Is Latex Still Worth Learning?

Bocaj asks: "I have start back to college and have to write a few technical papers. Right now it's mostly physics, but I'm a CS major and there will be many more papers to come. I've tried all of the office suites with little luck in getting them to format complex formulas correctly. I'm trying to learn Latex, but I am wondering if I should. Is Latex still the defacto standard for this kind of stuff? What about SGML or XML? What is everyone else using?"

Comments

Re:Front-ends for Latex

[Anonymous Coward]

And when you realize there's something essential that one of these can't do, you will either (a) learn LaTex or (b) Hack.

Either of which is not such a bad place to be - and I'm a satisfied TeXmacs user!

Yes and no. It depends.

[pmz]

For mathematics typesetting, I'm not sure LaTex/TeX can be beat. For large documents, it probably cannot be beat. LaTeX plus RCS/CVS and Make is truly awesome. Add gnuplot for graphs, GIMP for images, and you've got one hell of an Open Source solution for blowing away your academic buddies.

I also used to do resumes with LaTeX, which made for a distinctive look. However, once I got StarOffice, I started using it for resumes. StarOffice/OpenOffice.org does just fine, although it isn't quite possible to replicate the look of LaTeX output.

You should also ask around about Docbook, but I've never used it. Docbook, being XML-based, might be useful as a basis for web page output in addition to type-set output.

The best part about all this, is that it can all be done without Microsoft Office!!!

Re:LaTex is crufty.

[Rares Marian]

LaTeX is more readable.
LaTeX is already a layout language.
LaTeX doesn't need a backend to remind it it is a layout language.

Unless of course you'd like to create an application and hand that in instead of a report.

Learn it

[kurosawdust]

Right now it's mostly physics, but I'm a CS major...

If you're a CS major you should learn TeX regardless of whether you're going to use it in a paper or not. It's open-source and one of the few major pieces of software that is for all intents and purposes bug-free. It's part of the CS canon, and you should learn it and read the source for that reason alone.

Totally! LaTeX is still worth learning...

[RevAaron]

You should learn LaTeX. It is an awesome way to write documents. I've had a number of professors compliment me on my documents. They look really professional,

Why? My reasons:
#1) If you've done any HTML coding, or are a programmer in general, it is pretty easy to pick up the basics. You don't need to learn all that much to get the core of what you need to do- lists, bold/italic/underline, centering, paragraphs, tables, and some symbols.

#2) You can use tools like LyX to do the work for you. Even if you never learn a lick of real LaTeX code, you still end up with a beautiful document, and any of the other benefits.

#3) You can use LaTeX without having a GUI. Or a newer computer. Or a "full" word processor on a "full" OS. That is, you can write, compile and print out LaTeX docs on a DOS machine, from the console on a Unix machine, a PDA, etc.

I initially decided to learn LaTeX because there was a simple TeX compiler for the NewtonOS, my PDA platform until recently. There was also NewtonWorks- a good mobile Office suite- but there was no simple way for me to output the document and print it without docking with a Mac or Windows machine. With TeX for the Newton, on the other hand, I could export the text to any machine, compile the TeX on the machine itself or on the university mainframe, and then print.

I had to move on around a year ago from the NewtonOS, at least as my primary platform. On the Jornada 720, a Windows CE micro-laptop Handheld PC 2000 device, I started writing my papers using a real version of LaTeX- the same thing as I was using on my OS X machine. Editing the LaTeX code in emacs no less- all on a PDA! The whole cycle- editing, compiling, viewing (with WinDVI) and printing can all be done on a PDA. There are easy to install WinCE packages. I also had a PocketPC for a while, and the packages all worked very well there as well, but editing wasn't as nice as it was on the J720- it has a real keyboard. I've recently switched to the Zaurus SL-C760, and am a bit disapointed in that there aren't any easy to install ipkgs, along with a decent Qtopia LaTeX editor. Alas, I'll work on it soon enough- I'll need to be able to write up LaTeX docs and compile to PS before school starts. :)

#4) I had another reason, if I remember, I'll put it here!

#5) It's entirely free. Yeah, you could get OpenOffice. Or you could pirate/buy/get bundled MS Office. OO has generally just been a huge hassle for me; MS Office (I'm on OS X) is generally faster, more stable and less of a hassle than OpenOffice, but introduces its own set of problems.

Re:Microsoft Word can also do the stuff

[CompVisGuy]

No it can't.

If you work in the business world, and only ocassionally need to insert equations and don't care much about how they look, and you don't need to build bibliographies and use citations, then you can go ahead and use Word.

However, Word has a few serious problems that make it useless for academic scientific writing (people still use Word, but you can spot their papers a mile off as they look awful):

• Word has no built-in way to handle citations and bibliographies. You can use a footnote, but this is only good when you are not writing to a specific house style and don't want a bibliography. To do this in Word, you need to go and buy Endnote or something similar: you end up paying twice for a solution that is not as good as the free one (LaTeX).
• You edit Word documents visually. Most people use it to write short documents, and alter the formatting by highlighting text and changing its properties ("Hmm, I'll make this bold and make this italic, and oh, that's supposed to be a title, so I'll just make that 14 point..."). This is wrong -- the structure of the document should dictate the style, not the other way around. You *can* do this in Word, by setting up templates, but even these are flawed. This is essentially because when designing a GUI, you have to follow the 80-20 rule: devote 80% of the UI to make the features people use most often most easy, and devote 20% of the UI to those advanced features only a few people will use. Developer time and testing is also proportioned that way, so the advanced features are more poorly written and tested.
• Word does not allow you to easily make drastic changes uniformly to your document. Ever tried to renumber a large document's sections in Word? It is very easy to break things, and then you have to do the job *manually*! Changing the style of all text and layout is a breeze in LaTeX, but in Word seems to be virtually impossible. Most people don't know that equations can be written in different styles: changing styles in LaTeX is both possible and easy, doing so in Word is almost impossible (the equation editor is limited) and is a manual task.
• Word is a WYSIWYG editor. On the face of it, this is a good thing, but it has one major problem. The layout of the document changes as you type and as the display is updated. This has two significant problems. 1) You need to keep the whole document in memory. A friend who wrote a large Word document with many images in it found that Word would very frequently crash because of the huge memory demands being placed upon it. 2) When you open a document, its layout is recomputed and hence may change (because of a different internal state of the Word process or because of a different version of Word). So, I have had colleagues who have written a large document in Word, and then printed two copies at the same time. Because the internal state of the program changed slightly between prints, the layout was recomputed differently for the two prints and they looked different. This is not what one wants when you have to write a paper with "no more than four pages" -- it is easy to end up in a situation where on one version of Word you have a 4 page paper and on another you have a 5 page paper. In LaTeX, you edit a plain text document which is then parsed to form the final document (a PDF, say). The final document will then never change, and you can make as many printouts as you want. Because, when editing your LaTeX document, you do not need to keep all images in memory (they just sit on the HD), you can easily run many other programs without fearing that your document is going to vanish.
• Most critics of LaTeX say it is too hard to learn, that commands are harder than clicking buttons. However, to overcome all the problems of Word and use it properly, in order to create good documents, you need to know an awful lot. A friend is a Word user, and bought a reference book to allow him to produce documents properly using Word. Myself and another LaTeX user compared our LaTeX references with

Kinda like C, isn't it?

[Anonymous Brave Guy]

There are a lot of really crappy things about LaTeX, but it is definitely the standard.

Yep. I've always thought of LaTeX as being kinda C-like. Everyone knows it sucks in places, the syntax is hideous and the tricky bits require a minimum of guru status (and preferably demigod) to get right. And yet, it's awesomely powerful, it can do almost anything if you ask it nicely enough, and no-one has yet made anything with even close to the same level of power and a significantly nicer interface. For these reasons, it remains the standard for serious users.