You probably compiled the following minimal example:
\documentclass{article}
\begin{document}
Hello World!
\end{document}
which provided the following (partial) .log:
...
\c@part=\count79
\c@section=\count80
\c@subsection=\count81
\c@subsubsection=\count82
\c@paragraph=\count83
\c@subparagraph=\count84
\c@figure=\count85
\c@table=\count86
...
The above output stems from counter definitions within the default article document class. More specifically, the following set of commands (just search for \newcounter in article.cls):
...
\newcounter {part}
\newcounter {section}
\newcounter {subsection}[section]
\newcounter {subsubsection}[subsection]
\newcounter {paragraph}[subsubsection]
\newcounter {subparagraph}[paragraph]
...
\newcounter{figure}
...
\newcounter{table}
...
Each \newcounter{<newcntr>}[<oldcntr>] defines a new counter as a macro \c@<newcntr> - note the prefix c@. Here's the definition of \newcounter from the LaTeX kernel:
\def\newcounter#1{%
\expandafter\@ifdefinable \csname c@#1\endcsname
{\@definecounter{#1}}%
\@ifnextchar[{\@newctr{#1}}{}}
At first it checks whether a macro \c@<newcntr> already exists, and then defines it via \@definecounter{<newcntr>}.
\def\@definecounter#1{\expandafter\newcount\csname c@#1\endcsname
...}
\@definecounter issues the TeX primitive \newcount to create the actual count register that \c@<newcntr> points to.
The use of the c@ prefix is purely a conventional choice. In a similar manner, counter prefixes have a p@ macro prefix, while references (or \labels) in the LaTeX kernel have an r@ prefix. It attempts to avoid end-users from messing around with the counters, yet still allow them to create their own similarly-named macros. That is, making the following work without problem:
\newcounter{something}% <------- Creates \c@something
\newcommand{\something}{..}% <-- Creates \something
If the c@ prefix wasn't inserted by \newcounter, then \newcounter{something} may very well have created a counter macro \something, which is problematic when you want to create a real macro \something.
Anki uses some temporary directory for compilation, so you have to give the path to your images explicitly. Suppose the directory is /home/gernot/anki and contains an image duck.jpg. Set the preamble in Anki via Tools | Manage Note Types | your cloze note type | Options to
\documentclass[12pt,border=1mm,varwidth]{standalone}
\usepackage{graphicx}
\graphicspath{{/home/gernot/anki/}}
\begin{document}
Add a card of your cloze note type with the following contents in the Text field:
[latex]\begin{center}\includegraphics{duck}\\\relax {{c1::duck}}\end{center}[/latex]
Then Anki will generate the two cards
Some things to observe:
The path in \graphicspath has to end with a slash /.
Bare in mind that strings like {{c1::XXX}} will be replaced by [...] on one card and by XXX on the other one. You have to make sure that your code is correct in both cases. In the example above, writing \\{{c1::duck}} gives an error since \\[...] is interpreted as \\ with an optional argument. Writing \\\relax{{c1::duck}} still gives an error, since the string becomes \\\relaxduck on the second card and \relaxduck is an undefined command. Therefore one has to put a space, a line end or {} inbetween.
If the LaTeX code happens to contain double braces {{, it may be mistaken as Anki notation. Replace it by {%+new line+{ or, if acceptable from the LaTeX point of view, by { {.
If the path or the file name of the image contains special characters (spaces, tildes, dots) one may have to take extra precautions. See How to use \graphicspath? \includegraphics: Dots in filename; but remember the last point about avoiding double braces.
(duck image from Wikipedia)
Best Answer
This is defined in
IEEEtran.cls(see http://www.ctan.org/tex-archive/macros/latex/contrib/IEEEtran/) and set to true of a PDF flow is used: the class file checks whether\pdfoutputis (which seems to be the case inpdflatexas the comments suggest there):