惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

博客园_首页
H
Help Net Security
N
Netflix TechBlog - Medium
Apple Machine Learning Research
Apple Machine Learning Research
P
Proofpoint News Feed
A
About on SuperTechFans
V
V2EX
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
宝玉的分享
宝玉的分享
aimingoo的专栏
aimingoo的专栏
F
Fortinet All Blogs
博客园 - 【当耐特】
Microsoft Security Blog
Microsoft Security Blog
Martin Fowler
Martin Fowler
I
InfoQ
Google DeepMind News
Google DeepMind News
人人都是产品经理
人人都是产品经理
Engineering at Meta
Engineering at Meta
腾讯CDC
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
B
Blog RSS Feed
U
Unit 42
The Cloudflare Blog
Y
Y Combinator Blog

The LaTeX Project

TUG 2026 Talks workshop at DocEng'26 in Fribourg Webinar on Accessible Mathematical Content in PDF LaTeX 2026-06-01 released and distributed Time for testing: the final pre-release for 2026-06 is available New Best Practice Guide for Mathematical Content in PDF PDF Association’s press release: Accessible math in PDF – finally! Pre-release 1 of LaTeX 2026-06-01 is available for testing LaTeX 2025-11-01 released and distributed Accessibility of STEM documents - talk at PDF days 2025 in Berlin Pre-release 3 of LaTeX 2025-11-01 is available for testing Retirement of the LaTeX-L mailing list Pre-release 4 of LaTeX 2025-06-01 is available for testing Final pre-release of LaTeX 2025-06-01 is available for testing New TeXLive.net server & Book Discounts
LaTeX 2025-06-01 released and distributed
2025-06-02 · via The LaTeX Project
Jun 2, 2025

We sent the 2025-06-01 release of the LaTeX kernel to CTAN over the weekend. It will be appearing to most users via TeX Live about now, and shortly in MiKTeX.

News from the tagging project

Work continues on the tagging project, and this release sees major work on tagging mathematics, with automatic generation of MathML now available in LuaTeX. The MathML can then be used as an Associated File (AF) for PDF 2.0 output: a big step in providing accessible mathematics. There are (as usual) some caveats at present, but for many cases, this is already very usable.

Tagging is now easier to turn on, as we’ve improved the keys to \DocumentMetadata: all you need now is

\DocumentMetadata{lang = ..., tagging = on}

This change emphasises that we are now at the stage where a tagging is becoming less experimental and more routinely-usable.

We’ve adjusted the standard PDF output: it’s now PDF 2.0 when tagging. That’s because PDF 2.0 is much better for tagging than PDF 1.x: but you can still request the older output format.

Other improvements in tagging are there, of course, most notably addressing some spacing oddities following display math mode.

Improving marks and output routine

A few releases ago, we integrated a new marks mechanism into the kernel, but left the classical approach available as a separate code path. That has now been addressed: the older interfaces now use the new mechanism, and multiple mark classes can be used with, for example, multicol.

The marks code was originally developed many (many) years ago as part of work on a new output routine (OR). Whilst today we don’t plan quite such radical changes, we are improving the OR. To support accessibility, we need to integrate a number of hooks, and this is an opportunity to refactor the OR to use the socket/plug approach. This should make the OR much easier to control without needing to completely replace it.

Other improvements and changes

The newsletter also covers various other small enhancements or corrections that we provide with the June release.

Enjoy – Joseph