl (bad) |
| 615 |
Use sub-encoding 0 (not 5) |
| 616 |
----------------------------------------- |
| 617 |
\end{verbatim} |
| 618 |
thus it claims that |
| 619 |
all glyphs are provided, while in reality more than twenty are missing |
| 620 |
and sub-encoding 5, as declared in the kernel, is in fact correct. |
| 621 |
% |
| 622 |
\githubissue{1257} |
| 623 |
|
| 624 |
|
| 625 |
\subsection{Behavior when loading \pkg{textcomp} without options} |
| 626 |
|
| 627 |
When incorporating the \pkg{textcomp} package into the \LaTeX{} kernel, in |
| 628 |
the February 2020 release~\cite{39:ltnews31}, the default type of its package messages was changed from package info (\texttt{Package textcomp Info}) |
| 629 |
to \LaTeX{} kernel info (\texttt{LaTeX Info}). But if \pkg{textcomp} |
| 630 |
was loaded without options, the message type got restored to package info. |
| 631 |
This restoration has now been canceled. |
| 632 |
|
| 633 |
Note that loading \pkg{textcomp} with one of the options \option{error}, |
| 634 |
\option{warn}, or \option{info} still changes the message |
| 635 |
type to an error, warning, or info message from the \pkg{textcomp} package. |
| 636 |
% |
| 637 |
\githubissue{1333} |
| 638 |
|
| 639 |
|
| 640 |
\subsection{Rollback improvements} |
| 641 |
|
| 642 |
When requesting a rollback of the \LaTeX{} kernel and/or packages, |
| 643 |
several packages produced the error \enquote{Suspicious rollback date} |
| 644 |
because their rollback section contained only data about recent |
| 645 |
releases even if the package, such as \pkg{array}, was available since |
| 646 |
the first release of \LaTeXe{} in 1994. We now suppress this error and |
| 647 |
load the first release that is still part of the distribution (and |
| 648 |
hope for the best). This change was implemented for the packages |
| 649 |
\pkg{amsmath}, |
| 650 |
\pkg{array}, |
| 651 |
\pkg{doc}, |
| 652 |
\pkg{graphics}, |
| 653 |
\pkg{longtable}, |
| 654 |
\pkg{multicol}, |
| 655 |
\pkg{showkeys}, |
| 656 |
\pkg{textcomp}, and |
| 657 |
\pkg{varioref}. |
| 658 |
% |
| 659 |
\githubissue{1333} |
| 660 |
|
| 661 |
|
| 662 |
|
| 663 |
\section{Documentation improvements} |
| 664 |
|
| 665 |
\subsection{Further updates to the guides} |
| 666 |
|
| 667 |
We reported about the updated versions of \texttt{usrguide} and |
| 668 |
\texttt{clsguide} in \LaTeX{} News~37~\cite{39:ltnews37}. We have now |
| 669 |
revised \texttt{fntguide} as well to reflect the changes and macros |
| 670 |
added to the kernel over the last years of development. Note that the |
| 671 |
file name hasn't changed and there is no \texttt{fntguide-historic}. |
| 672 |
|
| 673 |
\section{Bug fixes} |
| 674 |
|
| 675 |
\subsection{Fix inconsistent expansion of the package option list} |
| 676 |
|
| 677 |
\LaTeX{} applies one-step expansion to the raw option list of packages |
| 678 |
and classes, so that constructions such as |
| 679 |
\begin{verbatim} |
| 680 |
\def\myoptions{opt1,opt2} |
| 681 |
\usepackage[\myoptions]{foo} |
| 682 |
\end{verbatim} |
| 683 |
are supported. |
| 684 |
%% CAR Rewritten sentence: |
| 685 |
But if a package declares its options using the new |
| 686 |
key/value approach~\cite{39:ltnews35} and it gets loaded a second time, |
| 687 |
then its raw option list will not be expanded and so an error might be raised. |
| 688 |
This has now been corrected. |
| 689 |
% |
| 690 |
\githubissue{1298} |
| 691 |
|
| 692 |
\subsection{Fix logic for first mark (page region)} |
| 693 |
|
| 694 |
In the new mark mechanism introduced in June 2022~\cite{39:ltnews35} the result of |
| 695 |
\cs{FirstMark} on a two-column page was incorrect if the first column |
| 696 |
contained no marks. In that case it should have returned the first |
| 697 |
mark of the second column but didn't. |
| 698 |
This has now been corrected. |
| 699 |
|
| 700 |
Documents using \cs{leftmark} are not affected, because that command is still |
| 701 |
using the old mechanism for now. |
| 702 |
% |
| 703 |
\githubissue{1359} |
| 704 |
|
| 705 |
|
| 706 |
\subsection{Struts at the end of footnotes or \texttt{p} columns} |
| 707 |
|
| 708 |
To produce consistent spacing in footnotes and tabular |
| 709 |
\texttt{p}-cells \LaTeX{} adds a strut at the beginning and end of |
| 710 |
the content. |
| 711 |
%% CAR Rewritten sentences: |
| 712 |
%% It assumed, however, that the footnote or tabular cell ended in |
| 713 |
%% horizontal mode and until now added the struts unconditionally. As a |
| 714 |
%% result, with vertical material at the end of the material this strut |
| 715 |
%% started a new paragraph consisting of a single line with just the |
| 716 |
%% strut inside. |
| 717 |
This assumed, however, that the content of the footnote or tabular cell ended in |
| 718 |
horizontal mode and so, until now, these struts were unconditionally added; |
| 719 |
as a result, if this content ended with vertical material then this strut |
| 720 |
started a new paragraph consisting of a single line with just the |
| 721 |
strut in it. This has finally been corrected and now the placement |
| 722 |
logic for the strut changes when vertical mode is detected. |
| 723 |
|
| 724 |
\emph{(First seen in a bug report for \pkg{footmisc} in combination with |
| 725 |
\pkg{bigfoot})} |
| 726 |
|
| 727 |
|
| 728 |
\subsection{Fix a ``missing \cs{item}'' rollback error} |
| 729 |
|
| 730 |
If \LaTeX{} is rolled back to a date between 2023/06/01 (inclusive) and |
| 731 |
2024/06/01 (exclusive), any list-based environment would raise an |
| 732 |
error: |
| 733 |
%(shown on two lines for \textsl{TUGboat}): |
| 734 |
\begin{verbatim} |
| 735 |
! LaTeX Error: Something's wrong--perhaps |
| 736 |
a missing \item. |
| 737 |
\end{verbatim} |
| 738 |
This has now been corrected as a hotfix in patch level 2, by enhancing |
| 739 |
the 2023/06/01 version rollback code of the new paragraph mechanism. |
| 740 |
% |
| 741 |
\githubissue{1386} |
| 742 |
|
| 743 |
|
| 744 |
|
| 745 |
\section{Changes to packages in the \pkg{amsmath} category} |
| 746 |
|
| 747 |
\subsection{\pkg{amsmath}:\ Correct equation tag placement} |
| 748 |
|
| 749 |
If there is not enough space to place an equation tag on the same line |
| 750 |
as the equation then \pkg{amsmath} calculates a suitable offset and it |
| 751 |
places the tag above (or below) the equation. In the case of the |
| 752 |
\env{gather} environment this offset was not reset at the end, with |
| 753 |
the result that it also got applied to any following environment, |
| 754 |
resulting in incorrect spacing in certain situations. This has now |
| 755 |
been corrected. |
| 756 |
% |
| 757 |
\githubissue{1289} |
| 758 |
|
| 759 |
|
| 760 |
|
| 761 |
%\section{Changes to packages in the \pkg{graphics} category} |
| 762 |
|
| 763 |
\section{Changes to packages in the \pkg{tools} category} |
| 764 |
|
| 765 |
\subsection{\pkg{array}, \pkg{longtable}, \pkg{tabularx}:\ Support tagging} |
| 766 |
|
| 767 |
%% CAR Rewritten sentences |
| 768 |
These three packages have been extended so they can now, on request, |
| 769 |
produce tagged tabular. This is done by adding a number of sockets (see |
| 770 |
\cite{39:ltnews38}) that, by default, do nothing; but when tagged PDF |
| 771 |
is requested they get equipped with appropriate plugs. |
| 772 |
|
| 773 |
In the previous \LaTeX{} release this was handled in \texttt{latex-lab}, by |
| 774 |
patching the packages when tagging was requested. |
| 775 |
|
| 776 |
|
| 777 |
\subsection{\pkg{array}:\ No \cs{unskip} in math cells} |
| 778 |
|
| 779 |
Math cells in the standard \env{array} environment of the kernel are |
| 780 |
not subject to space removal at the right end of the cell, i.e., explicit |
| 781 |
spaces from \cs{hspace} or \verb*|\ |, etc.\ are honored |
| 782 |
(normal spaces are automatically ignored in math). In the \pkg{array} |
| 783 |
package all spaces got removed by calling \cs{unskip} unconditionally, |
| 784 |
regardless of the type of cell. |
| 785 |
This difference in behavior has now been removed by correcting the processing of |
| 786 |
math cells in \pkg{array}. |
| 787 |
% |
| 788 |
\githubissue{1323} |
| 789 |
|
| 790 |
|
| 791 |
\subsection{\pkg{verbatim}:\ \cs{verb} showed visible spaces} |
| 792 |
|
| 793 |
A recent change in the kernel was not reflected in the \pkg{verbatim} |
| 794 |
package, with the result that \cs{verb} showed visible spaces |
| 795 |
(\verb*/ /) after the package was loaded. This has already been corrected |
| 796 |
in a hotfix for the November 2023 release. |
| 797 |
% |
| 798 |
\githubissue{1160} |
| 799 |
|
| 800 |
\subsection{\pkg{verbatim}:\ Support tabs in \cs{verbatiminput*}} |
| 801 |
|
| 802 |
Mimicking the November 2023 kernel update that allowed \cs{verb*} |
| 803 |
to mark tabs as spaces, the \pkg{verbatim} package |
| 804 |
has now been updated so that \cs{verbatiminput*} also marks tabs as spaces. |
| 805 |
\githubissue{1245} |
| 806 |
|
| 807 |
|
| 808 |
\subsection{\pkg{multicol}:\ \cs{columnbreak} interferes with mark mechanism} |
| 809 |
|
| 810 |
The \pkg{multicol} package has to keep track of marks (from |
| 811 |
\cs{markright} or \cs{markboth}) as part of its output routine code |
| 812 |
and can't rely on \LaTeX{} handling that automatically. It does so by |
| 813 |
artificially splitting page data with \cs{vsplit} to extract the mark |
| 814 |
data. With the introduction of \cs{columnbreak} that code failed |
| 815 |
sometimes, because it was not seeing any mark that followed such a |
| 816 |
forced column break. |
| 817 |
|
| 818 |
This has now been corrected, but there is further work to do, because |
| 819 |
as of now \pkg{multicol} does not yet handle marks using the new mark |
| 820 |
mechanism\Dash see the discussion at the beginning of the newsletter. |
| 821 |
% |
| 822 |
\githubissue{1130} |
| 823 |
|
| 824 |
\subsection{\pkg{showkeys}:\ Allow \cs{newline} in \pkg{amsthm} to work} |
| 825 |
% |
| 826 |
Previously \pkg{showkeys} added an extra box layer which disabled the \cs{newline} |
| 827 |
of \pkg{amsthm} theorem styles. This extra box has now been avoided. |
| 828 |
% |
| 829 |
\githubissue{1123} |
| 830 |
|
| 831 |
\tubcommand{\newpage} |
| 832 |
\subsection{\pkg{xr}:\ Support links and properties} |
| 833 |
% |
| 834 |
The \pkg{xr} package implements a system for eXternal References. |
| 835 |
The \pkg{xr-hyper} package (in the \pkg{hyperref} bundle) |
| 836 |
extended this to also support links to external documents. |
| 837 |
%% CAR: Rewritten sentences: |
| 838 |
Using last year's extension of the \cs{label} command, which unified |
| 839 |
the label syntax of \LaTeX{} and \pkg{hyperref}, |
| 840 |
it became possible to merge the two packages and thus make |
| 841 |
\pkg{xr-hyper} obsolete. |
| 842 |
With this change it is also possible to refer to properties |
| 843 |
that are stored in external documents using \cs{RecordProperties}. |
| 844 |
% |
| 845 |
\githubissue{1180} |
| 846 |
|
| 847 |
\section{Changes to files in the \pkg{cyrillic} category} |
| 848 |
|
| 849 |
\subsection{Correct definition of \cs{k}} |
| 850 |
|
| 851 |
Ages ago, the encoding-specific definitions for various accent |
| 852 |
commands were changed to guard against altering some parameter values |
| 853 |
non-locally by mistake. For some reason the definition for \cs{k} in |
| 854 |
the Cyrillic encodings \texttt{T2A}, \texttt{T2B}, and \texttt{T2C} |
| 855 |
didn't get this treatment. This oversight has now been corrected. |
| 856 |
% |
| 857 |
\githubissue{1148} |
| 858 |
|
| 859 |
|
| 860 |
\begin{thebibliography}{9}\frenchspacing |
| 861 |
|
| 862 |
%\fontsize{9.3}{11.3}\selectfont |
| 863 |
|
| 864 |
|
| 865 |
\bibitem{39:Lamport} |
| 866 |
Leslie Lamport. |
| 867 |
\newblock \emph{{\LaTeX}: {A} Document Preparation System: User's Guide and Reference |
| 868 |
Manual}. |
| 869 |
\newblock \mbox{Addison}-Wesley, Reading, MA, USA, 2nd edition, 1994. |
| 870 |
\newblock ISBN 0-201-52983-1. |
| 871 |
\newblock Reprinted with corrections in 1996. |
| 872 |
|
| 873 |
\bibitem{39:ltnews} \LaTeX{} Project Team. |
| 874 |
\emph{\LaTeXe{} news 1--39}. June, 2024. |
| 875 |
\url{https://latex-project.org/news/latex2e-news/ltnews.pdf} |
| 876 |
|
| 877 |
\bibitem{39:ltnews31} \LaTeX{} Project Team. |
| 878 |
\emph{\LaTeXe{} news 31}. February, 2020. |
| 879 |
\url{https://latex-project.org/news/latex2e-news/ltnews31.pdf} |
| 880 |
|
| 881 |
\bibitem{39:ltnews33} \LaTeX{} Project Team. |
| 882 |
\emph{\LaTeXe{} news 33}. June 2021.\\ |
| 883 |
\url{https://latex-project.org/news/latex2e-news/ltnews33.pdf} |
| 884 |
|
| 885 |
\bibitem{39:ltnews35} \LaTeX{} Project Team. |
| 886 |
\emph{\LaTeXe{} news 35}. June 2022.\\ |
| 887 |
\url{https://latex-project.org/news/latex2e-news/ltnews35.pdf} |
| 888 |
|
| 889 |
\bibitem{39:ltnews37} \LaTeX{} Project Team. |
| 890 |
\emph{\LaTeXe{} news 37}. June 2023.\\ |
| 891 |
\url{https://latex-project.org/news/latex2e-news/ltnews37.pdf} |
| 892 |
|
| 893 |
\bibitem{39:ltnews38} \LaTeX{} Project Team. |
| 894 |
\emph{\LaTeXe{} news 38}. November 2023. |
| 895 |
\url{https://latex-project.org/news/latex2e-news/ltnews38.pdf} |
| 896 |
|
| 897 |
\bibitem{39:deims} Frank Mittelbach and Ulrike Fischer. |
| 898 |
\emph{Enhancing \LaTeX{} to automatically produce |
| 899 |
tagged and accessible PDF}. \textsl{TUGboat} 45:1, 2024. |
| 900 |
\url{https://latex-project.org/publications/indexbyyear/2024/} |
| 901 |
|
| 902 |
%\bibitem{39:blueprint} Frank Mittelbach and Chris Rowley. |
| 903 |
% \emph{\LaTeX{} Tagged PDF \Dash A blueprint for a large project}.\\ |
| 904 |
% \url{https://latex-project.org/publications/indexbyyear/2020/} |
| 905 |
|
| 906 |
\end{thebibliography} |
| 907 |
|
| 908 |
\end{document} |
| 909 |
|