LaTeX documents look like they have been professionally
typeset. LaTeX will take care for fine-tuning spacings,
line and page breaks, subscript and superscript placements,
etcetera with little user input needed.
The better quality compared to MS Word and similar documents
is usually greatly enhanced because the latter documents are
normally not created that well. Most users are not
professional typesetters and know very little about it.
When a user does the formatting, he or she will often make
less than optimal choices, and be somewhat inconsistent from
one place to the other. LaTeX will ensure optimal and
consistent formatting throughout the document.
Because it is faster, easier, and more reliable for typing
When you are used to it, it is a lot easier and quicker to
type in the names of characters than to pick them from
LaTeX takes care of proper placements of subscripts and
superscripts, spacing, putting one expression above or below
Professors often create their lecture notes in LaTeX not
because it looks better, but because doing the mathematics
is so much quicker.
Because the source document that you type contains much more
In your LaTeX source, you say that some string is the title
of the document. You do not say that it is bold and
centered on the page; LaTeX takes care of those details.
Similarly you say what are the authors, the abstract,
chapter titles, section titles, quotations, citations,
This simplifies format conversions. The title may look
different in a book than in an article, but it remains the
You can search for these quantities. For example, you could
search your document collection for a paper cowritten with
Because LaTeX takes care of numbering of chapters, sections,
equations, figures, citations, etcetera.
You use mnemonics of your own choosing, not numbers.
That saves you a lot of hassle for big documents. For
example, suppose your chapter has 300 equations and you need
to insert a new equation between numbers 100 and 101. If
you used numbers, you would now need to renumber all 200
subsequent equations. And also change their numbers
wherever they are referred to in the text. And while
writing you would constantly have to search through your
document for "now, what was the number of that equation
again." As an example, have a look at my 1,600 page book
Quantum Mechanics for Engineers. Just try to imagine
that I would have to number everything in this ever changing
It is much more reliable. Are you really going to change
say 400 equation numbers without missing one and without a
single typo? Even if the numbers were correct to start out
with? If you make a typo in a LaTeX mnemonic, you get a
message that it does not exist.
Because LaTeX does so many other things for you all by itself:
Table of Contents
List of Figures
List of Tables
Index (You just place markers on the words you want to be
Sorting and formatting the References/Bibliography list.
(Too bad that you still have to provide the data on the
In the pdf document, hot links from contents items, index
entries, equation, figure, and table references, etcetera to
the thing they are pointing to.
In the web page version, even more and better links are
Because a journal may require or prefer LaTeX.
Even if they allow MS Word, they may process LaTeX
submissions more quickly. Remember that TeX is a
typesetting language already.
And if they use your document instead of typesetting
everything from scratch, the risk of additional errors
slipping in that way is eliminated. If you proofread it
very carefully before submission, you do not have to start
from scratch again.
Because it gives you programming capabilities.
You can set variables, use if statements, and do other
"programming" type operations in LaTeX.
As an example, consider my undergraduate thermodynamics
class. This is a required, out-of-interest-area, difficult
class, so cheating is common. I give everybody the same
questions, but I do give each student the questions in a
different order. Makes it harder to ask "What is
the answer to question 3?". And I print each student's name
on each of the three or four sheets of the test. I
sometimes sort the tests by name, other times by e-mail.
Better if you do not know who is going to sit next to you.
The answers are already in my source document, I just print
it out without them for the students. Then after the test
is over, I print it again with solutions for both grading
(in the random order that I gave them) and for posting. I
could not reasonably do these things without the programming
capabilities of LaTeX. Especially with 70 students in the
There is a learning curve associated with using LaTeX.
L2h tries to minimize it by providing examples. But you do
need to get used to LaTeX.
Do not try to learn LaTeX when your document is due tomorrow
You might be cowriting with someone who does not know LaTeX.
However, if you are the principal author, like for your
thesis or dissertation, someone else, like your advisor,
will probably not have much trouble making minor changes.
The LaTeX source is readable text. Just tell them to ignore
the formatting commands.
As another option you can send someone else a pdf of the
document. They can make annotations on it using the full
version of Adobe Acrobat or some free software. (On Linux
however, I know of only one: evince Document Viewer. Using
Document Viewer, open the side pane, then use "Annotations",
"Add", click on the pencil picture and then click on the
location of the annotation. Slightly resize the yellow
window to get it active.)
If someone else wants to do some real editing of the
language, (maybe because you write in a language other than
your own), l2h allows you to remove all LaTeX, leaving only
the normal text in a web page file. This can be loaded into
MS Word and edited that way. When they send it back to you,
you can restore the LaTeX.
Finally, if the other author does know LaTeX, using LaTeX is
ideal. LaTeX is normal text, compact, easy to e-mail. No
viruses can hide in it. There are no inconsistencies in
equation editors and other software to worry about. You can
easily see what changes have been made using simple text
comparison tools. You can check mnemonic labels instead of
numbers in the changed parts. Etcetera.
It is less easy to comply with nit-picking.
The fact that LaTeX does the formatting for you also means
that you have less immediate control over it. The problem
with that is that you often need to deal with nit-picking
regulations. Journals have requirements for articles.
Universities have requirements for theses. Etcetera. They
often require you to format things in some particular way
that they happen to like. Typically, you will then need to
spend some time messing around if LaTeX does things slightly
different (and probably better) by default.
The ability to do a web search has made doing those
modifications much easier than they were a couple of decades
ago. See also the links in intro_to_latex to FAQ lists,
If the journal or university provides a LaTeX "template", as
many do, there is of course no problem.
Finally, I should point out that usually I struggle at least
as much with MS Word when I need to comply with regulations.
But of course I have much less experience with MS Word. And
worse, if I have to use MS Word, they often give me a
template to use that they have set up. And since they do
not understand the hidden structure of MS Word documents,
those templates are typically completely corrupt. If they
give you a MS Word table with a row that looks
exactly like a true table row, but is not a true
table row, you struggle. Or hand in a mess.
The font or entire page can be resized at will. Yes, my
eyes are aging. The colors can be changed too.
The browser window can be resized. Studies show that people
typically find lines more than about 70 characters long hard
to read. A browser window can be resized to a width that the
reader is most comfortable with.
It is much easier to add multimedia to web pages.
My book on quantum mechanics has had animated wave packets
for many years in the web page version. (Nowadays, the pdf
has links to them.)
There is much more useful information in good web pages.
In a LaTeX document, title, authors, sectional headings,
emphasized text, etcetera is explicitly marked
as being title, authors, sectional heading, emphasized,
etcetera. This information can be transferred directly
to the web pages, and by and large, LaTeX2HTML does so.
So software does not have to guess whether, say, some bold
string is maybe a sectional heading. It is told it
is. That allows automated software to understand the
content much more reliably.
Not only does this allow web searches to be more accurate,
it also allows for increased access. For example, it allows
software to be written that can tell a blind person a very
great amount of accurate information about my book on
quantum mechanics. It can say what the title is, who the
author is, what the chapter headings are, what the sectional
headings in each chapter are, what web page is the table of
contents, which one the index, which one the bibliography,
etcetera. All this information is taken from the LaTeX
source and transferred explicitly into the web pages.
The web pages are much easier to access remotely.
To read a section in my book on quantum mechanics requires
the entire 20 MB pdf to be downloaded. The same content,
however, is divided into hundreds of web pages. Someone
interested in one particular section only needs to download
the one section they want. And the text can be downloaded
and read before all the images are there.
In addition, after you download the pdf from a google link,
you would have to search through it to find the section you
want. A google link to the relevant web page gives you
exactly what you want with no searching and no confusion.
Hot links are much more accurate on the web pages than in the pdf.
There is no particular reason that I know why the hot links
in the pdf are not that accurate. I am sure it can be
fixed. But at the time of writing the reality is that they
are often not precise. They may put you on the right page,
but not on the right spot on that page.
The images are much better at default resolution.
Like the previous item, there is no compelling reason why
this should be so. But at the time of writing, the reality
is that pdf viewers do a lousy job of anti-aliasing images
for screen resolution. They interpolate, or even just take
the nearest pixel, where they should be integrating. I have
seen images where entire lines completely disappeared.
Note however that if you magnify the images in the pdf,
they will become much more accurate than the web images.
Simple menu-based processing can handle even large documents
Works great on the 5 MB LaTeX source of my book on quantum
A quick way to learn the real LaTeX.
Example documents show how to do every part of creating a
book, a scientific article, or a thesis at selected
universities. That even includes explaining how you can
create plots with gnuplot! You do not have to piece all
these things together from separate sources.
Often, publishers also give you an example or template of a
correct LaTeX paper for their journal. If you have learned
the real LaTeX, you can deal with those.
Menu-driven processing into final form requires just a key
press. And l2h will usually tell you what key to press too!
If LaTeX errors occur, l2h will provide help on them.
Remove LaTeX from your document, leaving only the text. Run
that through your favorite spell/grammar checker. Then have
l2h put the LaTeX back in.
Make web pages in addition to pdf files of your documents.
This is based on
bug-fixed version of LaTeX2HTML. It routinely processes
my 1,600 page (in pdf form) book on quantum mechanics
without any error. That comprises hundreds of separate web
pages and close to 10,000 images, many used repeatedly.
L2h checks your LaTeX source for common LaTeX2HTML problem
areas before processing them.
A mechanism has been added to easily set web page file
names. (The major problem with the standard LaTeX2HTML
names node1.html, node2.html ... is that if you add or
remove a section, all external links to your document are
broken. They are also not easy to remember.)
A mechanism has been added to set web pages headers and
various other settings from within the LaTeX source. That
eliminates the need to edit perl files to set these.
You can now select one of several themes for your web pages.
See the examples web page for
how the provided ones look. More themes might be added later.
In addition, the l2h menu allows extensive further
customization of the theme fonts, colors, buttons, etcetera.
For images with more than 256 colors, jpg images can be
Single-language web pages can now be hyphenated in 70 or so
Web pages can now be finalized to prevent bad line breaks at
XeLaTeX is now supported, allowing support for multiple
languages, of any kind, in a single document.