TeXmacs typesetting practice: Thermodynamics textbook

Created by Steven Baltakatei Sandoval on 2022-01-02T19:32Z under a CC BY-SA 4.0 license and last updated on 2022-01-03T09:23Z.


In 2020, I decided to retypeset an open/libre textbook called "Thermodynamics and Chemistry" authored by Dr. Howard DeVoe, Associate Professor Emeritus of University of Maryland. A PDF draft containing everything but the bibliography and problem solutions is available here.


For some time I have wanted to explore all the features the math typesetting program TeXmacs has to offer. Also, I have wanted to share what I've learned in Chemical Engineering with others. Recent improvements in remote collaboration technologies have convinced me that it is worth investing significant amounts of time sharing what I know and helping others share what they know.

My experience with early digital texts of the 2000s

Story time.

I grew up at the end of an era when textbooks did not require maintenance of an internet connection. While attending university around 2009, I recall that physics textbooks offered extra features such as access codes to private web servers that provided dynamic content such as simple physics simulators. These were amusing curiosities that I don't think many students took advantage of. The graduate students acting as teaching assistants in my freshman physics course relied more on physical demonstrations and verbal discussions rather than attempting to insert a computer screen into the mix. The access codes printed on cardstock inserts in textbooks we purchased from the university bookstore were never used.

My experience with industry-specific digital texts

I graduated university but my desire to acquire knowledge still remained. While working to maintain a supercritical carbon dioxide injection plant in remote Southern Utah, I gained a more intuitive understanding of the laws of thermodynamics. I spent many hours analyzing pressure, temperature, composition, and flowrates of real fluids. The product of the entire process was hydrocarbon liquid that people indirectly paid me to extract so they could burn it in their automobiles. It was here that an engineer named Gary Brom recommended I try using the chemical process simulation software VMGSim (which I understand has since been acquired by Schlumberger). The company I worked for paid for the license (several thousand USD per year) since I was able to use the software to make predictions to justify certain equipment setpoints. There were temperature, pressure, and flowrate setpoints that were critical in minimizing problematic issues like hydrate formation, sulfide stress cracking from H₂S, and lost compressor capacity due to high improperly stabilized condensate. VMGSim helped me explore the possibility space of different equipment configurations without needing to risk real equipment. The software, however, did not include user education in thermodynamics within its scope. One could learn much from playing with different settings and seeing how simulations responded but in order to make most of the software features useful I had to educate myself. The money I earned from solving straightforward problems (e.g. corrosive RO permeate) allowed me to purchase industry-consensus "standards" containing knowledge sold by engineers before me to organizations such as:

I understand that each of these organizations has a committee who decide on what facts and recommendations to include in each journal issue and each standard document they publish. I worked with an engineer named Philip Dickinson who worked with one of these committees in his spare time. From what I can tell, participation in these committees is unpaid. Historically, someone who wanted a document from the libraries of these organizations could receive a physical copy via a mail-order catalogue.

Reappearance of subscription libraries

However, one trend I noticed after graduating from university in 2011 was that access to knowledge from such libraries was becoming more restricted, not more available over time. I know API, NACE, ASHRAE, ASME, and ANSI only sell digital copies of their standards as PDFs protected by Digital Rights Management mechanisms such as encryption or by printing the first purchaser's personal information as a watermark on every page. Additionally, even if you purchase a physical copy from ASME or API, the copy is simply a printout of the DRM-protected PDF. This reminds the reader on every page that unless they personally purchased their particular instance, they are criminals. I imagine this also means that a university engineering library that once was able to provide students and the public with access to physical volumes are now unable to do so without violating the DRM license. I may have seen indirect evidence of this when I visited visited the Columbia University in the City of New York's engineering library in 2018-05; I found that the library had been converted into a Wi-Fi hotspot with some token bookshelves of miscellaneous texts; in order to access content you had to be enrolled as a student and have an internet connection. I had visited the library and paid for a visitor's pass thinking that while I was visiting Manhattan I would pass some afternoons sampling various fields of knowledge. I was disappointed and ended up enjoying New Jersey's Institute of Technology's engineering library (still physical as of 2018-05) instead. Similarly, My alma mater's engineering library also suffered a fate similar to that of Columbia's in 2010.

Subscription libraries are not new. Objectively speaking, the policy of engineering libraries such as those of Stanford University to permit public access to texts was a historical aberration. A private entity should be able to withhold physical access to its possessions. My dissatisfaction with Stanford's move to restrict access to physical texts (and restricting access to non-physical texts via DRM) was because these physical restrictions didn't exist when I was an undergraduate student there; Stanford Libraries seemed functionally indistinguishable from what I would call a public engineering library.

The old books-resting-on-a-shelf model at Stanford could afford to be egalitarian because the costs of maintaining public physical access (books being lost, copyright violation required someone physically taking photographs or xeroxing copies) were sufficiently low. I am not aware of students wholesale xeroxing textbooks being a significant issue. However, once the prospect of books becoming digitized arose, the initial digital infrastructure Stanford adopted for storing, indexing, and serving these new books were proprietary platforms such as JSTOR. A person browsing a wide selection of journal articles used to be able to walk down the library stacks and pick random volumes off the shelf to read. Now with JSTOR, this person would have to basically have to pay to own a private copy of each article they wanted to view. This flies in the face of the fact that upon digitization, distribution costs for books drops to near zero. Digital books, be they photographs or digitally typeset, can be copied and transmitted without needing to create and transport glue, cellulose, and ink composites across the planet. In theory, a digital engineering library should cost less to browse and explore than a physical one.

Yet, as of 2022, I know of no engineering library that meets such criteria. However, I have identified some close digital analogues may meet my requirements.

Free digital libraries

Digital libraries that actively work to keep their texts open include include:

What these organizations share in common is use of "Copyleft" licenses, specifically, Creative Commons Attribution-ShareAlike 4.0 (CC BY-SA 4.0). The idea is that an author should be able to make sure anyone who shares or resells their work cannot become a middleman that attaches new legal strings to buyers of the work. A common example of such additional legal strings is a prohibition by a publisher against copying a work or circumventing copying restrictions (see DMCA). Such strings are useful tools to improve revenue to the publisher by requiring the publisher be a middleman in every transaction involving the work. However, this requirement can be abused for purposes besides simply improving sales of a work. For example, if a middleman can make a work unavailable unless a continuous subscription fee is paid (i.e. a subscription library), then the middleman has the ability to punish users based on matters unrelated to the purchase and therefore exert political control over the users.

To illustrate, let me give an example involving Lord of the Rings.

Lord of the Rings analogy of Copyleft

Imagine if Sauron had a benevolent brother named Miło who forged a magic ring that not only gave its user supernatural abilities but also the ability to create perfect copies of itself for others to use. Let's call this Miło's Ring 1.0. Intellectual property law would be analogous to how copies of Sauron's One Ring could not be made unless they were lesser rings whose continued use required continuous acceptance of Sauron's terms and conditions that he could change at any time. Sauron used his control over users of the lesser rings to convert some of them into his servants, such as the Nazgûl ("Ringwraiths") who lacked free will.

Without some Copyleft mechanism, any one of Miło's rings could be converted into a One Ring with its own set of manipulative lesser rings bound to it; this would be analogous to releasing Miło's ring into the public domain. Unlike public domain, however, Copyleft prohibits attaching mechanisms of control to manipulate users of further copies. This prohibition is made explicit in section 2.a.5.C of CC BY-SA 4.0:

No downstream restrictions. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material.

Also, the "ShareAlike" section (3.b) offers an interesting benefit. Let's say a wizard improved Miło's Ring 1.0 to make a 2.0 version that enabled the wearer to fly in addition to being invisible. The ShareAlike section would require that the improvement not be subject to the wizard's will. For example, a wizard couldn't assassinate someone by waiting until they used Miło's Ring 2.0 to fly high above the ground and then turning off the flying ability so they came crashing down to earth.

Some would argue that publishing under copyleft licenses is inferior to publishing to the public domain. Couldn't Miło release many copies of his ring in order to outcompete Sauron's lesser rings? This might be a viable strategy were Sauron not also capable of changing the terms and conditions of his lesser rings to permit, for a time, free duplication. Sauron could then make his lesser rings indistinguishable from the public domain Miło Ring 1.0 or market them as walled garden improvements. Then, when Miło Ring 1.0 has disappeared from Middle Earth marketshare and the last users hunted down, Sauron could gain power by changing the terms and conditions of his lesser rings to control everyone by disabling free duplication and requiring obedience to his will in order to unlock features that had previously been available. This is an outcome that could have been avoided if a copyleft license had been used.

Digital publishing reduces barrier to entry

The hypothetical library I desire should have books that do not subject readers to mechanisms of control. Mechanisms of control have historically been used to enforce payment to those involved in the book production or reselling process. Therefore, my hypothetical library, unlike Stanford's subscription library, cannot rely on DRM to pay for electricity, bandwidth, server hardware, and IT staff that are all required to make copyleft texts available. However, thanks to digital technologies of today in 2021, only a small fraction of an individual person's lifespan is required to produce a book; typesetting can be performed on a computer and plant fibers need not be pulped and bleached to produce paper; research collaborators can Email or video chat instead of physically transporting their bodies via passenger ships or airliners to conferences.

These digital technologies lower the barrier to entry for more people to share their knowledge. The success of Wikipedia in 2007 is proof of this. As of 2021, 5,000 volunteers provide more than 100 article edits per month and 64,000 provide more than 5 per month (Wikimedia.org permalink). This is evidence that many people are willing to collaborate to publish what they know on a volunteer basis.

Creative Commons as collaboration tool

I read Atlas Shrugged, a book by Ayn Rand, several times years ago in order to better understand people I knew who identified themselves as libertarians.

The story is a verbose caricature of ignoble socialists ("Moochers") who enslaved noble capitalists ("Producers") through blackmail. After several re-readings I realized Rand had given Moochers superior cooperation and collaboration abilities that enabled them to orchestrate absurdly effective laws that coerced/forced Producers to become slaves. The Producers, on the other hand, ended up not collaborating with eachother to legally defend themselves but instead encouraged one another to go on strike in order to starve the Moochers.

I bring the story up because it is a story that defends property rights of the individual which the Creative Commons BY-SA license arguably diminishes. In the John Galt chapter, a causal (not "casual"!) chain of inventor to investor to manufacturer to operator is described through which goods are produced. Between each link a higher entity dictates terms and conditions that restrict the actions of the lower entity. For example, an inventor may sell a computer program critical to the operation of a power plant to an investor under the condition that the program not be shared with anyone else without the inventor's permission; if the investor violates this contract then they may be required, ultimately under threat of government-sanctioned violent force, to pay fine. The contract could be encoded in a license that the inventor writes and transmits to the investor. To enforce their conditions, the inventor may choose to utilize a DRM scheme that remotely shuts off the program if they aren't paid a monthly subscription fee.

As I described with Miło's ring, DRM, a methods of control, can be made into a method of coercion. To illustrate in the context of Atlas Shrugged, imagine a modern rewrite in which Producers extinguish the lights of New York City not by a decade-long campaign to convince engineers to retire but instead by neglecting cloud authentication servers necessary to operate critical infrastructure. If a power plant requires programs running in a third party datacenter (e.g. Amazon or Microsoft) then that power plant operates at that third party's pleasure. Similarly, the inventor in my example could shut down the power plant by activating the program's DRM killswitch. DRM preserves property rights of upstream entities (e.g. inventors, programmers, writers) at the cost of restricting freedoms for downstream entities.

However, if the inventor used a Creative Commons BY-SA license, then such methods of control would not be necessary. Instead of using a copyright license to secure direct compensation, a Creative Commons license could be used as a collaboration tool. Like-minded Producers could freely contribute improvements to the program. Each Producer's experience with the operation of the improved program allows them to demand more compensation when selling their time implementing the program. For example, Red Hat, Inc., is a company that sells commercial support for free software such as GNU/Linux. Red Hat itself does not sell software so much as it sells support for software. Profits are used to purchase closed software in order to release them under free licenses, such as the GNU General Public License (GPL). Thus, newly freed software can be improved by anyone. Producers of software improvements can then charge support fees as consultants not necessarily tied to Red Hat. Although the Creative Commons BY-SA license is not recommended to be applied to software as the GPL is, both promote collaboration by giving recipients rights to copy, modify, and redistribute works.

Free and Open Source engineering textbooks

When I imagine my ideal public digital enginerring library, the texts it contains are licensed Creative Commons BY-SA. For reasons stated above, the quality of the texts could be incrementally improved by anyone. If I were to contribute or improve a text in such a library, I would not want my improvements to be used by a corporate Sauron to lure users into being controlled against their will.

As indirectly mentioned earlier, a nascent form of this library exists on Stack Exchange in the form of the Engineering section. Similar to Wikipedia, Stack Exchange requires that users submit their articles and comments under a Creative Commons BY-SA license. As of 2021, their website supports MathJax rendering of equations, permitting clear communication of math equations encoded with LaTeX markup. I have made some contributions (e.g. "How to calculate kinetic energy of gas flow in pipe given tempearture, pressure, pipe diameter and velocity") to this forum but the question-and-answer format promotes narrow-scope answers to specific questions. This format helps readers quickly find specific answers if they know how to form a query for a search engine. In contrast, an educational textbook is typically a more comprehensive work that provides a sufficiently broad background of a topic so the reader can answer a wide array of questions they may not have been able to articulate for themselves.

My Contribution


I have a desire to share my knowledge because I believe my life can be personally improved if more people such as myself share their knowledge. I have given some thought into how I can make this contribution. As of 2022, given my background in chemical engineering and knowledge about free licenses, I've decided to search for and improve a free and open source thermodynamics textbook.

The Book

I identified a textbook called "Thermodynamics and Chemistry, 2nd Edition" by Howard DeVoe, Associate Professor Emeritus of the University of Maryland. The textbook is published under the Creative Commons Attribution 4.0 International License (CC BY). This means I can create and even sell a derivative work so long as I attribute DeVoe as the original work's author.

So, I created a derivative work here on GitLab (PDF).

Transcribing the Book with TeXmacs

Dr. DeVoe courteously provided me with a copy of the LaTeX source code he used to compile the PDF files published on his website. Throughout 2021, I transcribed this source code into the markup used by GNU TeXmacs, a WYSIWYG typesetting program developed primarily by a researcher named Joris van der Hoeven.

I chose TeXmacs over a more generic LaTeX editor because:

  • It is part of the GNU project, meaning it is free/libre open source.
  • It uses a tree-like data structure for its source code.
  • The source code is not meant to be modified by a generic ASCII editor. Edits are meant to immediately provide visual feedback of typesetting details.
  • It has Emacs shortcuts available for fast keyboard navigation.

I had heard from a friend in graduate school that there wasn't a clear choice of software for creating LaTeX documents. As I understand the situation in academica, one writes LaTeX source in a text editor then renders it to a PDF to see the result. Additionally, the core typesetting software can be extended with plug-ins for which I could not find a package manager. It seems every graduate school lab has their own particular traditions for submitting typesetting source code to publishers. My friend recommended Overleaf, a browser-based WYSIWYG editor, but I have yet to be convinced that it can handle the textbook features that I have found TeXmacs to be capable of handling (e.g. automatic table-of-contents genration, automatic footnote and equation numbering).

On 2021-06-29, I published a blog post about how I reached the milestone of transcribing all fourteen chapters.

Current status

As of 2022-01-02, I am still fleshing out the Bibliography, Biography, and Appendices. However, the bulk of the main text, Table of Contents, and Index are functioning. A PDF draft is available here.

If you are interested in helping contribute changes, you can contact me via the GitLab project page, Twitter or snail mail.

Future plans

In 2022, I plan on transcribing the Solutions Manual. I skipped these sections to save time.

I also plan on integrating examples and solutions that a reader can verify using free/libre open source chemical process simulation software such as DWSIM. For instance, without such software, a problem about a steam electric generator would likely require the engineer to solve energy and mass balances by looking up values from large tables called steam tables. As I mentioned earlier, such software helped me daily to tune setpoints for heat exchangers, gas compressors, and other process units of chemical engineering. The source code for such process models would be as important to the textbook as captioned images or tables.


In 2021 I transcribed a thermodynamics textbook to satisfy a desire I had to share some of what I learned in the field of chemical engineering. The textbook source code is published on GitLab under a Creative Commons BY-SA license so that others can improve it. A compiled PDF draft is available here.