Introducing non-technical people to LaTeX | Hacker News Reader

Introducing non-technical people to LaTeX

texblog.net

46 pointsgnosis25 comments

I've taught around 20 non-technical people LaTeX. I would say that these are all good tips. But really, the only thing a non-technical person needs to learn LaTeX is to have someone sit down with them and teach them 1 on 1. I have a fairly refined lesson at this point, and can successfully teach anyone LaTeX in 45 minutes to an hour and a half, depending on how quickly they get it.

I've taught mostly undergraduates (while I was one), studying Biology, Chemistry, Linguistics, Psychology, Economics, English, Classics, and Architecture, so I'd like to try to answer the question "Why?"

LaTeX is worth learning if you're regularly writing papers of any length. In all of the disciplines above there's at least one problem with using word that makes composing a paper longer than five pages a huge pain. More often than not LaTeX is able to relieve some of the pain of that process. For example, in Linguistics, sentence trees are fairly easy to create using LaTeX[1], or for a classics paper, being able to switch between typing in Greek or English with a single command is a huge win.[2]

Besides the field specific features, there are a lot of other benefits to LaTeX. It's true that just about every person that I've taught LaTeX to has been a little skeptical at first as to what exactly it could provide. The thing that I tell them is that it is incredibly helpful in structuring their document. They end up loving being able to do an outline with section headers, and fill in the body paragraphs. They like having automatic reference and bibliography insertion, and they love being able to comment out sections of their document.

I think there's also a reason why it isn't hard to teach a non-technical person LaTeX. When teaching someone a programming language, the student will need to learn or already know a fair amount about how a computer works, and what programming can accomplish. With LaTeX, students generally already know how to write a paper, and will be able to easily map that process in to their learning process.

I'm a huge LaTeX evangelist. If you don't already know it, find someone who does, and have them sit down with you for an hour. Ask them to go through the steps of creating a document that has: Sections, subsections, tables, lists, figures, floats, references, and math. By the time you're done, you'll have a good enough knowledge to google for anything you can't reference in the document you just produced. You'll be amazed at how easy it was.

[1] http://www.ling.upenn.edu/advice/latex/qtree/qtreenotes.pdf

[2] http://www.tug.org/texlive/Contents/live/texmf-dist/doc/gene...

> Biology

What would be the problem? I've both written papers in Latex and Word, but cannot think of a discipline-related advantage.

The big one, as I understand it, is inserting nice looking tables for ANOVA results. Both Stata and SAS can output their results tables to LaTeX.

To add to this, R has fantastic LaTeX interoperability through Sweave, which makes it pretty simple to write R syntax whose results are directly rendered in LaTeX. I wrote about 100 pages worth of stats problem sets this way, and it was a life-saver. Any science that requires stats can benefit from LaTeX.

My response to this post is, what on earth for? While you can use LaTeX for a ridiculous number of things [1][2][3], LaTeX is primarily a tool for typesetting technical documents. Note that word "technical" in there; why would a non-technical person want that? It's complex; you have to type ridiculous things like "\textbf" to get bold text, and "\begin{enumerate}" to make a simple list. It's complicated; you have to look up what header argument to make to simply change the margins. It's also difficult to debug; try to explain to a non-technical user what this means:

  ("C:\Program Files\MiKTeX 2.9\tex\context\base\supp-pdf.mkii"
  [Loading MPS to PDF converter (version 2006.09.02).]
  )
  ! Undefined control sequence.
  l.66 Your text goes here. \agweagsd

The whole idea doesn't make sense to me. Let them use Word or whatever, it'll save both you and them a lot of headache in the long run.

[1] Crossword puzzles: http://www.tex.ac.uk/CTAN/macros/latex/contrib/gene/crosswor...

[2] Music: http://www.mab.jpn.org/musictex/mtex_lib_e.html

[3] Animations: http://pages.uoregon.edu/noeckel/PDFmovie.html

Journal editors who accept LaTeX source from authors, desktop publishers/layout specialists who may know their field of expertise and are familiar with Word/DTP software, but not LaTeX syntax - it's more efficient to quickly fix a small item on-site and request the author's OK than ship the source back, have them fix it, take another look, send it back again, etc. Sometimes, you find an issue in the nth hour before publishing, so the ability to make a small change and get an assent makes or breaks getting that article into the run.

Making scientific authors who include a lot of equations in their work use Word because the journal can't handle dealing with LaTeX is just cruel. :)

I agree that the error messages are dreadful, but I've found a pretty simple pedagogical approach to make them less terrifying for a new learner. The key is that you reinforce a habit of frequently hitting the "typeset" button that way, when something goes wrong, they know to look at the commands that they just typed.

If you combine that habit with an explanation of the more common errors (like forgetting an "end" tag, or using improper nesting), then the error messages become a lot less scary. The learner, at least initially, doesn't have to try to understand what the message means, just how to deal with it.

Even for non-technical documents, I would still choose Latex for its advantages in areas like references and index, amenability to line-oriented version control, and the quality of the typesetting. I agree that non-technical users might make a different choice and that Latex error messages are dreadful.

I wholeheartedly agree with you.

I use LaTeX exclusively (with Emacs as editor) for non-technical documents (I study languages), and for me, the most relevant advantages are tightly integrated bibliography and citation management, index and shorthand capabilities and the chance of "line-oriented version control", i.e. LaTeX in git.

And the thing for introducing beginners, help them set up a preamble with the packages they will need. My experience with LaTeX has been that no matter how long you have been using it, there will always be something you will have to alter or look up. But a basic setup is not hard to deduct and by picking the packages relevant to you, you keep your files lean and get to understand what each part does. If you then have other problems you can dive into weird code snippets that only make sense since someone published them saying "Ermh...look...it's complicated...but...works, so copy & paste & be happy".

Once you've set up a basic preamble, people can actually concentrate on the content, and if they have problems, let them write their content first, the beauty of LaTeX is that all formatting woes can be addressed after the content is done.

I regretfully have to agree. When I was pursuing a masters in math, I used latex all the time, and I loved it. When I write anything about math or computer science, I prefer Latex to any other tool I tried. I got to a point I could type in Latex faster than I could hand write the equations.

But then I switched to law school, which required the use of MS Word. I found Word quite amenable to writing nontechnical documents, and it was much easier to explain how to make a table of contents in Word to my non-technical classmates than it would have been to explain Latex.

I love Latex within its field, but I do not see the point of evangelizing it to non-technical users.

I have had multiple people tell me that my site (https://www.sharelatex.com) has helped with this, firstly more advanced people are able to help learners, also the environment is already set-up.

The LaTeX wikibook is he best place I went to understand the basics of LaTeX http://en.wikibooks.org/wiki/LaTeX

Don't forget Lyx! In fact, I will take this opportunity to give a huge thank you to the developers and communities of Lyx, BibTex, BibDesk, Tex, Latex, Open Office, and NeoOffice. With their tools and online documentation, my wife just self-published a book, including a correctly formatted Kindle edition. She is generally not technical, but was able to figure out a workflow that worked for her and produced a really good looking product. I think she dealt directly with "code" exactly once (separating "front matter" from "main matter" in Lyx). Thanks!

(sorry for the self-reply) For the curious: Lyx working document + (BibDesk+BibText+LaTex+Tex) -> really nice PDF for printing. Lyx -> ODF -> NeoOffice -> Amazon Kindle toolset -> Kindle book. For backup/recovery, she relied on the built-in revision control in Lion (based on Time Machine, I think).

The not so Short Introduction to LaTeX2e (http://tobi.oetiker.ch/lshort/lshort.pdf) has been my go-to source for a long time. Works both as a manual and for reference.

I never "learned" LaTeX, despite being an academic mathematician and using it all the time. My documents have headers which I don't understand. Indeed, learning it from scratch seems like a hassle, and it has features far beyond what I ever use.

I simply obtained a document that was similar to what I was trying to create, made a copy, and then changed the content.

Sounds a lot like JCL.

I've long felt that LaTeX is unnecessarily complex for most users. The output is undeniably attractive but there's considerable pain involved in getting there.

It's because LaTeX is old and its ubiquity has caused the field to stagnant for decades. It's an area ripe for disruption, but it's tricky due to the amount of complexity involved.

My suggestion: teach people how to use Markdown and then use markdown2pdf (which first converts to Latex and then to PDF).

That doesn't work too well when writing for a conference or journal with a predefined template and style file. The obvious solution -- do all your writing in Markdown and then integrate at the very end of the workflow -- doesn't work too well because you can't tell how long the document is until you use the real template.

Same goes for org-mode's Latex export.

Actually, it is acceptably straightforward to include a custom header in org-mode. Basically you can create your own latex class in customize-group org-export-latex and then add a

#+LATEX-HEADER: \input{stuff.tex}

to your file to include the appropriate header magick. I'm currently using this approach to prepare my thesis in org-mode.

Pandoc (http://johnmacfarlane.net/pandoc/) is also very good, allowing you to embed LaTeX inside markdown. So you might do

    The greek for the letter _A_ is $\alpha$

mashing up the simple markup for bold, italic, lists, headers, and quotes from markdown with the more fully-featured LaTeX support for equations.

The big thing that is missing from most "here's how easy it is to do Latex" articles is integrating with pre-existing styles and templates. When learning Latex I was able to write my own documents with no problems. When I had to deal with (eg) A0 poster templates or conference-dictated styles I had a lot more problems.

i sense a disconnect in the premise. do non-technical people really need latex? - to invest the time to study prettyness of presentation... depends on definition, scope and degree of technical'ness but generally i'd think they (should) focus on quality of content in their own field because it's just much cheaper to outsource the nonessentials (latex being one of them) to more competent experts who dedicate lifetimes to 'making the presentation pretty'

I've been using LaTeX for over 20 years and I still do not exactly understand what I am doing. I use LaTeX for papers in mathematics and computer science myself. But even at my relatively high-functioning level of misunderstanding, I have been unable to persuade environmental scientists and chemists to use LaTeX instead of Word.

I'd argue that anyone using styles in Word is in the same boat.

Reply on news.ycombinator.com