Mrthfx

The most common mistake is spending too long on TeX coding and not getting the document written.

Here are some common, non-esoteric mistakes I help people fix on a regular basis:

1. Ending each and every paragraph in the document with \ (or even \[10pt]) instead of a blank line.




2. Ignoring warnings and errors until there's hundreds of them, and the new material isn't getting typeset at all. "Why can't I format my file? I have to finish chapter 2 by tomorrow."




3. Using font attribute commands (or old-style font commands) as if they take arguments: This bfseries{important} message... why is my entire dissertation bold?




4. Cobbling together a complex series of commands for some text element (a symbol, a way to format headings or tables, etc.), then using it over and over and over instead of defining a macro. Then realizing it needs to be modified.




5. Carrying along enormous preambles, donated by well-meaning friends, with no idea what they're for. Eventually there's a mysterious incompatibility, I'm asked to help, and I have no idea which packages are actually being used-- and neither does the author.




6. Relying on begin{table}[h!] (resp. {figure}), along with ad hoc page breaks etc., to typeset content that must not be floated. (But I hesitate to call this a mistake, because getting proper captions outside a float is not trivial: every package I've tried came with surprises.)




7. Not using floats when they would be appropriate, given the size (and content) of tables or figures. This definitely seems like a Word holdover...




8. Not capitalizing content words in BibTeX titles, and not protecting words that should never be lowercased: title = {A grammar of Late Modern English} will come out wrong under both capitalization styles.




9. Pet peeve: Setting entire words in math mode subscripts, ignoring the fact that they look awful. S_{easy}, S_{difficult}, S_{the worst}. Way too many linguists do this.



10. 
    

    This one's not common but I've seen it happen: Formatting section headers by hand, using literal numbers and manual spacing:

    
    
    

    \[12pt] 
    
    textbf{2.3 The importance of semantic mark-up} 
    
    \[10pt]
    
    

    
    

Not making enough use -- or (sadly all too frequently) not making hardly any use -- of the opportunities afforded by LaTeX to separate the content of a document from its visual appearance. In particular, too many attempts to engage in visual formatting at early to intermediate stages of writing a working paper, a technical note, or whatever.

Addendum: The answer by @Alexis gives quite a few examples of mistakes that arise when one engages in (too much) visual formatting...

DONT Read tutorials that advise best practices from the stone age.

Especially german LaTeX tutorials advise something like this in the preamble:

usepackage[latin9]{inputenc}

usepackage[T1]{fontenc}

Then i see a lot of people that are writing umlauts like this: "u. Please forget this!

DO In times of UTF8 use usepackage[utf8]{inputenc} and you're done. No need for magic to write foreign characters!

DO Another good advice is to use xelatex instead of pdflatex to compile your document. The advantage is, that you can change the fonts in your document quite easily, without shooting yourself in the foot!

DO Read and try to understand the build log! Looking for the source of compile errors is quite hard for TeX beginners (sometimes even for more advanced users), thus it's very helpful to get a look for relevant error messages!

UPDATE

DO Use the nag package (usepackage[l2tabu]{nag}) to get warnings when using bad practices or obsolete TeX-style commands. It is also helpful to read the documentation ( german "Sündenregister", english). After reading this, you can decide if a tutorial is using obsolete commands or other bad practices.

Not using version management for your document.

I've introduced some colleagues to LaTeX, and the most common questions I get relate to how suddenly something doesn't compile and how it can be fixed yet, when I ask them about reverting to a previous version, they consequently say they never thought of putting their work into a version management system (funny enough they manage all their other CompSys code related tasks in it). I now mention this even to beginners as one of the major benefits of working in plain text.

Often beginners are not aware of the different whitespace and possible linebreaks:

Cancelling an "end of sentence" space

A space right after a period following a lowercase letter by default ends a sentence and LaTeX inserts an extra whitespace. There are several occasions where you do not want to have the default behaviour. For example,

Ms. Bean is ldots\

Ms. Bean is ldots

To make the Ms. Bean unbreakable, use ~ instead of <space>. <space> and a ~ differ only in their line-breaking behaviour, the whitespace is the same. See also: When should I use non-breaking space?.

Enforcing an "end of sentence" space

A space right after a period following an uppercase letter by default represents an acronym and LaTeX does not insert an extra whitespace. There are several occasions where you do not want to have the default behaviour. For example,

I left at 12:00 P.M. In ldots\

I left at 12:00 P.M@. In ldots

produces

Enforcing a space after a control word

Control words eat spaces that follow.

LaTeX is fun.\

LaTeX is fun.

The <space> here is necessary to produce space between LaTeX and "is".
An alternative is to use braces to terminate the command. For example, LaTeX{} is fun. and {LaTeX} is fun. are equivalent to the above. See also Spaces after Commands.

Using $$ for italics. I find this error so often that I am suprised it has not been already mentioned.

Not using LaTeX itself, but rather obsolete TeX. For example, ${rm text}$ in math mode, or bigskip {bf Section} instead of section. The list goes on. This is a result of learning TeX from copying ancient documents without understanding, following one of the many contradictory guides online, or simply not caring as long as it looks as you expect.

A very common mistake for LaTeX newbies : they don't ask questions on tex.stackexchange.com.

I will hammer home my case for books! This is LaTeX!

Besides not using Emacs, I guess 90% of newcomers don't buy books, but try to learn "LaTeX by Google". What a waste of time.

Edit: OK, borrow books, also possible. Somebody else with such proposals? Maybe I'd betters say that studying a book as a beginner seems the crucial thing.

»Works pretty well«: Sure, but it takes much more time than take the book from the shelf and read. How many users never heard of »texdoc packagename«? And I'm quite convinced that "LaTeX by Google" works for nerds (but it surely costs much more time), but if you are studying the humanities, a book explaining the concept is really helpfull.

Edit 2: »Books are sooo last millenium«: Yess, and probably 99% of the users here as well. Besides that, courtesy of F. Mittelbach we present: The companion as ebook!

• Not realizing an empty line is a par and trying anything and everything to work around the spacing problems. This is probably among the top five and I would patch that out if I could.


• Using inline math instead of display math and screwing up the line spacing in progress. Points for using displaystyle in an inline math in a normal paragraph.


• 
  

  Using random solutions/templates/code pieces from the internet. Yesterday I saw someone actually confused about why using

  
  
  

  documentstyle{foo}
  
  usepackage{bar}
  
  

  
  
  

  throws an error, after he found the first line on the internet. I wish there were a way to get rid of the old pages that still show up with latex 2.09 code, yes weinelt.de you first. This goes further than @klingt.net. @Keksdose's solution is a remedy to a certain point, because the printed material is usually of better quality, but even there outdated solutions are advertised sometimes.

  
  


• 
  

  Using babel without the shorthands turned off and then being surprised that it messes up math, ipv6 adresses, listings or any other random thing. This 'feature' is my personal pet peave.

  
  
  

  Examples:

  
  
  
  
  
  • 
    Problem with spanish babel package
  
  
  • Conflicts with Datatool and babel (french)
  
  
  • tikz declare function and babel french option
  
  
  • Why does the package babel[french] destroy @for loops?
  
  
  • Tikz and babel error
  
  
  
  

  I've also seen problems with other packages in the past, listings and syntax from mdwtools come to mind immediately, but there are more.

  
  



• Accidentally redefining commands with def.




• Using incompatible packages or order sensitive packages in the wrong order. To be honest latex should make that harder or prevent that. Bonus points for obscuring the order to scrutiny by including a package multiple times.



• 
  

  Including redundant packages. Recently seen preamble:

  
  
  

  usepackage{lmodern}
  
  usepackage{mathptmx}
  
  usepackage{ae}
  
  usepackage{courier}
  
  usepackage{mathptmx}
  
  usepackage{fontspec}
  
  

  
  


• Manually trying to set variables instead of using proper packages. Most common with parskip, setspace and geometry.


• Declaring one encoding and using another or mixing different encodings.


• Ignoring Italic corrections.


• Using the wrong quotes (e.g. "" instead of ``'').


• Floats where they make no sense at all. Today I saw someone combine a table with Absolutely, definitely, preventing page break just to get a caption and latex was not happy.


• Not using non breakable spaces, where needed Dr.~Foo
  


• Not knowing when to use math and when not. Also not using text{} in math mode for non math.


• Not escaping _, #, ^ and friends.

In my experience, people come from Word and expect a LaTeX editor to work like MS Word, with the exception that they have to click a button to generate the document.

Because of this, many do not understand the concept of compiling properly. They do not understand what happens in the background; That first, pdfLaTeX is run, then BibTeX, then maybe another two runs of pdfLaTeX. They do not even understand what pdfLaTeX or BibTeX are, or the meta concepts of these, namely that there can also be LuaLaTeX, biber, or xindy!

This commonly leads to them being unable to fix problems. In my opinion, such basics of editing a document in the world of TeX should be treated better in existing documents. I think it would actually help if people would start out in NOtepad (or equivalent), and compile through the command line.

Fight typesetting with tooth and nail.

Many try to emulate what they know no matter what. Examples include:

• Switch to Times New Roman. What, you can change the math font?


• Use \ liberally to break lines and delimit paragraphs.


• Ignore the concepts of floating environments.


• 
  $...$ is the only math environment you know; spacing with \.


• Never use -- or ---, let alone . and , in acronyms.


• All math delimiters have the same size.

Not exactly typesetting, but related in spirit:

• What are label and ref?


• Instead of using the ones from babel, use wrong or hacked quotes (language dependent).


• Literature references hardcoded in footnotes.


• Re-use pixel graphics created with some drawing tool.


• Write everything in one huge file.


• Use a *TeX distribution that is older than your PC.

Of course, the most common mistake is to write stuff yourself. In 95% percent of the cases, there's a package that does what you want!

Using left and right indiscriminately before all delimiters (parentheses, brackets, braces). Some front-ends to LaTeX do it by default, but this behavior must be disabled.

Spaces around binary operation and relation symbols are flexible (for operations they are both stretchable and shrinkable, for relations only stretchable), so they can participate to the space adjustments made on a line for achieving justification. However, left and right create a subformula with the material in between them and, by rule, spaces inside a subformula are “frozen” to their natural width.

Another aspect worthy being mentioned is that a subformula is not breakable across lines; TeX is reluctantly willing to break a line after an operation or relation symbol, but not when this symbol appears between left and right.

These two commands do have their proper usage, but too often they are misused.

1. Compiling too early and too often.
Now you are spending your time working on fixing formatting and such that will change as soon as you add more text anyway and all the floats will move anyway.

2. Using floats when you don't have to. Floats are the devil, yet all tutorials use them. Most of the time I know where I want the image, at least roughly, and didn't know for ages that I could just insert the image without wrapping it in a float, so I'd spend ages trying to get stuff to stop floating down multiple pages, or into the wrong section, etc.

Edit: As this is attracting a lot of debate: When I started every example shows you to insert an image in a float, so I thought you had to use float. Even a couple of months ago I thought you needed to put a table in a table environment; I'd still have to look up how to make a non-floating table. This is a problem as a lot of the time I need an image in a certain place. For example, if I'm typesetting a homework problem (Something I did a lot as an undergrad, and am now doing again when I'm writing questions) I can't have that image floating down into a different question; my TA or prof isn't going to go hunting for it, they are just going to dock me marks (And I don't want to confusing anyone working on my problem set!). Also as the h option is a joke they usually float far away. I don't mind if a figure floats a few paragraphs, but three pages away is bad: Same page as the discussion or facing page only. I should always be able to see both figure and text at the same time, unless there are exceptional circumstances or a very large text (I cut some slack when writing my thesis, as there were so many figures they wouldn't all fit on the same page as the text).

There are times floats are the right choice; I'm using one in the document I'm working on right now. However, they are just as often the wrong choice, and a lot of new users don't realize this.

3. Using outdated packages that some website gave you My group drives me crazy since they use very outdated packages that have been superseeded. subfigure instead of subfig, stuff like that.

4. Relying on GUI text editors instead of understanding what is happening People at work also drive me crazy with refusing to use anything but the compile button in WinEdt. If that doesn't automatically work, they refuse to use the package (Thus no biber, no biblatex, etc.

5. Using MikTeX That install package on compile thing NEVER works right. Hard drives are large, dammit.

Edit: Ok, not never. But it fails every time I've given my coworkers a new package to use, causing us to have to go into the interface and have it manually install.

6. Limiting yourself to what chemistry journals allow for everything I dream of one day being editor of a big chemistry journal, just so I can go to the AMS or APS and borrow their code. This has things like not using macros, not redefining things, not using excess packages (or in one case, not using packages AT ALL).

7. Not being willing to put in time understanding LaTeX The people I work with just want a solution now. They aren't willing to put in any time into making it work; they want something that works now, since the journal will reformat it anyway, so if it works and gets vaguely close to what they want, good enough.

8. Making DocumentNameV3.tex and forgetting to open the new PDF, then wondering why nothing you do is working Not that I've done this recently, after I was too lazy to set up a new code repository.

Incorrectly choosing the compiler

• 
  

  Compiling

  
  
  
  
  
  • an input file importing PDF, JPEG, PNG
  
  
  
  

  with latex->dvips->ps2pdf sequence. The compiler should be either pdflatex or xelatex (among other possibilities).

  
  


• 
  

  Compiling

  
  
  
  
  
  • an input file containing PSTricks code
  
  
  • an input file importing EPS
  
  
  
  

  with pdflatex. The compiler should be either latex->dvips->ps2pdf or xelatex (among other possibilities)

  
  

Forgot to break the paragraph before the closing }

When a paragraph has font size change, the par must be invoked before }

documentclass[preview,border=12pt]{standalone}

deffoo{%

These dummy texts will span more than 2 lines. If you see they do not span more than 2 lines then you have to inform me now!}



begin{document}

parindent=5emrelax

foo



{huge foo}



foo



{huge foopar}



foo

end{document}

Not understanding the compile sequence when using GUI tools like TeXmaker and LyX was something I was personally guilty of. Understanding this better, helped me understand when references, citations and page numbering went wonky. Switching to emacs and using a custom scripted compile helped sort out the kinks.

My own take on this question is more about the mistakes beginners make because of us. I made too many shameful mistakes (and still do at times), but with my first private beta, I realised that many beginner mistakes were in fact my beginner LaTeX developer mistakes.

They are still beginner mistakes – but I have a feeling of responsibility for a lot of them.

1. Not teaching users to first read the documentation. Not teaching users to look at the code if the documentation is unhelpful (code does not bite). I have had someone tell me "I didn't dare to modify your package because I didn't know what I was doing"… well, just back it up somewhere and play, code is not sacred).


2. Not teaching users to send you the log when they want to report a bug or any other issue, even if they do not understand what the error message means. On that matter, I feel like it is helpful to teach all beginners how to read the most basic errors (command already defined, command not defined, underfull and overfull boxes).


3. Not writing appropriate documentation files for users (we all boast about "literate programming" and we mostly do a good job at writing docs, but not all of them are perfect). We have some beautiful files, and also some barely-commented "implementation" docs with just an introduction and a list of macros that no beginner would ever understand.


4. Not teaching users the difference between form and content (as mentioned several times before). But I also mean it literally: whether they want to code something for themselves or whether they are requesting a new feature, they need to determine if it is just a stylistic preference or if there is a general purpose behind it (e.g. do we need a simple command or an option for customisation). There is nothing more stupid than creating a whole system that works really well except it is totally impractical in real life.

I hope it is not off-topic… but thinking back, I would think that most of my own beginner mistakes were also often developer (or documentation-writer) mistakes. We are generally good at writing documentations, but we sometimes have a hard time finding the right words for both beginners and experts.

Most of the beginner mistakes mentioned here are not really in the books… or at least they don't show up as sections, tutorials, etc. It is no wonder beginners first look things up on google. We don't write documentations on debugging what we assume will work (it works in the hands of a TeXnician, not in the hands of someone who does things at random).

The usage of raster graphics (JPG, PNG, ...) instead of vector graphigs (SVG, TikZ pictures, ...) could be a mistake since this increases output PDF file size and slows down rendering in several readers.

Compile early, compile often as with 'test early, test often' is good advice when spending a lot of time working with complex documents and multiple included files. A common mistake is to work too long on a particular section, not noticing a missing } and then having to scour multiple documents for hours, when this could've been fixed by either an automated constant compile, or just ensuring every section or paragraph you work on still leaves a coherent document.

• Not understanding the difference between typesetting and word processing. The latter is simply using computer technology to replace a mechanical typewriter, primarily in the production of routine business documents. The former has to do with preparing documents in their final form for publication. TeX and Friends are typesetters, not a word processors.


• Thinking TeX and Friends are solely for preparation of documents intended for scientific/academic journals, classroom notes or textbooks with lots of formulae and/or footnotes and citations. TeX and Friends certainly excel at these tasks but also excel at typesetting poetry and prose -- in their original form, not just in textbooks or critical analysis (although TeX and Friends do an outstanding job of preparing those as well).


• Not understanding the difference between logical markup and WYSIWYG then spending inordinate amounts of time trying to make TeX and Friends behave like a WYSIWYG desktop publisher/word processor instead of putting their focus on the content of their document.


• In a similar vein, not understanding the difference between writing a document and designing a document. Two different disciplines, two different concerns. TeX and Friends allow a clean separation of the functions even if the same person is going to do both.


• Trying to make TeX and Friends do everything instead of using the right tool. Production of a consumer-oriented magazine is far better done in Adobe InDesign than TeX and Friends. Business presentations are far better done in PowerPoint than Beamer. etcetera, etcetera etcetera.


• Underestimating the amount of work it takes to produce a "beautiful" document. TeX and Friends will produce more beautiful documents than MS Word but it is not automatic simply because one uses TeX and Friends.


• Getting caught up in vim vs emacs vs texworks vs tex{studio | maker} vs notepad arguments on SX (or elewhere). :) They all work fine. TeX and Friends don't care which you use and the resulting documents don't either.

Doing by hands what LaTeX should handle. Like writing "on page 3" and thinking to learn "complicated things" like label-ref "later", i.e. never. On the other hand, it is not bad to use newpage before section instead of titlesec package --- this is what David Carlisle already answered about.

Basically this applies to all kind of systems. How many keyboard shortcut in Word to learn, how many packages to remember in LaTeX... It always depend on how much one is going to use some software.

Another one on meta-level: not asking computer support when you have one available. Many have used hours to find out things that I could have tell in a minute.

The most common mistake that LaTeX beginners make is that they start with the “stock” article class and enrich it with various packages to suit their needs, instead of picking a more specialised class providing all the functionality they need.

It has several drawbacks for the users:

• They spend a significant time trying looking for “tabular package XY” that can do “plop plop fizz fizz” instead of using what is provided by the specialised class they did not pick.




• They spend a significant time trying to make some packages work together and looking for answers to their problem.




• They spend a significant time trying to customise the class to achieve a pleasant layout, while they wether have the needed programming skills neither (most probably) the design skills.

Is there any LaTeX introduction or tutorial that emphasises the choice of a document class and provides useful resources (catalogue, comparison of features, application domain) to choose one? It seems to me, most of them pick one (article) and stick to it for the whole tutorial, while it could be more useful—once the basics of text and math typesettings have been covered—to drop that generic class in favour of a specialised one.

To put it in one sentance: “unexperienced users confuse the roles of designer and scripter (and tutorials tend to perpetuate this confusion).”

Judging by questions on this site - which I admit are probably not representative of beginners as a whole but which are, at the very least, of special interest to readers likely to peruse this answer - I would say that one basic mistake, which leads to a number of others, is not stepping back and getting a general sense of the capabilities of TeX & friends. I don't mean the details, but the 'big picture'. For example,

• TeX & friends are good tools for typesetting text, mathematics and bibliographies;


• TeX & friends are great for automatically ensuring consistent layout, accurate references and cross-references;


• TeX & friends are not the best tools to create a database, manage a spreadsheet or draw elaborate 3D scenes with accurate perspective and lighting;


• TeX & friends are not good replacements for tools such as sed, (g)awk and grep, although many editors provide much of their functionality;


• neither TeX nor friends are good replacements for a sheepdog;


• TeX & friends make at best indifferent tea.

In particular, TeX & friends are much better than human beings - and much better than many other programmes - at the things they do well. They are not generally good - and typically much poorer than alternative programmes - at the things they are not intended to do.

One key strategy for getting the most out of TeX & friends is to use the right tool for the right job. Some questions boil down to requests for TeX solutions to the problem of brewing the perfect cuppa. This is a perfectly reasonable aim but the best answer involves advising a non-TeX solution.

Other questions essentially ask for TeX-based cat-herding solutions. The best answer in this case involves recommending against adopting the goal set out in the question.

However, these types of answers are rarely appreciated and frequently interpreted as either lacking in imagination ('I don't believe it is not possible/advisable') or as admissions of TeX's failure to rise to the needs of twenty-first century users ('Super Programme can make tea while simultaneously sheering sheep, sending a letter to my gran and paying the gas bill'). This is a shame since these really are the best answers in these cases and appreciating TeX's limitations, as well as its strengths, is key to using it effectively.

[That is, I am extremely sceptical of solutions of this kind.]

Cup of tea is from openclipart.org.

'Like Juggling While Herding Cats' is by Robin Catesby and available here under this CC licence.

Cleaning too early

• Suppose there is an input file importing EPS images and it will be compiled
  with latex->dvips->ps2pdf. Some users think that the EPS images are no longer needed after running latex so they can be deleted. Actually it is wrong because dvips still needs the EPS. (it must be documented in arara)




• Suppose there is an input file using cross-referencing with label. Some users aggressively remove any auxiliary files generated by (pdf)latex compiler before the next invocations. It makes the cross-referencing always end with ??. These auxiliary files must not be deleted until the cross-referencing is properly done. (it must be documented in arara).

Another common misbehavior is reading package or class documentations from TeX mirrors (the worst case, they read from sites with obsolete contents) even though they already installed the complete, up-to-date packages and classes.

I mean that they don't know that texdoc <package-name> invoked in their own machine can launch the documentation in question.

Actually it is not wrong but the output might make your eyes a bit itch. "Wrongly" representing a series of unaligned equations using [..] instead of using begin{gather*}...end{gather*}.

documentclass[preview,border=12pt]{standalone}

usepackage{amsmath}

begin{document}

section*{Not recommended}

The solution of $x^2-5x+6=0$ is

[

(x-3)(x-2)=0

]

[

x-3=0 text{ or } x-2=0

]

[

x=3 text{ or } x=2

]



section*{Recommended}

The solution of $x^2-5x+6=0$ is

begin{gather*}

(x-3)(x-2)=0\

x-3=0 text{ or } x-2=0\

x=3 text{ or } x=2

end{gather*}

end{document}