This book describes the Tectonic typesetting system. The goal of the Tectonic project is to empower people to create beautiful, effective technical documents.

Technical documents

What do we mean by “technical documents”? While the Tectonic project seeks to cast its net as widely as possible, common examples might be software manuals, scientific papers, or analytical reports. What features do such documents include that traditional authoring frameworks have trouble supporting?

  • Mathematics. Any kind of mathematical typesetting is challenging. Beautiful mathematical typesetting is extremely hard and requires deep integration with the overall typesetting system.
  • Abundant cross-references. Technical documents often involve extensive internal and external cross-references (links), and managing such links is a nightmare without extremely good tooling support.
  • Rich content. Technical documents also generally include a great deal of rich content beyond their text, such as figures, tables, and code. In the best documents, this content is seamlessly integrated into the document presentation, with precise author control over that presentation.
  • Integrated computation. In the 21st century, it is finally possible to integrate computation — runnable code samples, interactive graphics, live-updating data, and so on — into documents, and it is becoming clear that this new capability is not just evolutionary, but revolutionary.
  • Hackability. Finally, we also believe that technical documents should ideally be “hackable,” meaning that people should be able to see how they work “under the hood” and use them as a basis for their own creations.

Tectonic and TeX

At the core of Tectonic is a modernized, complete, self-contained TeX/LaTeX engine, powered by XeTeX and TeXLive.

For those new to it, TeX is a programming language. While most programming languages create software, TeX creates typeset documents. TeX is quite archaic in some ways, but in many fields it’s still the tool of choice for authoring the kinds of documents described above.

  • TeX is absolutely unparalleled in its ability to typeset math. Workers in virtually every mathematics-heavy field use TeX to create documents.
  • The TeX ecosystem provides infrastructure for deep and rich cross-referencing with programs like bibtex.
  • Another hallmark of the TeX ecosystem is longstanding support for complex figures, tables, and other forms of rich content included in the document.
  • Because TeX is a programming language for creating documents, TeX-based documents can be hackable in exactly the same way as the open-source programs that underly so much of the modern software ecosystem.

The fundamental principle underlying the Tectonic project is that TeX is, and can continue to be, the best language out there for creating the beautiful, effective technical documents that the world deserves. The TeX language is an amazingly clever piece of engineering, and the fact that it’s still in use 40 years (!) after its creation speaks for itself. But by the same token, there is a lot about the TeX software ecosystem that is archaic and outdated. The goal of Tectonic is to build on the good stuff and leave behind the things that don’t make sense anymore.

In particular, Tectonic is derived from the source code that powers the XeTeX engine, and the bulk of its code is the same core engine that implements the complex, Unicode-aware typesetting performed by XeTeX. Tectonic provides both a new user experience around that engine, and several key interventions that enable the engine to be used in fundamentally new ways.

The goals of Tectonic

As stated above, the overall goal of the Tectonic project is to enable people to create beautiful, effective technical documents. In particular, there are several elements of the existing TeX ecosystem that Tectonic aims to improve upon:

  • User experience. Many aspects of the classic TeX user experience (UX) are bizarre and unpleasant to modern users. In particular, its error messages and diagnostic output can be utterly mystifying. Tectonic chooses to break compatibility with classic TeX when doing so offers the chance to improve the UX.
  • Embeddability. The modern TeX software system consists of a suite of interacting command-line programs modifying a complex tree of thousands of support files. This makes it extremely unpleasant to embed the TeX engine within other software systems, which prevents a whole host of exciting use cases. Tectonic delivers its engine as a reusable software library and aims to make that library easy to embed and reuse anywhere code can run.
  • Reproducibility. For the same reasons that the classic TeX experience is difficult to embed, it is difficult to guarantee reproducible document builds. For many technical documents, reproducibility is a highly-respected virtue if not an outright requirement. Tectonic aims to enable easy, byte-for-byte reproducible builds.
  • Web output. Modern displays and Web browsers are incredibly powerful, versatile tools. One of the motivations for the founding of the Tectonic project was the belief that current Web-based technical documents are falling far short of what should be possible, and the belief that some changes in the core TeX engine are necessary to fully unlock its ability to produce excellent Web-based output.