|developer||Donald E. Knuth|
|programming language||Web , Pascal|
|License||own liberal license|
TeX ([ tɛx ], in German and space [ tɛç ] ), in-house spelling , is a by Donald Knuth in 1977 developed and 1986 finished typesetting system with built-in macro language (which also TeX is called). In a narrower sense, TeX is a single program (an executable file or the underlying source code on a computer ) that reads simple text files and font files and creates a binary document file (a graphic format , more precisely a page description language ) that is printed out and displayed by other programs or otherwise processed for an output; the functionality of this program has been “frozen” since 1990, and nowadays it is mainly emulated by further developments such as pdfTeX . In a broad sense, the TeX typesetting system consists of further developments of this program, character set files, thousands of packages that can be used as input texts , and many other programs for pre, post and intermediate processing (sorting register entries, generating the bibliography from a database, Screen display, ... ). The easiest to learn and also best maintained access to TeX for different combinations of requirements is to use the macro package LaTeX (including its extensions), so that for the majority of users “TeX” and “LaTeX” are in a sense the same.
Pronunciation and origin
TeX is an abbreviation of the Greek Τέχνη (from ancient Greek τέχνη (téchne), ability, artistry, craft). The letters stand for the Greek letters tau , epsilon and chi (Τ, e and Χ). In the representation, these are often identical to the Latin T, e and X, but the latter differs in the pronunciation.
According to Knuth, the X should be pronounced as an Ach sound, such as in the German word ach , ie: [ tɛχ ]. This sound is unusual for many German speakers after a front vowel like e . That is why the pronunciation [ tɛç ] with ego sound as in tech nik and other words derived from Τέχνη has become common in German.
TeX can be used for all types of texts, from short letters to multi-volume books, although TeX was originally developed for longer texts and academic papers. Many large scientific publishers use it for book printing or work typesetting . A particular strength is the mathematical formula set as well as the generated typeface.
From a technical point of view, TeX is an interpreter that knows about 300 built-in commands (so-called primitives ) and provides a complex mechanism for defining your own macros . TeX creates so-called DVI documents from appropriate source files ; newer versions can also generate PDF files directly . The text is set paragraph by paragraph, with all line breaks (including word separation ) in the paragraph being determined at the same time using a separate optimization process. A simple algorithm then determines the optimal page break. The optimization procedures work with evaluation points (penalties), on the basis of which the possible breakpoints are evaluated. Line and page breaks are controlled by a large number of parameters, especially registers and elastic lengths, which can also be changed in the document (a more detailed description is given below ). TeX itself gets by with a minimum of resources and only provides the basic functions.
TeX is supplemented by a large number of software packages that enable an efficient and comfortable use of TeX. Plain TeX , which Donald E. Knuth developed himself and which is described in detail in Knuth's introductory book The TeXbook , is a relatively simple set of software that fulfills the most important general basic needs . The LaTeX, originally developed by Leslie Lamport , is far better known and more user-friendly . Based on this, there are several thousand additional packages for every imaginable area. With the hyperref package , for example, a hypertext can be created without much additional effort, which enables navigation in the table of contents and in the index in a PDF reader, as well as links to the Internet . The BibTeX program enables the creation and management of source information in TeX texts, MakeIndex generates page number lists for the respective term entries from “raw” index entries of the form (term, page) . In the meantime, modern utf8-capable alternatives to these tools have been developed.
"Extensions", "Formats", macros
"Extensions of TeX" can be distinguished as follows:
- Further developments such as pdfTeX , which exist in a single program (as a binary file) and extend the functionality of the TeX program (see below );
- Formats in the sense of typically extensive macro collections such as the above-mentioned plain TeX and LaTeX (“format” should not be confused here with “ file format ”, more follows);
- Additional packages in the sense of smaller macro collections (text files), which are mostly tailored to a single format , such as LaTeX extensions ; the mentioned hyperref is such a “LaTeX extension”.
A macro is identified by a “command”, usually beginning with a backslash, and an associated replacement rule . For example, under plain TeX and LaTeX
\unhbox\voidb@xreplaced by. When a format is referred to as a “macro collection”, it is a collection of corresponding replacement rules. Only with such a format does TeX become manageable for practical applications. ("Lifelike" examples of macros can be found under LaTeX # Structure of a document .)
A TeX format is typically identified as follows:
- with a text file containing the macro replacement rules - for plain TeX / LaTeX the file
latex.ltxline contains a
- by a binary file with the ending
latex.fmt), which represents TeX's memory after reading the macro collection file (but also contains memory size parameters and language-specific separation rules);
- through additional packages tailored to the format (e.g. LaTeX extensions) - in the "TeX archive" of the Comprehensive TeX Archive Network , the directory contains
macros/subdirectories for individual formats in which the additional packages tailored to them are listed;
- through available guides - there are books about plain TeX like The TeXbook and books about LaTeX and its extensions like The LaTeX Companion , but no books about plain TeX and LaTeX.
- A TeX format is something in the middle between a text file and a binary file. In principle , TeX could simply read in the “macro collections” plain.tex or latex.ltx , then the additional packages and finally the files with the text of the document to be displayed. In the years around 1990, however, reading and processing the macro definitions took so many minutes that it made sense to read in the memory image resulting from the value assignments instead of having this done for each TeX run from the macro collection file.
- Appendix E of The TeXbook lists examples of smaller TeX formats . A " human-readable " representation from plain.tex is printed as Appendix B (with technical omissions).
- In contrast to format-specific additional packages, there are also packages that can be used with different formats; such can be found in the “TeX archive”
Besides the TeX formats plain TeX and LaTeX , ConTeXt is also well known. Compared to LaTeX, it has more functions (such as hypertext support) built in, on the other hand, fewer additional packages are available than for LaTeX. However, ConTeXt is not an example of a format that can be used with Knuth's TeX program , it only works with "binary extensions" from Knuth's TeX, see ConTeXt # Mark II and Mark IV .
The following example shows the steps to set text in plain TeX .
First a text file (for example
Text.tex) with the following content is created:
The text to be set is in this case
Hallo. The instruction
\byeis a TeX command that specifies the end of the file and does not appear in the output.
Then you give the command in a command line interpreter
a. TeX creates the file
dvistands for "device independent", ie "device-independent").
dvifile can be printed out directly from the display program or converted into a printer format ( e.g. PostScript , with the
History and background
Volume 1 was published in 1968, volume 2 in 1969. The artwork was set using the monotype technique. This type of formula set was time-consuming. After Volume 3 was published in 1973, Knuths Verleger sold his Monotype machines. The corrected new editions of volumes 1 and 3, which appeared in 1975, had to be set in Europe, where some monotype systems were still in use.
The new edition of Volume 2 was supposed to be produced in 1976 with photosetting , but the quality of the first samples disappointed Knuth ( Digital Typography , Chapter 1). He had put 15 years of work in a row and only wanted to continue when the books were properly set. In February 1977 there was a way out when Knuth was presented with the output of a digital printing system with 1000 dpi resolution as part of a book evaluation . Pat Winston had written a book on artificial intelligence with it. When Knuth realized this, he interrupted work on Volume 4, of which he had completed the first 100 pages, and decided to write the programs himself and his publisher needed to re-set Volume 2. TeX design began on May 5, 1977.
Knuth estimated the necessary work to be a few months. In May 1977 he wrote to his publisher that he would have the first templates ready in July.
After four years of experimenting with a Xerox typesetting system, he had not yet achieved a result that surpassed phototypesetting. Knuth didn't give up and met well-known font designers , including Hermann Zapf (who was developing the hz program at the time ), from whom he learned. After five more years of work, he had reached a level with which he was satisfied. The completion of TeX was celebrated on May 21, 1986 at the Computer Museum, Boston , Massachusetts .
In the books of the TAOCP series, Knuth deliberately wanted to record part of the knowledge of computer science, which he assumed had already reached such a level of development that this knowledge would still be presented similarly in a hundred years. It was therefore very annoying for him that the typographical quality deteriorated with the techniques of the time from edition to edition. It was not planned that the TeX project would then take almost ten years of his time.
“Ever since those beginnings in 1977, the TeX research project that I embarked on was driven by two major goals. The first goal was quality: we wanted to produce documents that were not just nice, but actually the best. [...] The second major goal was archival: to create systems that would be independent of changes in printing technology as much as possible. When the next generation of printing devices came along, I wanted to be able to retain the same quality already achieved, instead of having to solve all the problems anew. I wanted to design something that would be still usable in 100 years. "
“Since the beginning in 1977, the TeX research project that I started with had two main goals. The first was quality: our printing units should not only be 'really nice', but actually the best […] The second major goal was archivability: the system should be as independent as possible of changes in printing technology. Even with the next generation of printing presses, I wanted to maintain the previous output quality and not have to solve all the problems from the beginning. I wanted to design something that could still be used 100 years from now. "
Knuth wrote a five-volume series on the programs of the TeX system and their use:
- One volume each for the source code (with comments) for the typesetting program TeX and the font generator Metafont ,
- one volume each with user instructions for the use of TeX and METAFONT,
- as well as a volume with the METAFONT source programs for the letters of the Computer Modern fonts.
The version number of TeX approximates the circle number by adding a digit to each error correction. The current version number is 3.14159265. According to Knuth, there will not be a version number 4 because - although he himself has already made some suggestions for additional functions for TeX - he considers it more important to have a stable system that will still deliver the same outputs 100 years from now today. However, this did not prevent Knuth from spreading the appearance of a new version with the name “i-TeX *” in a satirical speech in front of the TeX Users Group on the occasion of TeX's 32nd birthday.
Furthermore, Knuth has decreed that TeX should not be further developed by anyone after his death. Also, the version number should be changed to, and any bugs that would still be present at the time of his death should be recognized as features. Analogously, the version number of approaches Metafont the Euler number e .
Novel aspects of TeX typesetting
The typesetting generated by TeX had some aspects that were novel at the time of TeX's release or of lower quality in other typesetting systems. Some of these innovations are based on interesting algorithms that were the subject of several theses by Knuth's students. The rules for spacing in mathematical formulas are unique to this day; others of Knuth's discoveries are now used in various typesetting systems.
Distances in formulas
Since Knuth's main focus during the development of TeX was on high-quality text typesetting for his book The Art of Computer Programming , he attached great importance to well-functioning rules for the choice of distances in mathematical formulas. As a basis he used what he believed to be examples of high quality mathematical typography: the books set by Addison-Wesley , editor of The Art of Computer Programming - especially the work of Hans Wolf - editions of the mathematical journal Acta Mathematica around 1910 and a copy of the Dutch journal Indagationes Mathematicae . From the precise analysis of the text set in these works, Knuth developed a set of rules for setting distances in TeX. While TeX only provides a few basic rules for spacing, the exact parameters depend on the font used for the formula set. The rules for Knuth's Computer Modern fonts, for example, have been fine-tuned over many years and are now frozen. But when he first started using other fonts like AMS Euler , new spacing parameters had to be defined.
Word separation and justification
Compared to manual typesetting, the problem of typesetting in justification can in principle easily be solved with an automated system such as TeX. Such a system can automatically adjust the word spacing in a line if it has been defined where a line break makes sense. So the real problem is finding the line breaks that create the best overall impression. Many line break algorithms use what is known as a first-fit approach: the line breaks are set line by line and no line break point is changed once it has been set. Such a system cannot judge a line break point by the effect of the break on subsequent lines. In contrast to such algorithms, the Total Fit algorithm that Knuth and Michael Plass developed for TeX takes into account all possible line break points in a paragraph and finds the combination of those that produce the best-looking arrangement overall.
Formally, the algorithm associates a value called badness with every possible line break. The badness is increased if the word spacing has to be stretched or compressed too much to get a line to the correct length. Penalty points are added if a line break is particularly undesirable, for example if a word has to be separated, if two lines in a row end with a word separation or if a very stretched line immediately follows a very compressed one. The algorithm then finds the set of break points that has a minimum sum of the squares of the badness values. Contains a paragraph n possible breakpoints, so would a naive algorithm 2 n possible newline combinations are checked. Using dynamic programming methods , the complexity can be reduced to O ( n 2 ) . Further simplifications (e.g. not checking extremely unlikely breaks at all - e.g. a word separation of the first word of a paragraph) usually lead to a runtime in the order n . In general, Michael Plass showed that the problem of finding the best page break can be NP-complete due to the increased complexity caused by the placement of figures and tables . A similar algorithm is used to place page breaks in paragraphs in a way that prevents whores children and cobbler boys .
If no matching line break can be found, TeX tries to hyphenate a word. The original version of TeX used a word separation algorithm that recognized word prefixes and suffixes, removed them if necessary, and inserted hyphens between the two consonants of a letter combination vowel - consonant - consonant - vowel, which is mostly possible in the English language. TeX82 uses a new word break algorithm developed by Frank Liang in 1983 that prioritizes break points in words. First, a list of word separation patterns is generated from a large corpus of separated words (such corpora contain 50,000 words and more). For example, when TeX needs to find an acceptable hyphenation position in the word encyclopedia , it creates a list of the partial words of “.encyclopedia.” (The period is a special character that marks the beginning and end of the word). This list includes all partial words of length 1 ( ., E, n, c, y etc.), length 2 ( .e, en, nc etc.), up to length 14, the word including the periods itself. TeX searches in its list of separation patterns, selects those partial words for which it has calculated the desirability of separation. In our case, 11 such patterns are found:
|1 c 4 l 4||1 cy||1 d 4 i 3 a||4 edi||e 3 dia||2 i 1 a||ope 5 d||2 p 2 ed||3 pedi||pedia 4||y 1 c|
For each word position, TeX now calculates the maximum value from all matching patterns, which is here
- en 1 cy 1 c 4 l 4 o 3 p 4 e 5 d 4 i 3 a
results. The odd numbers mark possible separation positions; so here the result is en-cy-clo-pe-di-a . This subword-based system allows the definition of very general patterns like 2 i 1 a with low break numbers (even or odd), which, if necessary, can be skipped by more specific (longer) patterns like 1 d 4 i 3 a.
With a clever choice of parameters it is possible to create separation patterns in such a way that
- the number of elements is as small as possible - in 1983 still of great importance
- almost all possible separation points can be found (over 90%)
- the algorithm does not recognize any wrong separation points (in conjunction with an exception list that can be expanded by the user)
The disadvantages of Liang's algorithm can be seen in languages such as German, where there is practically any number of compounds that cannot be taken into account when generating the separation pattern and are therefore often incorrectly separated.
Status and further developments
The definition of the command scope of TeX has been fixed since 1990. It was Knuth's concern to create a high quality program; the program is considered to be completed; only error corrections are made. Knuth attaches great importance to the fact that real further developments (for example with an extended set of primitive commands) are not referred to as "TeX" (see license ).
Graphical TeX editors (see below ) provide front ends for such command lines. To this end, pdfTeX used. This program has a more extensive instruction set than TeX; the original TeX commands should work in the same way as with TeX, so that the macro package LaTeX and its extensions can also be used with pdfTeX (
pdflatex). Knuth's macro package plain TeX can also be used with pdfTeX (
pdftex), and The TeXbook can also be used as a reference for this (but no longer applies with regard to the number of registers). - Today, pdfTeX is also used in current distributions to generate TeX's original DVI output format (e.g. via
latex), TeX is only used for command line commands
tex. In this way one has access to the instruction set of ε-TeX , an extension of TeX, which preceded pdfTeX and z. B. provides a larger number of registers than TeX. In fact, further, previously independent further developments of TeX can be emulated with pdfTeX today . With the additional use of the microtype package , a refinement of TeX's line break algorithm is also accessible.
Operating system character sets and Unicode
TeX initially used its very own system of character sets ( generated with Knuth's Metafont ), but could then also handle PostScript fonts . The latter is initially not easy for authors who do not want to use the system professionally, but rather for their own texts. The character sets offered by an Office word processing program on a computer are not accessible. The use of non-Anglo-Saxon glyphs and even more of non-Latin writing systems is also somewhat inaccessible from the perspective of Office users. This is especially true if you are restricted to the plain TeX macro package , while LaTeX provides well-documented interfaces as standard to overcome these obstacles. TeX's first character sets made fewer than 128 glyphs available, with LaTeX 8-bit character sets with the usual encodings including UTF-8 are easily available. pdfTeX does not represent any improvement with regard to these two points. More natural is the access to fonts from outside the TeX world with XeTeX and LuaTeX , which make OpenType available and support input files encoded in Unicode by default . The TeXbook no longer fully applies to XeTeX and LuaTeX.
"TeX" as a "generic term"
Although the program developed by Knuth is hardly used any more, there are still "TeX distributions" and "TeX user groups" (as follows), and working with the current developments of TeX is essentially considered working with TeX considered and designated accordingly. XeTeX is hardly widespread among authors who practically only use Latin fonts, and pure authors (in contrast to “TeX experts”) are also not aware of these different “engines”. Since pure users practically exclusively use macros from the LaTeX and plain TeX formats and their extensions (and not primitive commands), they are hardly affected by the differences. The fact that certain macro packages only work with further developments of TeX (especially ε-TeX) and not with Knuth's program is also of little relevance, since the latter (especially pdfTeX) are supplied with common distributions and their use (also in TeX editors ) is preset.
User groups were organized around TeX very early on, the first being the TeX Users Group (TUG), which operates internationally. Language -related and regional user groups were added later, such as the German-speaking user association TeX (DANTE) for the German-speaking area and GUTenberg for the French-speaking area. In total, there are currently around two dozen TeX user groups that work together.
There are many different distributions of TeX that should be fully compatible with each other . They contain all the programs you need to work with TeX. Optional additions such as LaTeX are also included.
- MiKTeX is a TeX distribution (program collection) for Windows , since 2018 also for macOS and Linux .
- TeX Live is a TeX distribution for various Unices (including Linux ), macOS and Windows. It is created by the TeX user associations and is included in many Linux distributions.
- teTeX was a TeX distribution for Unix created by Thomas Esser , which has not been further developed since May 2006. Thomas Esser recommends interested parties to turn to the TeX Live project.
The directory tree texmf, which is generated by the various current installations, is standardized as the "TeX Directory Structure".
The collection of TeX versions and extensions in the Comprehensive TeX Archive Network (CTAN) is even more extensive than the distributions mentioned . TeX Live is limited to TeX add-on packages that are strictly free licensed , so that even packages are not included that can be downloaded “manually” from the CTAN, but do not contain a sufficiently precise license. The other currently relevant distribution MiKTeX , on the other hand, is limited to software that can run with the Windows operating system , while packages can also be downloaded from the CTAN that can only run on Unix-like operating systems or even only on "historical" / "exotic" operating systems.
The source code of TeX is openly available, it can and should be further developed according to the will of Donald E. Knuth. The output of a certain version of TeX should be identical on all platforms. Further developments by other authors may therefore not bear the name TeX themselves, but must be named differently. Although it is sometimes said that TeX and LaTeX were placed under the public domain by Knuth and Leslie Lamport, respectively, this is not the case.
The license terms of TeX are not to be confused with those of LaTeX - although they are similar. The LaTeX Project Public License (LPPL) also allows copying and changing the source code of LaTeX packages only as long as the changed files are distributed under a different name than the original files. The LPPL does not apply to TeX itself. It only emerged from the license that Knuth had specified for TeX.
Supporting tools (editors)
The work of authors or their typists with TeX goes like this (first consider the short #example ): You enter a passage of source code (text, commands for formatting and for special characters such as mathematical symbols) (and save the text file) then run the TeX program (working with pdfTeX, for example ) with the new text file version , and finally look at the result with a previewer (with MiKTeX, for example yap , with Linux Evince ) (you may also create another tool a PostScript or PDF version from the DVI file). You have two windows: one for entering text and one for previewing. This is very different from working with an Office program , where text changes are made directly in the preview window. Due to the TeX commands, the appearance of the code differs greatly from that of the generated document, which is a certain challenge for the editor (cf. LaTeX # No WYSIWYG ).
After all, LyX is an editor for LaTeX with which you can make text changes in a window that does not show any TeX commands, nor does it offer a real preview, after all, an intuitive indication of the result.
The working environment with separate windows for input and output can be designed in different ways:
- In principle, a primitive text editor such as Notepad under Windows or Leafpad (or even ed in the console ) under Linux is sufficient to create the input files with text and commands . The TeX run can be initiated in the console ( command line ), then you need a previewer - this spartan equipment is sufficient for working with TeX, a special TeX editor is in principle not required. Notepad and Leafpad are very easy and intuitive to use.
- Text editors that offer syntax highlighting for various programming or markup languages, including TeX, such as Vim and, under Linux, gedit (or even nano ), are an improvement in this way of working . This makes it easy to see which code passages form text to be formatted, in contrast to TeX commands, and which passages are commented out; Finally, they help you close curly braces correctly
}for macro arguments (typical source of errors with complex consequences); also for the brackets in mathematical formulas , for whose formatting TeX is particularly popular. - However, Vim only becomes useful after a long period of familiarization; it is still based on “Stone Age” notions of graphical user interfaces .
- Even without a fully developed graphical development environment, there are a number of scripts and other tools that automate the creation of documents, so that external tools such as BibTeX and MakeIndex can be "started" if necessary without the intervention of the editor and you get certain similarities with WYSIWYG programs, especially for working with the mentioned text editors gedit and Vim ; one of those tools is rubber . - However, these tools are difficult to master for beginners.
- There are special TeX modes for the universal graphic integrated development environments Emacs and Eclipse : AUCTeX for Emacs, the plug-in TeXlipse for Eclipse and LaTeX . With AUCTeX, the preview-latex package provides an automatic preview of formulas, tables and graphics (in these situations the source code is particularly confusing and the preview is valuable). - This approach is not useful for beginners.
- There are graphic integrated development environments specially programmed for the creation of TeX documents, which are known as "TeX editors". In addition to the syntax highlighting, they usually offer help on the TeX commands and buttons to “set the TeX program or other tools in motion”. Examples are Kile ( for LaTeX ) under Linux, TeXnicCenter and WinEdt under Microsoft Windows as well as the award-winning TeXShop under Mac OS X , which inspired TeXworks now available for Windows, Linux and Mac OS X. Instead of an external previewer, an integrated one is used here, and there is (as with some other modern previewers) a function called inverse search to jump from a position in the preview to the corresponding source code. - The objective of TeXshop and TeXworks is precisely to offer beginners easy access to TeX.
A more detailed description of development environments for LaTeX can be found under LaTeX # development environments .
Automatic generation of formatted documents
- Victor Eijkhout: TeX by Topic . New edition edition. Lehmanns Media, Berlin 2014, ISBN 978-3-86541-590-5 ( CTAN [accessed on August 2, 2015] Access options and further information).
- Donald E. Knuth : The TeXbook . Illustrations by Duane Bibby. Addison-Wesley, Reading, MA et al. a 1986, ISBN 0-201-13447-0 (first edition: 1984).
- Norbert Schwarz: Introduction to TeX . 2., revised. Edition. Addison-Wesley, Bonn a. a. 1988, ISBN 3-925118-97-7 ( ruhr-uni-bochum.de [PDF; 2.4 MB ; accessed on August 2, 2015] In the "unchanged" PDF version, the author gives the year of publication of the 3rd edition first as 1991, then as 1988).
- Norman Walsh: Making TeX work . O'Reilly Verlag, Sebastopol, CA 1994, ISBN 1-56592-051-1 ( full text [accessed August 2, 2015]).
- Donald E. Knuth: Digital Typography . CSLI Publications, Stanford, CA 1999, ISBN 1-57586-010-4 (collection of articles, some on TeX).
- Michael Downes: TeX and LaTeX 2e . In: Notices of the AMS . tape 49 , no. 11 , December 2002, p. 1384–1391 ( ams.org [PDF; 822 kB ; accessed on July 26, 2015] 8-page representation of TeX, LaTeX and markup types including historical representation).
- Plain TeX Quick Reference (PDF; 138 kB)
- German-speaking user association TeX e. V. (DANTE)
- TeX Users Group (English)
- CRE podcast on the topic: TeX and LaTeX
- TTH: the TEX to HTML translator , Department of Nuclear Science and Engineering, Massachusetts Institute of Technology
- How should I pronounce “TeX”? In: TeX Frequently Asked Questions on the Web version 3.22. April 27, 2011, accessed September 2, 2011 .
- The TEXbook, Donald Knuth, Addison-Wesley, 1996
- hyperref - brief description on CTAN with links to documentation and downloads.
- For example BibLaTeX with biber and xindy .
See substitution (logic) . However, the replacement does not take place at the level of the entered characters, but in relation to the formal alphabet of the tokens in the sense of Chapter 7 of The TeXbook. In the example
\leavevmode, a single token is replaced by two other tokens; the representation is misleading in an effort not to get too technical. Individual characters can also behave like macros, e.g. B. the tilde
~to create a non-breaking space , and the double quotation mark with ngerman .
- tex archive / macros
- tex archive / macros / generic
- Donald E. Knuth: An Earthshaking Announcement. (PDF; 243 kB) June 28, 2010, accessed on January 31, 2017 (English).
- Donald E. Knuth: The future of TEX and METAFONT. (PDF; 15 kB) October 3, 1990, accessed on October 7, 2012 (English).
- Donald E. Knuth: Questions and Answers II. In: TUGboat. 17, 1996, pp. 355-367. Also printed as Chapter 32 of Digital Typography. P. 620.
- Donald E. Knuth: Typesetting Concrete Mathematics. In: TUGboat. 10, 1989, pp. 31–36, 342. tug.org (PDF; 584 kB) Also printed as Chapter 18 of Digital Typography.
- Michael P. Barnett. Computer Typesetting: Experiments and Prospects. MIT Press, Cambridge, Massachusetts 1965.
- Donald E. Knuth, Michael F. Plass: Breaking Paragraphs Into Lines. In: Software - Practice and Experience. 11, 1981, pp. 1119-1184. Also printed as Chapter 3 of Digital Typography. Pp. 67-155.
- Advogato: interview of Donald Knuth. In: TUGboat. 21, 2000, pp. 103-110. tug.org (PDF or advogato.org )
- GNU Project: GNU coreutils manual, version 6.9, 4.1 fmt: Reformat paragraph text. 2006.
- Franklin Mark Liang: Word Hyphen-a-tion by Computer. Ph.D. - Thesis, Department of Computer Science, Stanford University , August 1983.
- The TeXbook . Appendix H: Hyphenation, pp. 449-455.
- The CTAN package tex-overview offers a very comprehensive overview of further developments of TeX .
- For the main statements in the following, see the TeX Live documentation, especially the section on TeX extensions .
After the command line “
latex beispiel” the console output appears
This is pdfTeX, the version number then given contains the version number of the underlying TeX version.
- etex on CTAN. The program is usually referred to as "e-TeX" on websites, but ε-TeX in its PDF documentation.
- microtype on CTAN.
- inputenc on CTAN.
- Differences between XeTeX and earlier TeX variants are described in the CTAN package XeTeXref .
\catcode65536 input characters can now be used instead of the previous 256, and macro names can e.g. B. be composed of Chinese characters, the package is similarly unisugar .
- z. B. etoolbox
- TeX Directory Structure at tug.org.
- Cf. “systems” directory of the CTAN.
- Peter Gordon: Introducing Donald Knuth and Computers & Typesetting. In: TUBboat. Volume 7, 1986, No. 2, pp. 93-95, 94.
- Frank Mittelbach u. a .: The LaTeX companion. 2nd Edition. Pearson Studium, 2005, ISBN 3-8273-7166-X , p. 1009.
- tex.nanorc at wiki.ubuntuusers.de
- Change and show PDF with TeX “immediately,” and link I / O - private
- gedit LaTeX plugin on wiki.gnome.org
- LaTeX (Vim Tips Wiki) on vim.wikia.com
- tex.vim on vim.org
- Rubber - a wrapper for LaTeX and friends on launchpad.net
- TeXlipse on sourceforge.net
- preview latex on CTAN