So convinced was I that this was why scientists wrote in
LaTeX that I even had a go at writing a LaTeX paper of my
own (it never got finished; there was a lesson there).
So after successfully completing the task ZERO times, the author still felt fully qualified to dismiss the entire approach and whinge about the tooling. /facepalm
Indeed. I felt he presented a very limited argument, without much of an overreach.
He doesn't say "Never use Latex!". His opinion is that those who are picking it up under the false illusion that latex gets out of there way better than MS Word are doing themselves a disservice. Seems to me to be a fairly uncontroversial opinion. The latex document, especially one that's big, is not trivial to parse and edit. Making sure the 'flow' of the text is right is infinitely easier with WYSIWYG.
He absolutely seems to be OK with people who use latex because they care deeply about the beauty of the resulting document, and himself advocates for a post-writing conversion to latex. Markdown>Pandoc>Latex even. I feel many here are not reading the article in context.
I have an aside of my own to add to the "latex gets document design out of your way" crowd. I can bet any money that the types of people who will fiddle a lot with the typeface settings on MS Word will fiddle with themes/linespacing/monospaced font-du-jour on their text editor too. It has less to do with the writing environment and more to do with personality. I would know since I'm one of those people.
He slams Latex usage as a fetish. Then he actually has a hard time putting forth arguments against latex, ends up with very weak ones, for example he compares MacOS vanilla Emacs installation with a weird green-on-black colorscheme with MS Word as if it was the only editor you could use for Latex. In the end I feel I wasted my time reading his blog pot, because he just does not provide any substantial information.
I don't even have the impression that his proposed alternatives (conversion changes from other formats to latex) are really used by him.
Ah, I like writing documents in Docbook, so I am not even that kind of a latex enthusiast. But his criticism is kind of really off.
> Then he actually has a hard time putting forth arguments against latex
One of the main things he does point out is that the justification often used to promote LaTeX is complete bullshit. "You should use LaTeX because you spend too much time trying to worry about what size your title should be in Word." No, I don't.
He also points out that, even if you prefer writing markup yourself to WYSIWYG, LaTeX isn't exactly the simplest markup. Stuff like restructured text or markdown (hell, even HTML in many cases!) has much less intrusive formatting information than LaTeX does. LaTeX is designed for final typesetting, and it has too much overhead for the majority of writing cases where you don't care about final typesetting at all.
> He also points out that, even if you prefer writing markup yourself to WYSIWYG, LaTeX isn't exactly the simplest markup. Stuff like restructured text or markdown (hell, even HTML in many cases!) has much less intrusive formatting information than LaTeX does. LaTeX is designed for final typesetting, and it has too much overhead for the majority of writing cases where you don't care about final typesetting at all.
The larger hurdle (IMO), is that TeX can be extremely finicky even compared to HTML. To get a document that looks mostly the same as the one you got out of Word takes time and a fair bit of setup. I went through this same process with my CV after I finished my MA.
Once it's done however, updates are quick and require almost no effort.
That last point is the killer feature to me. I haven't updated my resume in years, but when I do, it is quick and I'm not worried about bitrot. The document I'm working on in wiki from last week? Odds are it is already broken. The document in Word that I shared with a friend? Almost certainly not going to be a trivial update.
This isn't to say that there are not benefits to word/wiki, mind. Just, "where everyone else is" is not the most satisfying of reasons.
Before I went the LaTeX route, I tried just doing it in HTML/CSS and although I'm well versed in both, it never really looked right.
Using the "print" @media type and converting from px to inches only gets you so far. For something that I expect to be printed at some point, TeX is a much better place to start.
Oddly, if you are willing to take complete control over the layout and not rely on any default flow of HTML, I suspect it would be easy enough.
That said, the stability of the flow of TeX and friends lets me not worry about anything having changed on me in the past few months/years. I think this has gotten better in the HTML world, but it is still a much shakier foundation.
I didn't take the criticism of editing LaTex as anything to do with color schemes, but rather that you had to mentally skip over all the formatting text. For some that will be easier than others, but it's silly to ignore that there are alternatives that don't require that skill.
It's not a very weak argument to say that learning LaTeX just so you can write your paper is a waste of time. The caveat is that, for some kinds of papers, it is less of a timewaster than other kinds. Especially if they have alot of math, or some wonky tables.
I think the blog post was okay, but could have used some editing as I ended up skimming some of the points he was trying to make but that's kind of what blogs are about.
> I feel many here are not reading the article in context.
Agreed, people are having a strong reaction to a criticism of LaTeX that the author isn't actually making.
> I can bet any money that the types of people who will fiddle a lot with the typeface settings on MS Word will fiddle with themes/linespacing/monospaced font-du-jour on their text editor too.
It pains me to agree with you, but I'll agree that at the end of the day, being concerned about typesetting is probably a bit superfluous.
I think you're missing the point though. There's a strong value to standardization in formatting when it comes to typesetting, can you imagine how annoying/distracting it would be if serious journals had minor differences in formatting? I'm not being sarcastic here, it's a necessity when it comes to any kind of repeated publication. LaTeX solves this problem better than anything else in existence, way better than the limited themes and styles available in WYSIWYG text processors.
I made a nice fancy template with tcolorbox for q/a homework problemsets and now that's a solved problem for me. It looks better than anything you could possibly make in Docs or Word, it looks exactly the same every time.
It's also scriptable. For example, I was working on an amateur radio and I needed to save off the calibration settings before I worked on it. I took a video of myself scrolling through the settings, watched it once with mpv repeatedly hitting 's' to take screenshots, and then was able to interpolate the screenshots with the actual content of the Yaesu manual, producing a beautiful manual that described the setting in question, as well as showing what it was set to before I started working. It took me like 15 minutes, and I'm glad that I did it because it saved me a lot of time later on. Try doing that in Word, Libreoffice, or Docs!
Not to bellitle your achievements (honestly!), but I do not think your last paragraph can be a commonly encountered scenario.
Either way, I'll back out a bit and ask a more elementary question. Is a workflow possible wherein one writes in anything-but-latex (Word/Markdown etc) and then uses latex to typeset? Because I (and the author) agree with you that a) standardisation is a good thing and b) latex is excellent at typesetting.
What latex is weak at is me writing my paper in it, and then trying to read back whether what I wrote flows well.
There is not, to my knowledge, a really commonly-used workflow that starts at Word and ends at typesetting. This is something Pandoc purports to do, but I would expect to lose most of your manual formatting tweaks from Word in the process.
Pandoc is a good tool for automating workflows that start with some other format and end with LaTeX.
Another good option if you like Emacs is org-mode, which has a LaTeX export, among other things.
Org-mode has a similar feel to Markdown, exports to LaTeX, and allows you to embed LaTeX snippets so that you can keep using the Org file as your source of truth, rather than eventually having to do a full switch and edit the LaTeX document full-time.
Another very neat feature is Org-Babel, which allows you to write code in (almost) any language that is then replaced with its output, which is great for example code. It can even pipe stuff through PlantUML, allowing you to embed your diagrams into the source code rather than having to manage them separately.
As an example of what it looks like at scale(-ish), I used this to write my thesis paper[1], and all the other related documentation[2].
I'm curious why latex is weak at that? Considering most authors probably just use a notepad equivalent, the idea is if you are finding yourself having to put a ton of markup into the document, you are probably relying too much on items other than the prose to make your points.
That is to say, yes, there is a workflow. It is basically to build the prose out separately from the layout. Latex is used to typeset the document you want to do. Not to author it. By the time you get to it, you probably should be at the point where the markup is what you care about.
That make sense?
Yes, there will be times where you care about how the document looks. That is likely an iterative process and many of us find value in separating the quibbles over the look from the authoring of the content. Both are valuable. Doing them both at the same time is ridiculously hard.
Latex is for typesetting. Remember that when tex was developed typesetting was typically done by someone other than the author. So rather than trying to write initially in latex, write in literally any program you want. Then, take that content and surround it with latex.
However, this won't always be the easiest method. For example, anything math-heavy will almost certainly be easier to type directly in tex markdown.
There is an unstated premise here that is very important: you care, or don't, about the final appearance of the document.
When people argue that latex gets out of the way, the road being spoken of is the one leading to a beautiful document, so the author is expected to care about the final appearance. LaTeX and Word both come with learning costs, the initial effort in to just bang out some words is clearly higher in LaTeX, but they will look good from the start.
Yes, if you don't much care about the looks of the end result, Word is more expedient, but then so is notepad.
Of course, his comparison shots w/ LaTeX and LibreOffice are misleading. While it's understandable to show what CM and the global style template look like, I don't think I've seen anyone - save for people writing package documentation - deliver using that formatting.
What sort of documents do you look at? Lots of actual reports are in the "basically default" template; I googled for "us annual car deaths", and followed the first "institutional" result found, clicking around to find an actual, recent*, report. The first I found was in "basically default" style: https://www.iihs.org/frontend/iihs/documents/masterfiledocs....
Next, I googled "russian pension reform pdf", and in the top three hits there were two reports with various levels of depressing Word defaults, and one that I'll give be benefit of the doubt and say was actually a conversion from a web page.
I'd say this is a fairly typical look for any report that isn't super official or actually a magazine trying to sell something.
Most of the time, it's CV's, Thesis papers, dissertations, that sort of thing. Really just looking for interesting ideas and packages I might want to "borrow" :). My current CV (and one other supporting document) started out as creatures of Word, so trying to keep to the format was a primary goal.
The IIHS report honestly isn't bad looking. To my eyes, it could be done in TeX or Word depending on how complicated the styles actually are.
I think it's actually not 'beautiful' but 'consistent'. LaTeX formatted documents, in my experience, have a higher mean appearance to them, but almost no variance - they are all very same-y. They're always fine, but rare is the LaTeX document I encounter that has actual artistry to it. Especially as people tend to work off templates for very understandable reasons.
Honestly, it's pretty rare to encounter any document that has actual artistry to it. The one I think of directly is a book about making books, though a few journals do good templates too.
I use LaTeX for my resume primarily and to be fair I edit the content as plain vanilla text first and only "typeset" it after that.
I view the two tasks as very different. One of the reasons I dislike Word and other Word Processors is that they crowd my writing views with formatting tools and suggestions.
A plain text editor is far more comfortable for me when I'm writing and Latex allows me to tackle the layout and typesetting work as a separate step.
I guess I don't see why that's a complaint about Word at all. When I'm writing doc, I fire up Word, set the view to Draft, set the Zoom to whatever is comfortable, and use the built-in styles for Normal, Headings, and Titles to structure my document. I have just as much screen real estate dedicated to Word as I would to VS Code. That automatically builds the navigation pane in recent editions, and since I'm creating a living document roughly 95% of the time, that's perfect. I literally care nothing at all about the exact formatting; just the document structure, content, and spelling. I'm doing about as much formatting as I would with a markdown document. If Word auto-formats something and you really don't want it to, you hit Ctrl-Z to revert it. If it keeps bothering you, you turn that auto feature off.
I get the argument that people -- that is to say, programmers -- like to write in their text editor of choice, too. However, I think that's what it is: preference for the familiar. Writing LaTeX feels like you're programming and lets you use tools you're familiar with, so it appeals to people who like to write programs. That doesn't mean the extra effort is really paying off.
I think arguments that the results from Word aren't good enough or professional enough don't really hold water, either. Take a look at any of the opinions from the Supreme Court of the United States [0]. Download any of those opinions and look at the metadata. Some of them are scanned documents, but many of them say: "Application: Acrobat PDFMaker N for Word". You'll find the same thing if you look at petitions that have been filed, and they get quite complicated [1].
Word doesn't support the editing tools that make me most productive so yes it's totally a preference for editing. But Word also by default does a lot of auto formatting and I just don't like chasing down the various toggles for it all.
It may have gotten better I haven't actually used Word in probably 10 years or more at this point. The fact is that I can edit my document in an editor I'm comfortable with and I can typeset it with a tool I'm comfortable with and the two steps are pretty distinct from each other and I like that explicit separation.
And that's fine. You should use the tools you prefer.
But it doesn't really make a LaTeX user more qualified to criticize Word any more than OP is for criticizing LaTeX, does it? Is "I just don't like chasing down the various toggles for it all" really a better argument than "I find reading all those codes aversely difficult"?
The difference with Word versus LaTeX fiddling is that I can see the markup and know exactly what I'm doing with LaTeX. Oddly, WordPerfect, back in the day, had a view that would show you the markup. It was actually quite nice. I could try to view things in a WYSIWYG view, but I could also drop into the markup, since that is ultimately capable of expressing where things are happening.
Contrast this with the current Word system. I have had a non-trivial number of times where it will start doing things and I have no idea why. Probably logical, but I can't see the inputs to the logic. Only the output, with some idea that I should be able to drag things. Doesn't always work, and sometimes "undoing" the drag of an items is a non-trivial task.
I don't think he did a good enough job justifying just how annoying LaTeX actually can be.
To write a conference paper for submission, the steps to start it out are generally:
a) Copy the journal's LaTeX shell file.
b) Copy some other Makefile to actually convert TeX files into PDFs. (Because it turns out that this is simply too complicated to actually make a single command, especially when you have a bibliography and several images to import).
c) Okay, now we can go through the process of actually setting up the master TeX file, separate TeX files for each chapter and start typing text.
d) Include an image. Weep at how annoying this is, and how inconsistent documentation is.
e) Time to tweak the article to get it to fit the page count. Don't you love all of those little cycles where you have to wait for a half-dozen commands to run to figure out if you've shortened a single paragraph enough to pull the section header up to the previous page? And this is assuming you have a PDF viewer that will automatically refresh the display for you if it detects the PDF file changed.
In my opinion, the two great things about LaTeX are its syntax for math and BibTeX. But I wouldn't contend that these are unimprovable (particularly BibTeX). But it spends too much time making you concerned with publishing concerns that trying to just jot down notes or other early document planning is surprisingly difficult.
Is he? For me, markup of plain text is easier because it fits with the other tools I use. Thus I get all the revision control, more powerful diffing capabilities, proper macrology (templates and styles).
Whereas I (not claiming a universal) can only use Word for small documents; when they get big, with revision tracking, chapters, etc etc I've found it becomes fragile. We had a 1600 page FDA submission garble itself shortly before we had to submit and had to go find older versions from backup tapes (!), re type stuff from old printed out copies, and check every number in every inline data submission which caused major panic and stress. Everybody we asked for help said, "don't use word for a large document"
Try to do a thesis in LaTeX + Sweave then, as I did. Reading books, browsing internet and building the thing from zero like a ram, just because I can (lol). In Emacs, of course (Learning to swim directly in the white waters, why not?) Fun times.
> In Emacs, of course (Learning to swim directly in the white waters, why not?)
Does Emacs have a reputation of being difficult to use? I remember that in the late 70s and early 80s MIT's CS and AI labs were proud that the (not specifically trained) secretaries used Emacs and wrote macros for it.
It's been decades since I was a novice user so I don't remember the experience, but was the first editor I'd ever used, and, years later, the first editor my (arts major) wife ever used, and I don't remember any particular anxieties or consternation from either of us.
Newer generation. Everybody used word. Couldn't be more disconnected from MIT. No unix gurus teaching fancy jedi tricks at sight. I was the anomaly sprouting in the middle of nowhere for no reason and doing all the walk backwards like a flapping salmon
Emacs was just more dificult than the other programs that I tried, and I tried a lot: word, antiword, abiword, starwriter, openoffice, ed, joe, elvis, nano, vim, xemacs... okay I will bang my fingers and head against emacs and see what happens
I realized later that I was doing it all wrong when I started to slowly unfurl the capabilities of the program. Couldn't be great if...? and, suddenly a wild major-mode appears. Not more viruses hiding around, not more big files that decide to crash or lose their format five hours before a presentation. Not more I updated the program and now I can't open my old files anymore.
Very well time spent in retrospective. I use it every day. Same for LaTeX that was much more dificult to grasp but even more rewarding in a lot of unexpected ways.
Emacs has always had a reputation of being a text editor for wizards. And it's even worse as modern computer editors have gone in a completely different direction. IMO, nano is probably the easiest of the standard editors on command line, and vim is somewhere in between.
> after successfully completing the task ZERO times
There's a Knuth quote about how the aim of TeX was to allow anyone to create properly typeset documents. I can't find the exact source, but https://www.tug.org/whatis.html should give you some idea.
This individual wasn't able to. Last time I looked at TeX it (or metafont specifically) often seemed to default to bitmap fonts. That's a poor default - a bug by today's standards - but Knuth has a different definition of 'bug' than you and I do. Maybe you should spend less time judging users and more time working on improving the software you're advocating for?
Metafont does produce bitmap fonts yes. However, it's quite good at generating these for arbitrary resolutions, so I don't think this is actually a problem. Though actually using a non-bitmap solution might work better nowadays since the computing power is available now, it's probably good enough as it is.
I think you should spend more time updating your knowledge of TeX :) All jabs aside metafont is separate from TeX in most regards, other than that Computer Modern was made using metafont (but is now widely distributed in a vector format).
You're missing the point. That metafont is seperate from TeX is irrelevant: the only time someone has to use metafont is with TeX, so metafont is considered part of Tex.
PostScript is its own can of worms. People who need typeset documents shouldn't have to concern themselves with it.
It's not 1994 anymore, alt.sysadmin.recovery is dead, nobody knows what a 'cluestick' is or what code ID-10-T is, and people generally have stopped blaming users for poor software - which TeX, by Knuth's own standards is.
This complaint is basically the reason XeTeX was written, which accesses your standard system fonts instead. LuaTeX also handles TTF and OpenFont without trouble. These both produce PDFs instead of DVI or Postscript output (per your comment below).
I'm not sure why you're not replying to that comment, but yes, there's a bunch of software created to have better defaults than vanilla TeX. This reinforces the point.
Not really, because when people say "use LaTeX", they don't really mean you should use only the original version of the software, they're talking about the overall suite of TeX derived programs.
> There's a Knuth quote about how the aim of TeX was to allow anyone to create properly typeset documents.
(Emphasis added by me)
I have probably read every single word Knuth published on the topic of TeX, and what you said seems very different from what I've gathered of Knuth's philosophy and style. Knuth intended TeX to be used by someone who cares about typesetting, and the program comes with a nearly 500-page manual (The TeXbook) that was very much intended to be read. In fact, he expected anyone who was preparing any special sort of book (a multilingual bible, or dictionary, or whatever) to change the program itself, and has expressed surprise at people using elaborate macro packages like LaTeX or doing everything in TeX macros.
(BTW the page you linked at https://www.tug.org/whatis.html is also pretty good and note that it says a lot about typography and the finer details of typesetting, and less about ease-of-use: search the page for "support" and "TeX is not the tool for you".)
> often seemed to default to bitmap fonts
Actually, TeX uses only font metrics, and the output of TeX is merely a set of instructions about where to place what characters from what font. (A compact binary encoding thereof, in the form of a DVI file.) A separate program is needed for actually using the font and rendering the output. (These days everyone uses pdfTeX or some other program that combines the two and uses vector fonts, but that's the original design.)
In any case, for the original task that TeX was designed for (completely specifying the appearance of a book, in a future-proof way however technology may change), it is absolutely crucial to combine it with a program for bitmap fonts (and that's Metafont -- it produces only bitmap fonts). Think about it: the rasterization (converting to bitmaps) has to happen somewhere: if you leave it up to some other technology (the one used by the printer or whatever), then you're back to the same problem that led to the creation of TeX, where in a few decades fonts may be produced by some completely different method and you may no longer be able to produce a book whose appearance is identical (except for the typos you fixed or minor rewrites or whatever) to the earlier one.
I'm aware rasterization has to happen at some point - the issue with metafont is the input, rather than the output.
Requiring someone to read a 500 page document to produce a typeset work does indeed indicate a conflict - I would spend some time finding a URL for the Knuth quote but HN really doesn't like people discussing TeX or Knuth in a negative way, and this hasn't been a very productive conversation so far.
I did read that piece from start to end and I always thought "the arguments against Latex must come soon", all that I read where kind of repetitions of their title and glorifications of MS Word et. al.
Linkbait trolling tier for sure. Slams LaTeX for being a markup language, then suggests we use org mode instead (org mode being a markup language which only produces interesting LaTeX output when you embed LaTeX in it). FWIIW I do use org mode as a sort of higher level LaTeX for note taking, mostly because the latex bits get compiled as I type, but when it comes time to turn my thoughts into something more organized, it gets turned into pure LaTeX.
There are a real reasons academics and quasi-academics (aka me) use LaTeX instead of something like libreoffice. Equations are basically impossible without it. I can't write down a complicated relation in LaTeX as fast as I can think of it, but it's a lot closer to that speed than laboriously constructing a relation in the equation editor of any word processor. If you do something like physics, machine learning or topology, trying to write a paper in libreoffice or whatever they call Word now is sort of like trying to write a novel on an iphone. The author never mentions equations. Yeah, it's more difficult for a n00b to construct, say, a letter to the editor in LaTeX than it is in Libreoffice -that's not what it's for! Though of course once you have paid the toll of learning the thing, it's pretty quick and in fact lives up to its promise of not getting in the way of writing the content.
Beyond that, my equations will be accessible for decades and will be completely platform independent. If you have an academic career, this is probably reasonably important.
There's other reasons besides "equations" and "long term stability" (BibTex is probably top third reason) -but those are the big ones.
Edit add: beamer mode is also a good reason to use it once you know it. Take your paper; turn it into a presentation -never leave the same application.
Microsoft Word's equation editor has improved a lot over the last years. I'm not saying it meets LaTeX at its strenghts but it is very unfair to call it "hell".
Also, I haven't used it but newer versions recognize handwritten equations and formulae and automatically transcribe them to the equation format.
My experience with the Word equation editor is that it causes the entire program to stutter, hang, and lag horribly if you have non-trivial equations. No idea why.
On the flip side, LaTeX definitely takes longer to write, but it's less frustrating than having to deal with Word's (well intentioned) autoformatting.
That is because ANYTHING on word beyond typing, italics, underline, and bold causes that. Try importing a citation from something like endnote or Zotero, I have time to take my hands off the keyboard and crack my knuckles before I can get back to typing.
Depending on proprietary stuff which changes all the time, or even complex FOSS software with single point of failure developers (I dunno, TeXmacs or something) is a huge risk which went away when I bought a cheap Leslie Lamport book on LaTeX. Doesn't really matter if a new feature in proprietary application-X allows me to also use it with LaTeX. I can just use LaTeX!
I wrote my dissertation in 2004, can still trivially use the LaTeX in it, and I sometimes do. Libreoffice, Word ... pretty sure I'd have a harder time of it. Heck Libreoffice was probably called StarOffice back then!
I was forced to submit a lab report in Word. I'd tried to use the LaTeX feature, and I couldn't even get it to properly understand limits of integration. After ~5 minutes, I just gave up and used the annoying Gui interface, since at least it worked. This was last year, and I haven't tried again (since if I need to typeset math, I'll just use LaTeX). Word has its advantages, mostly ease of use, but it just can't handle math and bibliographies as easily as LaTeX.
I haven’t actually used the LaTeX interface in Word, though I had used it in Pages on several occaisons. I was impressed and unreasonably wished that Pages was actually a full-featured, live updating LaTeX GUI editor.
The equations in the current equation editor look very ugly. The old Equation Editor produced nice-looking equations but it was a pain to use. In LaTeX neither of these is a problem.
> There are a real reasons academics and quasi-academics (aka me) use LaTeX instead of something like libreoffice.
I can attest to that from first-hand experience. Having done a thesis in LibreOffice once, I wouldn't want to repeat the experience.
That's not an absolute negative against LibreOffice, though. I still use it often and like doing so. But a large document with many illustrations, many text-internal references, code and a non-trivial bibliography? Never again.
Counterpoint: I'm an academic, and almost no one I work with uses LaTeX, and I've found that the pain points between the two are different, but not particularly better.
And if I never saw another beamer presentation again I'd be a happy man.
I'd venture to say that None of those are beamer/LaTex problems. They are all exclusively at the hand of the presenter. Someone who doesn't care enough to learn how to properly give an engaging presentation, or was never given the opportunity to do so will make those kinds of presentations.
In my undergrad Honors Physics program, I was required to attend two different workshops on giving presentations (on top of the university requirement of Speech & Comms class) and a technical writing class. Suffice to say, my beamer presentations are not identical, nor so they include "extra" slides. It doesn't take much to drop slides, so it should be easy to comment out a number of slides for a tailored presentation. That's just a lazy presenter.
> I'd venture to say that None of those are beamer/LaTex problems. They are all exclusively at the hand of the presenter.
My personal experience with LaTeX is that it is actually rather annoying to do anything with images, so you seek to explain stuff in text (or equations) instead of trying to develop good diagrams. For presentations, this is exactly the wrong mindset to be in. A good presentation software should nudge you, if not outright force you, to making diagrams and pictoral representation over bulleted text slides. So I suspect beamer really is contributing to the problem of bad slides by making it even more time-consuming to develop good presentations.
When I was an undergrad at UT Austin I remember hearing someone who worked with Steven Weinberg mention that Weinberg told people that he “thinks in TeX”.
One of the things I love about TeX is I can write it in my editor _without_ WYSIWYG. The tooling in Emacs to write TeX is amazing. I can never go back to a graphical writing tool.
In fact, the one place you should never try to do any more substantial writing than designing the poster for your niece's 4th birthday part is a word processor.
Word processors are one of the worst environments I can imagine for actually writing. Lots of people surely write in LaTeX in part because they can use a reasonable environment for writing (i.e. a decent text editor).
Also, if you're in a decent text editor, writing LaTeX can be pretty easy (though for lots of things Org mode or markdown etc. may be more appropriate), and you can set up macros to auto-fill in things like the preamble etc.
I essentially write for a living, and I would not have survived had I not adopted LaTeX.
I wrote my dissertation in Word. 200+ pages, dozens of captioned and numbered figures and tables, over 100 equations all numbered and referenced, table of contents, cross references, works cited. I found it quite pleasant.
I wrote a book in Word. It was a nightmare, but a useful one: a version conflict across users or something in the copyediting phase glitch-deleted all the math from a chapter ... but in laboriously reconstructing them I found and fixed an error in the original. So thanks Microsoft?
I'm glad it worked out for you. I would have gone mad if I'd tried to write my dissertation in Word. (As it is, I'm, nearly a decade later, still a bit touched after trying to do much minor work involving lambda calculus in Word.)
On my old 486, I wrote about 80 page document about signal processing in Word, full of painstakingly created equations, until one day, all equations basically disappeared or became uneditable. I learned my lesson there about wysiwyg.
They did completely overhaul the equation system in Office 2007 so that you edit them with unicode text instead of the proprietary OLE editor from previous versions. The document [1] describing how they use unicode characters to structure mathematical expressions is pretty interesting, even if you never touch MS Word again.
Word is fine if you really, really, know what you're doing with it. By that I mean understanding how to use styles for formatting, how to properly insert footnotes/figures/tables etc, and what the implications are if you don't.
When my wife wrote her doctorate she did none of those things, and I ended up spending the best part of two days laboriously going through the entire document to make it conform to the university's style guide for thesis being submitted. Thankfully she uses Scrivener for writing now which as near as possible enforces doing things the right way the first time.
You mean that it will be perceived as so difficult to use that people will ignore it in favor of anything else? Calling that an advantage is a rather unique perspective. General document creation is not exactly something only a few specialists need to do.
I was a little tongue in cheek with that comment, but there could be a filter effect where people who aren't prepared to do it properly run away screaming from TeX and its many shells to the loving embrace of Microsoft Word, where they then do a terrible job and give Word a bad reputation.
I've done plenty of work where the customer insists that the final deliverable be in Word format, because that's what they're comfortable with. You can do proper formatting in Word once you learn the tools, and it's not even all that hard, although it can be somewhat unforgiving of screwups. The one thing I wouldn't attempt is to convert a TeX document to Word. In theory that shouldn't be entirely possible, but in practice there are about a million ways it could blow up in your face.
There was also a time when we paid people to operate an elevator. Then we improved the technology enough to make it simple enough for everyone to do it themselves. It has many benefits, and have the user more control over what they got and when they could do things.
I don't know about you, but I really appreciate being able to instantly get the results of a calculation rather than having to submit it to be worked out by a human calculator.
Agreed. I am a power user of Word. All of my formatting is painstakingly performed by applying styles rather than manually formatting things. Headings, subheadings, captions, etc. are all tagged as such. Everything appears in the table of contents, table of figures, etc as it should, and styles can all be updated in one easy sweep.
But if you have even one collaborator who goes in and hammers something by applying manual formatting, the whole thing can be ruined. Actual images can end up showing up along with its caption in the Table of Figures. Entries can go missing from tables. Lone paragraphs can randomly end up with a different style.
Word can successfully manage large, complex documents, but it requires discipline. (as does LaTeX)
> Word is fine if you really, really, know what you're doing with it. By that I mean understanding how to use styles for formatting, how to properly insert footnotes/figures/tables etc, and what the implications are if you don't.
And I've yet to find a guide which instructs on how to set it up properly, how to share styles across documents etc. Do you know of one?
20 years of writing HTML and CSS has spoiled me; many times I'm using Word and thinking "I could do this with one rule in CSS" (often some crazy positioning nonsense with images/tables). If only!
Personally I've never found a single guide that you could hand to someone, but if you're coming from a background of HTML and CSS then you should pick up formatting rules pretty quickly - they're essentially the same, but implemented via a GUI rather than explicitly specified selectors and tags. The key for me is to just pretend that none of the explicit styling options exist - so no use of Bold/Italic/Centered/Whatever buttons, and instead turn on the styles pane and apply all formatting from there.
I wrote one thesis in LaTeX and one in Word. There were some challenges each way, but all in all, doing it in Word was easier. WYSIWYG helped a lot, Word had good documentation, and LaTeX was hard to customize to meet the university's formatting rules.
I will admit, however, that LaTeX run on default settings looks good. Whoever designed that bit had taste.
I first became aware of TeX generally when towards the end of writing my comp exam responses, I came across @kjhealy's templates. While this looked interesting, I didn't have the time to dive in (I was on deadline for revisions, and borrowing a family member's MacBook as the 8600M in my 15" died).
With that said, I wonder if the people who actually did write a thesis/dissertation in LaTeX had to do all that customization themselves.
For example, I'm in NJ and although I didn't go there, I could imagine Rutgers OIT telling people, 'If you want to use TeX, just download our package, it will install everything you need and will output the correct format for submission'
I wrote mine before word was really a thing. I used LaTex, and had to create my own style setup. It wasn't too difficult. I had programs generate EPS documents for inclusion of figures, resize jpg/png images, run the multiple pass latex compilation.
I am pleased to report that the makefile for my thesis still works, and generates close to the original doc, 20 something years later. I can barely open a 10+ year old office doc thanks to incompatibilities within the product.
Back in those days, word processors would not do complicated multi-line equations very well; type setting for print, display, etc. This was a problem for me.
To each their own though. I wound up writing some user manuals for products I developed in the late 90's in LaTeX as well. I've not used it much since.
E. Allen Emerson once came knocking at the unix cave door, complaining that his older papers, written in LaTeX, were formatting differently.
Upon examination, the only difference, literally, was that we had replaced the ancient 300dpi printers with less ancient 600dpi printers. And yes, CMR is a spindley little font.
People who need to produce Word documents and are writing things like books or scripts may find Scrivener very useful. It's a lot more stable on very large documents.
I. Love. Scrivener. My workflow is to sketch out ideas in MindNode, drag the outlines into Scrivener, and work in their to flesh them out. I find this infinitely easier and more flexible than outlining and editing in a word processor.
I've heard good things about Ulysses, too, but haven't tried it. I think that whole class of application is inherently better for editing large documents.
Ulysses was fantastic when I last used it to do a multi-chapter document. I separated the chapters into their own files, it remembered each chapter's scroll position so it was easy to look things up in each, and the default templates looked superb.
I stopped using it when it went to a subscription, though. I can't accuse a company that lets you write in Markdown, basically plain text, as rent-seeking, but I dislike my workflow being held hostage like that. Haven't really found a suitable, multi-document replacement yet.
There are actually a number of options - given tools like pandoc and org-mode conversion inside of Emacs, there are lots of choices of sane working environments even if you ultimately have to have a Word doc at the end.
Scrivener's great right up until the moment when you need to collaborate with someone else, at which point it falls over completely. The usual practice in these scenarios is to write in Scrivener and then when finished export the document to Word format to send to these other parties, but then if they come back to you with edits/feedback/etc. you're stuck with the problem of getting those changes merged back into your original Scrivener document, which is no fun -- especially if you've got multiple rounds of revisions among multiple collaborators to deal with.
Don't get me wrong, I like Scrivener a lot, but in lots of scenarios it just doesn't work as a drop-in replacement for a more general word processor.
In my experience the worst part about Word is collaboration. Have you ever seen the track changes interface? It's a sea of incomprehensible, badly segmented, changes where every switch from bold to italics is counted as a separate change mixed in with actual substantive changes and has some little red line among thousands of indistinguishable little red lines pointing to some unclear place in the document. It's a totally unusable interface, and yet some highly collaborative industries (ahem, law) are stuck with it. It's horrifying.
I have seriously considered offering to pay people with whom I'm forced to exchange documents (especially academic law review articles which tend to have thousands of edits) to learn markdown and git just so I don't have to eyestrain my way through that track changes screen
How well does git perform for usual text (not code)? I only use git for code and as far as I know it compare only whole lines, which is not very useful in a text where a paragraph has no line brakes and is about half a page long.
I write in Word as well and merging edits from collaborators with my in-the-meantime-updated document is really anoying and error-prone. I found https://www.simuldocs.com/, but haven't tried it yet.
There are a few tweaks you can make to gits diff so it shows changes in a line better, but it's not great.
You can also write each sentence on its own line - output will still put them in the same paragraph. But it's not as nice a reading experience of the markdown.
markdown and git can work together. I've set that each sentence gets its own line, and configured git to use wdiff instead, highlighting the changed words instead of lines.
> How well does git perform for usual text (not code)? I only use git for code and as far as I know it compare only whole lines, which is not very useful in a text where a paragraph has no line brakes and is about half a page long.
As long as you limit your lines' lengths git is just as good for text as it is for code. Of course you're correct that using it for paragraphs without line breaks wouldn't be that great, but my question would be why would you ever do that?
I'm going to make a contrarian argument, so I should preface it by noting that I write technical documentation in Markdown for a living, write an awful lot of fiction in Markdown as well, and write the rest of my fiction in Scrivener, which only superficially resembles a word processor. Furthermore, I've typeset several books with LaTeX.
Having said that, an awful lot of tech nerds (arguably like myself) seem to seriously underestimate what word processors can do.
Back when I did write fiction in a word processor, and even now when I have to send fiction to editors who expect things in Word format because that is the lingua franca of the non-technical publishing world, I don't have to spend an awful lot of time formatting manuscripts. Word processors have templates, and hierarchical styles. I wouldn't need a macro to auto-fill in the preamble because I'd have a template that included that.
But if I did want a macro, I could have it! A lot of word processors do have macro languages! Many of them also have search-and-replace capability which is basically regular expressions with an easier syntax. For all the crap Word gets, much of it well-deserved, it's actually pretty good at editing prose. It's almost like they've been developing it for over thirty years or something.
If you're in a position -- which I am -- to use a plain text editor for all your prose writing, and you prefer it, that's terrific. And you could totally make the argument that if your final copy is going to be LaTeX, you should start in plain text, whether or not you're using LaTeX markup from the beginning. But the notion that Word is some kind of underpowered little toy no real author should ever possibly use is just, well, wrong.
> But the notion that Word is some kind of underpowered little toy no real author should ever possibly use is just, well, wrong.
I wouldn't characterise it as underpowered so much as wrongly-powered. Just the ability to use macros etc. is not the same as being a real text editor.
I actually started out using word processors, and really wish I had discovered proper text editors sooner. For me, word processors are just miserable working environments.
This isn't me trying to convince you of anything, just adding some anecdata.
I find Word a miserable working environment. On the other hand, I find Pages a delight to work in. It's one of those "it just works" programs for me, and I find documenting editing so easy and intuitive.
It reminds me of those who swear by WordPerfect and wouldn't dream of using Word unless forced to; it's not the class of application they care about it; it's the specific set of tools, their presentation, and how the app matches their work habits.
(Same reason that I just can't use Google Docs, no matter how many workplaces make its use a soft requirement)
The simplest way to format a Word document is to select the text and format it. Unfortunately the simplest way is also the worst. If you constrain yourself to using Styles instead you get all the benefits of templates and consistent styling.
There should be a mode in word processors that disallows on-the-fly formatting and forces consistent use of styles.
I wrote a 200 page technical book in Word. After a couple days of customizing Word (few macros, hotkeys, customizing the styles to my liking), everything went well and I was able to do everything I wanted w/o trouble.
Word does have a learning curve (e.g. put all pictures inside a text box, it makes layout 10x easier), but once learned it is quite nice for a lot of purposes.
> Word does have a learning curve (e.g. put all pictures inside a text box, it makes layout 10x easier)
Woah, good tip! As I write more specifications and reports for management I'm using Word more, and there's a dearth of good advice online for getting decent results (it's drowned in Google by the "how to type a document" kind of guides).
Many times I've wished to write HTML and have it formatted like Word, as my styles go haywire.
Talking of styles, I'm surprised there's not a Syle interchange format for Style Sets - Microsoft's online resources don't have any to download. I was looking for a Latex-equivalent default template.
This article is embarrassingly bad. It attempts to do a literary critique of written statements of the benefits of latex, and analyzes zero actual use cases or implementations.
Tex is such a powerful tool. It makes me sad to see someone dismiss it likely because of unpleasant first learning curve, which is also why a lot of people dismiss vim or emacs.
When I was a freshman in college, an upper classmen friend advised me to learn TeX, and helped me through writing several documents. That advice & help was absolutely priceless.
Today I work in a typical tech company environment, and my team writes most of our technical documentation, sales engineering presentations, in-house tech presentations, and even recently some supporting documents for patent applications, all in LaTeX, all stored in version control in one of our repositories.
We can do real code review on technical documentation changes, have the power of many latex packages, beamer, etc., and even build versioned PDFs via our continuous integration tooling.
In some cases we also combine latex source with Mako templates and have actual software that can “push button” update all sorts of charts for new data, client-specific requests, etc.
I'm maintaining some documents in VCS+CI. But it feels like writing code in a really bad programming language: can never get the tables, titles, whitespace and page breaks right. There is always some package that does the thing I want while breaking other thing I want.
I don't think Markdown alone is powerful enough for most LaTeX use cases. What's the markdown for a numbered formula? How do you refer to it? How do you do proper footnotes and endnotes in markdown and keep the references up to date?
Thus you need to use some flavor of Markdown that covers what you need, which may not also cover another use you have. It still doesn't address output, so you need to eventually process the Markdown into something readable or typesettable, like LaTeX, which you'll need to understand anyway if you want to get the output you desire or fit into a style set given to you by a journal.
Since LaTeX is the lingua franca of math and CS, you're better off just writing in LaTeX and then your advisor/professor/reviewer/editor can see or edit your intent as well as your output.
Unfortunately, if you're in Biology, they're all over the place with their styles, and Word is the default. Perhaps then you will benefit from doing things in your own Markdown flavor and translating it to Word in the end using a little script you've written.
I would say your comment is more embarrassing than the author. Your comment smacks of "I know latex, so I'm chomping at the bit to drown out any voices against it!" It seems to me that you didn't fully read what the author wrote. Your entire comment is about your workflow, without much in the way of engaging with what the author wrote.
He is not arguing that latex is useless. Not even that isn't supremely powerful. His argument, in essence, is that a word processor is superior for writing and editing what you wrote than latex is, because there is nothing but text on the screen to get in your way. So, editing is easier.
He directs this argument in direct opposition to those who tout that the _main_ benefit of latex is that it lets you forget about formatting. The author is NOT contesting the typesetting benefits of latex. Nor is he saying "Only use a word processor, version control be damned". He is perfectly happy with markdown, for instance.
He is merely saying don't write in latex just because you think it's better at getting out of your way.
> “He is not arguing that latex is useless. Not even that isn't supremely powerful.“
I did not presume the author made either of these points. However, the scope of latex’s capabilities does undermine the author’s point about editing in a traditional word processor — that is an argument someone unfamiliar with latex’s capabilities might make, and so referencing the fact that latex is a more powerful tool than this is a valuable perspective for refuting the author’s hasty dismissal.
> “His argument, in essence, is that a word processor is superior for writing and editing what you wrote than latex is, because there is nothing but text on the screen to get in your way.”
I actually dispute what you claim here, because the author spends a lot of time talking about the claimed benefit of TeX that it decouples presentation from content, analyzing it linguistically from written statements of that benefit rather than pragmatically (for example, like separating written text out entirely and using \include or other features to inject text into formatting code).
I think you misunderstood the article because one of the weakest points about it is that it starts by discussing this idea that TeX has a benefit from decoupling the visual presentation, but then the article conflates that topic with the topic of what is on the screen while you’re editing the source. Those two things are not actually very related, and you can edit content in TeX with as much or as little interlaced formatting or style commands as you please (in fact, that’s the whole point).
Not to mention that it facilitates actual code review as a means of collaborative editing of either content or presentation logic or both. Something quite hard to achieve with wysiwyg editors.
Your comment reads like a classic knee-jerk HN contrarian response to me. You’re in a rush to criticize.
For example, you say,
> “Your entire comment is about your workflow, without much in the way of engaging with what the author wrote.”
as if this invalidates any points or makes the original post less embarrassing. But that was my whole intention. I was trying to explain how TeX is useful, because readers of the post might mistakenly think that editing in a traditional wysiwyg editor will be easier based on this post’s mistaken complaints.
A useful reply to a post like that is to talk about how the tool can be used productively, so that people may feel it’s worth trying to edit latex source directly (since that works really well in practical cases, like my workflow which I wrote about).
You seem to have some strange internal standard that the only type of valid critical response has to break down the OP directly and isolate responses to specific points.
Not so. And in fact, this post isn’t good enough to warrant that. It’s a lot more efficient to just say, “nope, you’re wrong, here are real world workflows where, had you believed the advice of the post, you’d be in a much worse situation.”
Further, while even after self-reflecting on the tone of my post (i.e. the choice of “embarrassing” to hopefully connote how short-sighted the OP’s perspective really is), I feel confident that my post was fair, not overly harsh, and useful. Meanwhile, the tone of your post feels like you’ve just got an ax to grind for some reason and want to be negative.
You’re certainly entitled to that, I don’t mind. But your post definitely entrenches me further into the position that my post was pretty fair and that it’s reasonable to characterize the OP that way.
The author's found arguments of separating content from presentation is only part of the reason I use LaTeX.
I use LaTeX and other markup languages (mainly AsciiDoc) when I can, as opposed to using a word processor, because I have found that Word manages to find hidden ways to mangle any document that does more than include text in paragraphs. What I like about using LaTeX is if the document is broken, it's usually my fault and I can see the underlying code that is breaking it. If you've ever had to unzip a Word document to fix the underlying XML (I have), you'll understand how frustrating it is working with hidden markup.
This, this, and this a thousand times. As a PhD student about to write up his thesis I'm on the cusp of deciding whether to go down the Word or LaTeX route. I cannot describe how unbelievably angry I have gotten at Word screwing up document after document.
My last effort was a journal paper submission, and the journal in question supplied the (80MB!!!) template which was so crammed full of macros, custom styles and miscellaneous instructions that my AV program quarantined it. When I eventually managed to use it, I literally transferred the content from Notepad++ (my go-to for the act of actually writing - take note, OP) and images etc. from file and spent DAYS trying to get the format right. A hateful experience.
Word is great for casual use. For academic or professional use beyond the barest of basic typesetting, absolutely not.
Not sure if you saw my above comment, but if you did go down the LaTeX path, would you have to set up the environment yourself?
I realize people may want to customize things later, but making the initial install as painless as possible seems highly desirable - esp. for students.
For students, especially undergrad, overleaf is an excellent way to dip your toes into latex. I wouldn't want to use it for anything large or serious, but it will handle a typical undergrad length document and doesn't even require installation.
I used Overleaf for my PhD thesis without any issues, including a lot of figures, etc. In fact, I almost only use Overleaf when it comes to LaTeX - it takes care of most of the pain points and is a great service that I pay for.
I tried Overleaf briefly when first converting my CV a couple years ago. Having a live preview helped so I didn't have to wait while recompiling.
Although things like [Skim][1] exist, Basictex does most of what I need, so I've never really tried wiring it into my existing install.
Pipe dream - If the people who work on macOS's Preview application were smart, they would steal it (or [Marked's][2]) ability to have a custom processor for displaying output. In that case, I could point it at my TeX install and just run my Makefile from there.
Very cool tools/perspective, but I think we have pretty different workflows/setups. I love not struggling to install TeX or its packages, and for me Overleaf has taken care of so many things that to reproduce it on my own seems much more effort than I'm willing to do.
To add to this, I find immense value in dealing with text files. In other words, treating them like I treat my code. It plays well with my choice of text editor, version control system, and distribution. Word just doesn't fit all these criteria. Being non-free and proprietary makes it a pretty awful choice for collaboration and distribution.
For the same reasons, I even switched to making all my presentations in latex even though I believe Google slides and Powerpoint are feature-wise unarguably superior.
> LaTeX is a typesetting system and a markup language. Typesetting systems are not customarily used for writing in, and while markup languages such as XML and HTML often are, this is generally recognised as a bad idea.
This is dumb. One can argue that LaTeX is not a convenient markup language enough to write in. But stating that a markup language in itself is a bad thing for writing, while Markdown was invented in 2004 and wildly popular at the time this article was written (2016), means that the author himself is not completely rational when defending his preference, using XML and HTML as strawmen.
He _does_ mention Markdown, and specifically how it was designed to be human readable. But he lumps Latex with XML and HTML in that they are rubbish to write in. He mentions also how he hates himself for writing in even his simplified version of HTML when writing long posts.
His argument is restricted to those looking to use Latex for questionable benefits that they do not fully understand.
I read this in 2017, and..early in 2018 spent months getting good at LaTeX/TikZ (using TexShop), and getting some idea about TeX. It's been amazing, well worth the time invested already. For the first time, I have organised notes on everything, including my programming. And it looks lovely, it's a pleasure to look over, makes me want to do more.
as a physics student I learned latex and became pretty proficient in it. I do write in latex often and the author misses a few major points. The main point it misses is the issue of how to write a paper fluently if i have a lot of equations or symbols. Here is the case when just write in word is not sufficient.
How so? The equation editor in word is great and even understands LaTeX formatting. I too write scientific papers with lots of equations and find Word more than sufficient.
The built-in Equation Editor in Word used to be awful until about Word 2007-2010 when it started interpreting LaTeX-like keywords (it doesn't really support full LaTeX, only a pseudo-LaTeX-like subset) and fixed some aesthetic issues. I would say the latest Equation Editor is pretty decent and suffices for most scientific papers.
Even so, it does stumble on more complicated equations (the kind of equations that pure mathematicians write -- the symbol set in LaTeX is just much bigger, and being a typesetting language, it gives the user much more control over element positioning). Equation Editor also doesn't support more esoteric math objects like say Feynman diagrams, but in LaTeX there's usually a package for it. Also, aesthetically, Equation Editor output, though functional, looks a bit off compared to true typeset math. LaTeX-heads are sensitive to aesthetics that way.
In my time, those who outgrew Equation Editor moved to third-party software like MathType, which is much closer to LaTeX in functionality. I'm not sure if people still do that.
I'm surprised to hear this claim: about equations being easy in Word. I've written long documents in both, and (1) the math layout in Word is crude/inelegant; (2) there seems to be no easy way to change notation in Word. If you define macros for variables in Latex, you can change notation as needed by changing a couple of macros.
This pair of problems has driven me away from writing any but the simplest of mathematics in Word.
I've also found the symbol set in Word to be too small, but maybe with the universal adoption of Unicode this has been fixed.
It's not just in Word, it's Unicode in general. Unicode doesn't even have all of the letters available for superscripts and subscripts. I think q is one that doesn't exist as a subscript in Unicode. It's a good system but isn't math-ready.
Although I suspect it isn't as powerful, Pages will do TeX and MathML these days in its editor. For people coming from Word however, having to port their styles to another app is probably still a challenge.
I write primarily in Emacs and use markup. I don't try to convince anyone else to do the same. If MS Word (or TextEdit) work for you, great! I tend to write large documents for which I find those tools to be clumsy, but if your milage varies, no problem.
Latex lets you write equations quickly. Whether that makes it worthwhile to learn depends on how many equations you write and how quickly you want to be able to change them.
I also like being able to write my papers in vscode, with all the shortcuts and vim settings that I’ve gotten used to from coding. If Vim is second nature to you, it’s frustrating to have to switch over to ms word. Vscode has a good latex plugin.
That said, it did take some getting used to. Anything hard and potentially useful can become a “fetish” whereby we fool ourselves it’s the usefulness that motivates us but really it’s the hardness. I recall a time when I tried to use matplotlib to format all my plots, with labels and such. Horrible experience but I was really proud of these ugly graphs I’d make entirely from code. Then one day i just started saving them and opening them in illustrator for all the bells and whistles. Haven’t looked back.
This is often underestimated. It is possible to write complicated equations more quickly in LaTeX than in handwriting. I have done it.
However, to achieve this, there are provisos: 1) Macros already defined for commonly used structures; 2) Lots of repeated elements that can be copy-pasted; 3) Fast typer with a good text editor. 4) Keywords from amsmath, etc. committed to memory.
In college I sometimes livetex'd a class, i.e. wrote (full, not terse) notes for a course while the professor was drawing on the board. The above is all necessary, but this is easier than it sounds because speakers pause to let words sink in and to write on chalkboards.
The only thing I've found missing from the Word equation editor is: 1) macros, and 2) the output just looks less nice than in PDF form. To be fair I've never exported but I still think Computer Modern has better glyphs.
There are cynical and practical reasons to use LaTeX.
The cynical: While most conferences in CS advertise both a LaTeX and Word template for submissions, if you submit an article in Word, it seems quite likely it will be desk-rejected for being quackery, possibly without even being read. Is this fair? No. Has the scientific/mathematical community lost out on any great ideas due to this prejudice? Probably not. Think of this as the ultimate, final form of the clear plastic binder from Calvin and Hobbes, except Calvin was right all along and you probably can get away with saying that bats are bugs if you write it in LaTeX. This says more about the process of science than it does LaTeX.
The practical: I can think of three pragmatic reasons.
It seems that nowadays, good science is collaborative. Single author papers are rare. How is it that multi-author papers are written? Writing in a text-based markup language allows co-authors to easily send each other changes and keep a versioned history without having to worry about "track changes" and using file names as tags, and worrying about cross-compatability between ancient version of Office on Mac and Windows and that one person who only has Linux and so opens things in Libre.
In my experience, people do want to get into the weeds with formatting in LaTeX for many reasons, good and bad. For example, how can we shave off two lines of text so we're within the page limit (yes, even though the proceedings will all be published electronically we will get desk rejected for being one page over the page limit where the only thing on the last page are two lines from the bibliography, c'est la vie). That is a bad reason. But good reasons could include improving the look and feel of a document, or controlling where figures are placed, or drawing figures using TikZ. Additionally, having the sum total of the formatting in plain text has advantages. Do you think that undo really un-does everything in Word? Can't you remember some times where you do something, but then un-do and then do "the same thing" again and a different result occurs? LaTeX has many foibles but that is not one of them. Though, of course, sometimes I will wish you luck in understanding exactly why what you wrote in LaTeX works the way it does now. But for example, it is possible to define new mathematical symbols, operators, etc. in LaTeX in a way that is not possible in word processors.
The third pragmatic reason is that everyone else already uses it and some day or another your text will need to be turned into LaTeX for publication. I used to keep a lab notebook in Markdown and sometimes I would feel the desire to copy and paste some text from my lab notebook into a paper (if what I wrote was particularly lucid or well written, or was at least a good place to start from). This almost never ended well. If you have a draft of the paper in Word you might easily spend many days converting it from Word into LaTeX. If it's sufficiently complicated you might never get it converted. This could be a barrier to getting collaborators or submitting at all.
I think that everyone that works with LaTeX, knows, deep down, it is absurd, and the arguments the author runs into are not really arguments anyone would stand by if pressed. I think if anyone is going to write something entirely in prose, that they will never care about the versioned history of or want collaborators on, and don't care if the finished product looks good or bad, you can use a word processor, or at least, this is my criteria for whether or not I use a word processor. Am I drafting an outline of how we're going to write the proposal? That happens in Google docs. Am I actually writing the proposal? That happens in LaTeX, because I want control over the typesetting to fit in the page limit, I want it to look good, and I want collaborators to be able to edit the document and send me plaintext patches rather than docx with track changes maybe or maybe not enabled.
> if you submit an article in Word, it seems quite likely it will be desk-rejected for being quackery, possibly without even being read.
I've never had to submit source files for review. Maybe it's different in your field? In my field (robotics) we submit PDFs for review, and only submit source for publication.
What journal/conference are you submitting to? I haven't come across any in CS/Math/EE that even have a Word template to submit with. Every last one has a latex template.
Think of this as the ultimate, final form of the clear plastic binder from Calvin and Hobbes
Brilliant. Almost everyone in economics presents using LaTeX (Beamer) for exactly this reason. The outcome is that almost everyone's slides are composed primarily of tiny writing, and are therefore awful.
LaTeX indeed seems overly obscure and the fact it mixes the content with formatting does feel bad but I still can see no viable alternative. It seems the only code-based typesetting system (am I wrong?), the only typesetting system that provides such a degree of power and arguably the only way to produce beautiful PDFs with reasonable ease. So far I have chosen the way of writing well-structured HTML or Markdown (to separate content from formatting), rendering it to LaTeX with a Python script and using LuaTeX to produce final output.
Your workflow is very in-line with the author's recommendations though. His idea is to use something like pandoc to convert word-processor or markdown writing into latex, and then letting latex do what it's actually good at: Typesetting.
I am pretty sure the author of that blog post has not done so. Everytime I used pandoc or similar tools to convert text from one format into another, I had to put in a lot of effort to manually fix up things. Its not realistic for most cases. To-latex conversions are actually among the more developed paths, and all instances I can think of (pandoc, sphinx, dblatex) produce kind-of-weird output.
If you cannot stand latex but want to produce print-quality PDFs, Apache FOP (with Docbook) probably is your best bet. Otherwise your kind of bound to introduce a heavily manual processing step between writing and formatting for print.
I wrote 3 years of papers in LaTeX (ancient philosophy, so minimal math). I generally did markdown>pandoc>xelatex. I never really had to touch up the final output (except for a single script I made that removed some incorrect line breaks): I would just use inline LaTeX as needed to get things like figures into the output.
I wrote my thesis (200 pages) in LaTeX and I cannot imagine not having to do the final touch ups in LaTeX. You tend to get lots of line break issues, figure location and orientation and spacing issues, vertical spacing issues, that you have to touch up to make it look nice. It sounds incredibly impractical to do your writing in markdown and then use pandoc to convert to LaTeX, do dozens of touch ups, and then get your final output, send to your reviewers, get your reviews and recommended edits, do more edits (okay, now I can't go back to markdown anymore and redo the whole conversion process), more touch ups, rinse and repeat a few dozen times with your reviewers. Might as well stick with LaTeX in the first place.
That's a pity there is no visual word processor designed right for this. In my vision of a perfect word processor it should only manage the outline conveniently and let you write the content but have no features that let people use it the wrong way: no way to directly set font of individual elements (only let you mark them as headings, quotes, emphasized spans etc) etc. Of course I can do this by writing HTML or Markdown manually but there still are many people who won't (some because it seems unnecessary hackery to them, some because they actually lack mental capacity to get it) and even me I would appreciate a tool that could make it this more comfortable. I have described idea of a perfect writing tool here: https://news.ycombinator.com/item?id=18858538
Generally I would say that if your document doesn't include at least one equation it's not worth learning LaTeX to do. If your writing is going to involve lots of inline maths, or you're submitting to one of the journals that requires it, then it will make your life much easier.
(Word does now store equations as XML internally, but it's pretty horrible - 5 kilobytes for the quadratic formula, for example)
Edit: I've just remembered that when I was at university in the late 90s, there was a small community of non-maths LaTeX users ... who were doing things like New Testament Greek. There may be other specialised use cases where having fine control over WYG is more important than WYSIWYG.
Now I've gone looking for it, I can find an example : http://jcsu.jesus.cam.ac.uk/~mma29/essays/dissertation/ - I think the intent of using LaTeX here was really for Bibtex, the bibliographic system, which is an extremely important part of all true academic papers.
> There were also some development effort to “easy” LaTeX typesettings (for instance ConTeXt) but they not became widespread as “WYSIWYM” editors such LyX or TeXMacs, IMVHO because they are in fact less flexible than plain LaTeX and plain LaTeX is easier enough, at least it’s give you a easy and simple “learning path”.
ConTeXt is a substantially different beast to LaTeX. It's not LaTeX, although there are some LaTeX-inspired packages and tools to help migrate. A lot of ConTeXt is now written in Lua, and the versions in TeXlive are now sufficiently up-to-date to work from the manuals.
ConTeXt takes the view that it will start with sensible defaults and allow nearly all to be customized. Modules tend to play nicely together in an orthogonal way, but a lot of documents can be written without importing any modules.
The one thing I wouldn't say about ConTeXt is it's easier – you probably won't find the answers on stackoverflow like for LaTeX. It's a much more complicated system and you need the reference manual. But if you know how you want the page laid out and don't want to have to hack away at a style file, then it's a very powerful tool for serious typesetting.
Context seems to risky to use, its being maintained by a tiny number of maintainers, as of sometime I go, (I think) the source repos were not available publicly. Also it is difficult to install, and hard to find documentation for.
> its being maintained by a tiny number of maintainers
That is true, but it is quite stable and it works well. The last released version is from 2015. The last LaTeX release was 2e in 1994. LaTeX 3 is apparently still being developed by a small number of maintainers. [0]
> (I think) the source repos were not available publicly
From the homepage: "There is no offical repository there the development is going. The source code is released irregularly and placed at the http://www.pragma-ade.com website."
There's a git mirror of the releases.[1]
> Also it is difficult to install
It's in TeXLive and works fine. There was a time of intensive development a few years ago, and TeXLive decided to include older, more stable versions.
> and hard to find documentation for.
The main documentation is the reference manual, linked to from the home page [2]. That said, it's a real shame there isn't more web-friendly documentation. The Wiki has some commands documented, but others are either missing or out of date.
I have used ConTeXt a fair amount and I like it quite a bit. I would actually say it is easier than LaTeX, in the sense that it is more discoverable. After you use it for a bit, you see that the macros are all built on each other in predictable ways, so they all tend to take the same options and achieve similar effects. In LaTeX, the answer to your questions is usually "there's a package for that," whereas in ConTeXt, the answer is often "this macro takes the same options as this other macro, so just pass them in."
What is different about ConTeXt is the prevailing attitude towards using plain TeX from it. LaTeX really seems to consider plain TeX quite unsafe. I suspect this is because it is hard to build a resilient declarative system on highly weird and procedural TeX. ConTeXt, on the other hand, kind of encourages you to use TeX directly. So that led me to learning more about plain TeX, which now seems much less scary to me than when I was using LaTeX.
> I would actually say it is easier than LaTeX, in the sense that it is more discoverable.
Agreed. I think what I was trying to say is that LaTeX will have more answers on Stack Overflow. However, once you've found the command (and you can get an awful long way with the ten listed at the beginning of the reference manual) then the options are well documented.
> What is different about ConTeXt is the prevailing attitude towards using plain TeX from it. LaTeX really seems to consider plain TeX quite unsafe.
This is very interesting. I've found the same difference but in the opposite way. I've had to resort to using plain TeX in LaTeX quite a bit, usually buried in an environment or a command, to achieve what I want. And yes, it can be fragile, but sometimes it is the only way without using KOMA-Script.
I've found that ConTeXt has usually already considered what I want to do as a use case, because its scope is quite a bit more broad than LaTeX.
Most of my questions on tex.SE are about ConTeXt, simply because all paths in ConTeXt are less explored than LaTeX. The frontier is nearer. But I have never failed to get a useful and informative solution from one of the handful of experts.
I wish I had sources for my other comments handy. If I have time I will try to dig some up.
A perfect example of the Blub Paradox wherein the author cannot realize the power of a tool superior to what they're used to, and dismisses it as useless because "I don't need those features".
Counterfactually speaking, how do _we_ know that we are not the ones looking 'up' the power spectrum, as it were?
(I am joking of course; even though I would not say that LaTeX is the 'best' solution for all use cases, I can at least state with confidence that it satisfies some properties that traditional word processing tools do not have.)
I do feel there is bit of Latex worship going on. Now I think, markdown/asciidoc that boils down to DocBook is better that latex for most use cases. Markdown more clearly seperates logical formatting, from the rendering dependent physical formatting for different kinds of output. Moreover latex tables are a pain.
Maybe I this is my misunderstanding, is something the wrong with 'composability' of constructs in the LaTeX language? A concrete examples eludes me presently, it is like the latex functions/forms cannot be nested arbitrarily like normal programming language blocks/constructs.
I agree. I use and like LaTex but also use other things as well and am a bit surprised something better hasn't come along.
My overall sense is that the underlying reason for this is that supporting math in typesetting is an extremely difficult problem, much more so than people realize. Most typefaces don't come anywhere near being adequate for math, and those that are functionally ok are often really not very good aesthetically speaking. The equation typesetting can also be really difficult, in terms of "under the hood" implementation issues as well as the specification of standards. etc. etc. etc.
As a result, it's not so much LaTeX is good, as much as it is that it's been around for so long, and the domain is so much of a slog, that it's hard to develop something better.
I think a lot of people have a limited vision of what things could be like, and tend to base impressions on limited experiences with alternatives. The proprietary nature of Word does create problems in sharing written documents sometimes; the "black box" nature of WYSIWYG programs in general can be infuriating when you need to be careful with typesetting equations. On the other hand, constantly having every typesetting detail in your face with most LaTeX implementations can be intrusive and leads to "forest for tree" problems sometimes.
There are lots of beautiful texts typeset in LaTeX. But there's also a lot of LaTeX documents out there containing correctly typeset equations that are nonetheless really unpleasant to read or look at, for visual design reasons.
I'll bite as a LaTeX and Google Docs fetishist. Don't worry, I'm not taking this argument too seriously but if you happen to find some good arguments in there, please do tell!
> Seriously, anyone who believes that making people type this…
Just an introductory argument against it, not going into that. I'm looking for meat. Hmm... meat :)
> I know a number of academic authors who seem to spend considerable amounts of time doing that. I shan’t say that this is worse, but is it really better?
I did that too, but just because I wanted to be geeky. Was it useful? No. Would I have done this in a word processor? No. But the fact that I get to enjoy LaTeX by being geeky gives me a good feeling to return to it and write in it. Sounds like I have a fetish? I do! And not only one ;-)
> When it comes to stopping people from creating documents in purple 28pt Comic Sans, teaching them all to use LaTeX is a lot less efficient than stating that you will refuse to read anything that doesn’t match the style guide.
But LaTeX does have better alignment and justification algorithms. You won't see the difference, unless you're starting to compare it with a word processor [1].
> LaTeX does less to prevent authors from getting on with writing documents than TeX does. But if neither of the two existed, and you had to come up with something, right now, in 2016 – would it really be a markup language?
It depends, I see that people use Markdown for these reasons. And for my resume writing this is definitely true. The template also looks quite nice and I wouldn't know how to design a beautiful resume.
> we have LaTeX evangelism and the false implication that word processors don’t facilitate structured writing at all.
I wrote big documents in Word, Google Docs and LaTeX. I prefer LaTeX (my Gdocs fetish is with short documents or any psychology paper I'd need to write collaboratively).
> In sum, the case for writing in LaTeX is more than a little weak.
My reasons:
1. Automatic reference list generation (never saw psych. students use it in Word). I still don't know APA but I never got minus points for my bibliography.
2. The ability to comment. This has been huge for me (as in 10% to 20% better grades in writing huge) Yes, you can do this in Word and yes it also distracts me, when I do this on Overleaf it doesn't distract me. If I want to read, I read the PDF, if I want to write or edit and get meta information, I type.
3. Super quick restructuring thanks to include. There have been many times where I needed to quickly move +30 pages to a completely different location. It just takes one line.
4. Crash resistant: one time there was this psychology student who had her Word document crashed and unsaved. I had to use the command line to copy some temporary saved file that Word didn't recognize (apparently...) and tweak it in order to restore it. It took me an hour. LaTeX is a text file.
5. Tikz: I needed to create a graph and wanted to do it programmatically. Tikz is love, Tikz is life.
6. Knitr: I wanted to be ultra precise and focus on purely reproducable research, with Knitr you can put R formulas into LaTeX and calculate stuff. I handed in my source and PDF, my psychology teacher thought I was a wizard.
7. Different thinking cycle: by hitting compile and waiting for a bit you get some idle time to mind wander. For me it takes the pressure off and sometimes makes me anxious to see the result.
8. Fetish effect: you start to care about typography and the like (see [1]). Sometimes it isn't the programming language itself but the community behind it that makes it so powerful (Hello JavaScript! NodeOS is wonderfully hilarious! ;-) ).
9. Wizard effect: psychology students would leave me alone because they had no clue what I was doing. It also made me look smart :P Is this an argument? Hmm... I have a fetish, you pick out the good ones ;-)
10. Placing images neatly immediately: oh wait, nope this is a downside. Unless you...
> To cut to the chase, LaTeX documents are very hard to read until typeset,
Not my experience on Overleaf. I compile early and often just like in programming since it is a programming language.
> LaTeX is, as already noted, a markup language.
It is also Turing Complete and I have used it to do bibliographic database lookups, replacement and insertions. I used to program a bibliography per medium item and it would automatically figure out what medium item it was (e.g. journal, online blog post, online video, book, etc.).
> disruptive of the text for human readers (including editors and the original author): looking at the screenshot, we see the text mixed up with a lot of symbols that are not part of the text
I'd recommend you to use something like Overleaf and see if that's still the case (disclaimer: not affiliated with them, I actually used ShareLatex before that, I only went to them because I had to since they merged -- ShareLatex is open source by the way, so no internet connection needed).
> By the way, one of the words in the screenshot is not right. It’s a typing error that I deliberately inserted. I know where it is because I put it there, but looking for it is hurting my eyes.
Again, use something like Overleaf. They have spell-check.
> But that isn’t how most people want to write.
Most people also don't want to program. I won't recommend LaTeX to my girlfriend either, but to fellow programmers, why not? Maybe they like the program/debug cycle more.
> Never mind boilerplate code like \documentclass{article} or \begin{document} at the start of your document – running into something along the lines of \parencite[706]{lena_peterson_2008} in the middle of a paragraph and having to mentally parse it into ‘(Lena and Peterson 2008, p. 706)’ interrupts your train of thought and makes it harder to do what you really need to be doing: reading your punctuated words back to yourself to make sure that they ‘sound’ like what you really wanted to say and don’t have any mistakes in them.
I disagree. I never wanted to read my citations full out and I'm happy I can read them by aliases.
> This is what LaTeX is good for: not helping people to compose text, but helping them to make it look nice. If that is important to you, go ahead and give it a look.
LaTeX is good for long pieces of text. I never need to scroll millions of pages since everything is in a folder structure that I can immediately click on. Making things look nice is very important and even a slight competitive advantage, it makes you look more credible.
> I think you’ll probably agree that the LaTeX version looks better than the word processor export version. Whether it looks sufficiently better to justify the additional effort is a judgement call that you’ll have to make for yourself.
I think that students can invest the time. It took me about 40 to 80 hours to get a good flow with LaTeX. After that I didn't have any issues anymore. Also, whenever I got issues it was a fun procrastination activity (students like that from time to time).
I have to say your points 1 and 5 (which in my opinion are the same thing) is the killer reason why I write stuff in LaTeX (apart from the fact that I hate Microsoft Word with a loathing and find LibreOffice even worse). The other thing is that a LaTeX document is text - I trust my text editor to not mess it up - if I make a mistake with the format and break the document, all the text is all still there, so I can go in and fix it. With a Word/Libreoffice document, the software is doing all sorts of stuff behind my back, so I have no clue what it may have broken, or how to fix it.
> 1) can't have proper a version control with those binary documents word processors generate.
And even worse (maybe): no way of fixing errors later on; arguably, the error messages of LaTeX are obscure for new users, but at least you can always reset your document to a working state.
The article author should have tried LyX [1]. It's a graphical editor but still uses LaTeX to produce the output. (Its main downside is that its LaTeX import functionality is very patchy.) It has support for a large common set of LaTeX functionality, and there are a few ways to extend it e.g. user macros and modules. If all else fails you can directly insert raw LaTeX in your document; obviously this defeats the point so is best avoided where possible, but it shows that it can ultimately do everything LaTeX can.
The graphical representation shown on screen is somewhat approximate e.g. fonts and sizes are approximate, and line wrapping is totally unrelated to the final output. But it's good enough to understand what's what, and most importantly it's free of all the markup clutter, and that's what the author objects to. It's especially good at this for equations, but it makes a different even just for regular text. Plus it has autocomplete for LaTeX commands (I realise that there are LaTeX editors that do this too).
A lot of people mistake LyX as being a beginner's tool to avoid understanding LaTeX. I would say it is almost exactly the opposite: you still need to understand LaTeX to use LyX effectively. Rather than the markup being right in front of you it's hidden by a layer of indirection, which of course is the whole point, but it makes it harder to understand what's going on when something breaks. It breaks less often (e.g. you never get mismatched close braces) but the hard errors still crop up (e.g. incompatible packages giving really obscure errors about macro expansions).
I've written numerous mathematical documents of various lengths in LyX, and prior to that a few documents in LaTeX including my MSc thesis. When I first discovered LyX I was partway through writing up notes from a lecture course I was attending. In LaTeX it took me about three hours to write up notes for each hour lecture; in LyX it took me one hour per hour of lecture. A factor of three speed improvement! I can hardly imagine how long my PhD thesis would've taken without it.
Formatting a difficult equation using LyX is soooo easy because you throw the TeX in and see the result right there. Plain TeX with big equations has a lot of guess-and-check with it, and Word just fails. LyX is definitely a fantastic tool and I use it as my main document writer.
Org mode can render equations in line. It's got the easy-reading benefits of lyx but it's all plain text. Then easily compile to latex or pdf. By far my favorite way to do writing.
From what I've read, AUCTeX (which is also an Emacs extension) is better for writing full-blown LaTeX documents. But yes, the preview functionality in both is a big improvement on directly viewing the source markup, and if I couldn't use LyX then I would certainly use one of those instead.
In fact LyX also supports instant previews of equations. That might sound a bit odd since it already displays equations graphically, but it's nice to see a pixel-perfect display of them. It also supports wrapping instant preview of arbitrary bits of the documents, which is really handy when you have inserted explicit LaTeX e.g. a Tikz diagram.
Not to mention that LyX has lots of features a graphical editor can offer you that editing a plain text file cannot. For example displaying the document structure (tree of headings, subheadings, etc.), and allowing headings to be promoted/demoted/reordered (while taking all text and sub-headings and their text with them).
The idea of somebody critiquing Latex for being a difficult format to write in while writing almost exclusively plaintext is pretty baffling to me. It's like saying that Photoshop is too complicated because I can draw a square just fine in Paint.
My domain is pretty specific - typesetting chemistry papers, but for that it is absolutely invaluable. There's no tool on the market which can do it with anywhere approaching the ease. Even MSWord, which is now good enough that I can type simple maths about as fast as I can in Latex (although frankly even there I prefer Latex to futzing with a gui), doesn't have nearly the tools I need.
I'll bet this is true in all sorts of smaller domains, too. Typeset anything unusual enough not to have good specialist tools and the freedom of Latex typesetting will make up for the learning curve in weeks if not days.
>The idea of somebody critiquing Latex for being a difficult format to write in while writing almost exclusively plaintext is pretty baffling to me
Creative writers for instance don't really care about typesetting; they want to keep their manuscript organized and to be able to change it easily. Here's en example[1] of a writer showing his personal workflow for writing (in this case using emacs) and you'll see what I mean
A few points. First, he is an academic in the humanities. According to his article, most people around him talking up latex tell him the main reason for picking up latex is so you don't have to worry about design. There are not much equation-writing in the humanities, so that's outside his domain.
Second, he specifically does encourage using latex to typeset. Just not to type in the first place! Use Markdown, or a word processor, says the author, and then latex do what it is good at: typeset. But don't write long paragraphs in latex with attention-breaking code scattered around the text, since that places obstacles in the path of editing.
I recognise his area, but he's pretty general in his commentary "Don't use it to write (anywhere), it's just for typesetting", so I feel justified in using my similarly limited domain to critique.
But I don't know how you expect me to write a chemistry paper without writing the equations? Just leave them blank? Typesetting is instrumental to simply composing the paper, it's not something I can do after. As I said in my initial comment, nothing else will actually do the task as fast as Latex, so it'd be pretty perverse to type in something else. Maybe it works for your domain, but not mine.
I am a Mechanical Engineering PhD student, with a whole lot of equation writing to do myself. And as someone who has very briefly dabbled with latex, I have found MathType to be the better solution for myself. I just cannot get myself to type out latex equation flawlessly first time, and MS Word's equation editor is an abomination. MathType has a rather large list of shortcuts requiring memorising (made more difficult by my dvorak layout), but once learnt, equation writing proceeds at an enjoyable clip. It has the secondary benefit for resulting in WYSIWYG, to match with the rest of my MS Word workflow.
MathType looks decent for maths (and I assume engineering). I can see the appeal of the WYSIWYG, though I don't really see how learning a bunch of shortcuts is materially different from learning the Latex commands.
Nevertheless, it still doesn't have a bunch of the symbols I need, which was my point about smaller domains.
I'm glad it works for you, but given the advantage seems to me marginal and it offers no options for extensibility, I'm still unconvinced that we're at or even close to a stage where Latex has no place in writing documents.
> Free and open source software has a strong tendency towards being difficult to install and get up and running. TeX and LaTeX are no exception.
Setting up a LaTeX environment in Windows can be difficult if you do not want to install one of the very big suites (although being honest, who is worried about 2GB when you need more space for something like printer "drivers"?). However, in most Linux distributions, you just choose LaTeX in your package manager and it works. And this is the case for most free and open source software. I do not know about Macs.
There is a recent development that provides a much smaller subset: TinyTex by Yihui Xie (who has made many contributions to Rmarkdown). See https://yihui.name/tinytex/
One can add just the packages you need and it can be installed in user directories for people who do not have admin privleges.
253 comments
[ 11.6 ms ] story [ 260 ms ] threadhttps://pasteboard.co/HVEPxvy.jpg
Please consider using a simple ">" instead :)
(Which isn't to say that HN doesn't need to fix that shit. But it's still your fault if you don't account for it.)
He doesn't say "Never use Latex!". His opinion is that those who are picking it up under the false illusion that latex gets out of there way better than MS Word are doing themselves a disservice. Seems to me to be a fairly uncontroversial opinion. The latex document, especially one that's big, is not trivial to parse and edit. Making sure the 'flow' of the text is right is infinitely easier with WYSIWYG.
He absolutely seems to be OK with people who use latex because they care deeply about the beauty of the resulting document, and himself advocates for a post-writing conversion to latex. Markdown>Pandoc>Latex even. I feel many here are not reading the article in context.
I have an aside of my own to add to the "latex gets document design out of your way" crowd. I can bet any money that the types of people who will fiddle a lot with the typeface settings on MS Word will fiddle with themes/linespacing/monospaced font-du-jour on their text editor too. It has less to do with the writing environment and more to do with personality. I would know since I'm one of those people.
I don't even have the impression that his proposed alternatives (conversion changes from other formats to latex) are really used by him.
Ah, I like writing documents in Docbook, so I am not even that kind of a latex enthusiast. But his criticism is kind of really off.
One of the main things he does point out is that the justification often used to promote LaTeX is complete bullshit. "You should use LaTeX because you spend too much time trying to worry about what size your title should be in Word." No, I don't.
He also points out that, even if you prefer writing markup yourself to WYSIWYG, LaTeX isn't exactly the simplest markup. Stuff like restructured text or markdown (hell, even HTML in many cases!) has much less intrusive formatting information than LaTeX does. LaTeX is designed for final typesetting, and it has too much overhead for the majority of writing cases where you don't care about final typesetting at all.
The larger hurdle (IMO), is that TeX can be extremely finicky even compared to HTML. To get a document that looks mostly the same as the one you got out of Word takes time and a fair bit of setup. I went through this same process with my CV after I finished my MA.
Once it's done however, updates are quick and require almost no effort.
This isn't to say that there are not benefits to word/wiki, mind. Just, "where everyone else is" is not the most satisfying of reasons.
Using the "print" @media type and converting from px to inches only gets you so far. For something that I expect to be printed at some point, TeX is a much better place to start.
That said, the stability of the flow of TeX and friends lets me not worry about anything having changed on me in the past few months/years. I think this has gotten better in the HTML world, but it is still a much shakier foundation.
I think the blog post was okay, but could have used some editing as I ended up skimming some of the points he was trying to make but that's kind of what blogs are about.
Agreed, people are having a strong reaction to a criticism of LaTeX that the author isn't actually making.
> I can bet any money that the types of people who will fiddle a lot with the typeface settings on MS Word will fiddle with themes/linespacing/monospaced font-du-jour on their text editor too.
It pains me to agree with you, but I'll agree that at the end of the day, being concerned about typesetting is probably a bit superfluous.
I think you're missing the point though. There's a strong value to standardization in formatting when it comes to typesetting, can you imagine how annoying/distracting it would be if serious journals had minor differences in formatting? I'm not being sarcastic here, it's a necessity when it comes to any kind of repeated publication. LaTeX solves this problem better than anything else in existence, way better than the limited themes and styles available in WYSIWYG text processors.
I made a nice fancy template with tcolorbox for q/a homework problemsets and now that's a solved problem for me. It looks better than anything you could possibly make in Docs or Word, it looks exactly the same every time.
It's also scriptable. For example, I was working on an amateur radio and I needed to save off the calibration settings before I worked on it. I took a video of myself scrolling through the settings, watched it once with mpv repeatedly hitting 's' to take screenshots, and then was able to interpolate the screenshots with the actual content of the Yaesu manual, producing a beautiful manual that described the setting in question, as well as showing what it was set to before I started working. It took me like 15 minutes, and I'm glad that I did it because it saved me a lot of time later on. Try doing that in Word, Libreoffice, or Docs!
Either way, I'll back out a bit and ask a more elementary question. Is a workflow possible wherein one writes in anything-but-latex (Word/Markdown etc) and then uses latex to typeset? Because I (and the author) agree with you that a) standardisation is a good thing and b) latex is excellent at typesetting.
What latex is weak at is me writing my paper in it, and then trying to read back whether what I wrote flows well.
Pandoc is a good tool for automating workflows that start with some other format and end with LaTeX.
Another good option if you like Emacs is org-mode, which has a LaTeX export, among other things.
Another very neat feature is Org-Babel, which allows you to write code in (almost) any language that is then replaced with its output, which is great for example code. It can even pipe stuff through PlantUML, allowing you to embed your diagrams into the source code rather than having to manage them separately.
As an example of what it looks like at scale(-ish), I used this to write my thesis paper[1], and all the other related documentation[2].
[1]: https://gitlab.com/PicoNodes/PicoNodes/blob/master/docs/Repo...
[2]: https://gitlab.com/PicoNodes/PicoNodes/blob/master/docs/Pico...
That is to say, yes, there is a workflow. It is basically to build the prose out separately from the layout. Latex is used to typeset the document you want to do. Not to author it. By the time you get to it, you probably should be at the point where the markup is what you care about.
That make sense?
Yes, there will be times where you care about how the document looks. That is likely an iterative process and many of us find value in separating the quibbles over the look from the authoring of the content. Both are valuable. Doing them both at the same time is ridiculously hard.
However, this won't always be the easiest method. For example, anything math-heavy will almost certainly be easier to type directly in tex markdown.
When people argue that latex gets out of the way, the road being spoken of is the one leading to a beautiful document, so the author is expected to care about the final appearance. LaTeX and Word both come with learning costs, the initial effort in to just bang out some words is clearly higher in LaTeX, but they will look good from the start.
Yes, if you don't much care about the looks of the end result, Word is more expedient, but then so is notepad.
Tl;dr, That is ugly.
Next, I googled "russian pension reform pdf", and in the top three hits there were two reports with various levels of depressing Word defaults, and one that I'll give be benefit of the doubt and say was actually a conversion from a web page.
I'd say this is a fairly typical look for any report that isn't super official or actually a magazine trying to sell something.
The IIHS report honestly isn't bad looking. To my eyes, it could be done in TeX or Word depending on how complicated the styles actually are.
I view the two tasks as very different. One of the reasons I dislike Word and other Word Processors is that they crowd my writing views with formatting tools and suggestions.
A plain text editor is far more comfortable for me when I'm writing and Latex allows me to tackle the layout and typesetting work as a separate step.
I get the argument that people -- that is to say, programmers -- like to write in their text editor of choice, too. However, I think that's what it is: preference for the familiar. Writing LaTeX feels like you're programming and lets you use tools you're familiar with, so it appeals to people who like to write programs. That doesn't mean the extra effort is really paying off.
I think arguments that the results from Word aren't good enough or professional enough don't really hold water, either. Take a look at any of the opinions from the Supreme Court of the United States [0]. Download any of those opinions and look at the metadata. Some of them are scanned documents, but many of them say: "Application: Acrobat PDFMaker N for Word". You'll find the same thing if you look at petitions that have been filed, and they get quite complicated [1].
[0]: https://www.supremecourt.gov/opinions/slipopinion/18
[1]: https://www.supremecourt.gov/DocketPDF/18/18-306/63707/20180...
It may have gotten better I haven't actually used Word in probably 10 years or more at this point. The fact is that I can edit my document in an editor I'm comfortable with and I can typeset it with a tool I'm comfortable with and the two steps are pretty distinct from each other and I like that explicit separation.
But it doesn't really make a LaTeX user more qualified to criticize Word any more than OP is for criticizing LaTeX, does it? Is "I just don't like chasing down the various toggles for it all" really a better argument than "I find reading all those codes aversely difficult"?
Contrast this with the current Word system. I have had a non-trivial number of times where it will start doing things and I have no idea why. Probably logical, but I can't see the inputs to the logic. Only the output, with some idea that I should be able to drag things. Doesn't always work, and sometimes "undoing" the drag of an items is a non-trivial task.
To write a conference paper for submission, the steps to start it out are generally:
a) Copy the journal's LaTeX shell file.
b) Copy some other Makefile to actually convert TeX files into PDFs. (Because it turns out that this is simply too complicated to actually make a single command, especially when you have a bibliography and several images to import).
c) Okay, now we can go through the process of actually setting up the master TeX file, separate TeX files for each chapter and start typing text.
d) Include an image. Weep at how annoying this is, and how inconsistent documentation is.
e) Time to tweak the article to get it to fit the page count. Don't you love all of those little cycles where you have to wait for a half-dozen commands to run to figure out if you've shortened a single paragraph enough to pull the section header up to the previous page? And this is assuming you have a PDF viewer that will automatically refresh the display for you if it detects the PDF file changed.
In my opinion, the two great things about LaTeX are its syntax for math and BibTeX. But I wouldn't contend that these are unimprovable (particularly BibTeX). But it spends too much time making you concerned with publishing concerns that trying to just jot down notes or other early document planning is surprisingly difficult.
Whereas I (not claiming a universal) can only use Word for small documents; when they get big, with revision tracking, chapters, etc etc I've found it becomes fragile. We had a 1600 page FDA submission garble itself shortly before we had to submit and had to go find older versions from backup tapes (!), re type stuff from old printed out copies, and check every number in every inline data submission which caused major panic and stress. Everybody we asked for help said, "don't use word for a large document"
Maybe some people are able to do it.
It was worth it all the trouble? oh yes! :-)
Does Emacs have a reputation of being difficult to use? I remember that in the late 70s and early 80s MIT's CS and AI labs were proud that the (not specifically trained) secretaries used Emacs and wrote macros for it.
It's been decades since I was a novice user so I don't remember the experience, but was the first editor I'd ever used, and, years later, the first editor my (arts major) wife ever used, and I don't remember any particular anxieties or consternation from either of us.
Yes. And as it usually goes, it's an undeserved reputation caused by it being different from what most people are used to.
Emacs was just more dificult than the other programs that I tried, and I tried a lot: word, antiword, abiword, starwriter, openoffice, ed, joe, elvis, nano, vim, xemacs... okay I will bang my fingers and head against emacs and see what happens
I realized later that I was doing it all wrong when I started to slowly unfurl the capabilities of the program. Couldn't be great if...? and, suddenly a wild major-mode appears. Not more viruses hiding around, not more big files that decide to crash or lose their format five hours before a presentation. Not more I updated the program and now I can't open my old files anymore.
Very well time spent in retrospective. I use it every day. Same for LaTeX that was much more dificult to grasp but even more rewarding in a lot of unexpected ways.
There's a Knuth quote about how the aim of TeX was to allow anyone to create properly typeset documents. I can't find the exact source, but https://www.tug.org/whatis.html should give you some idea.
This individual wasn't able to. Last time I looked at TeX it (or metafont specifically) often seemed to default to bitmap fonts. That's a poor default - a bug by today's standards - but Knuth has a different definition of 'bug' than you and I do. Maybe you should spend less time judging users and more time working on improving the software you're advocating for?
PostScript is its own can of worms. People who need typeset documents shouldn't have to concern themselves with it.
It's not 1994 anymore, alt.sysadmin.recovery is dead, nobody knows what a 'cluestick' is or what code ID-10-T is, and people generally have stopped blaming users for poor software - which TeX, by Knuth's own standards is.
(Emphasis added by me)
I have probably read every single word Knuth published on the topic of TeX, and what you said seems very different from what I've gathered of Knuth's philosophy and style. Knuth intended TeX to be used by someone who cares about typesetting, and the program comes with a nearly 500-page manual (The TeXbook) that was very much intended to be read. In fact, he expected anyone who was preparing any special sort of book (a multilingual bible, or dictionary, or whatever) to change the program itself, and has expressed surprise at people using elaborate macro packages like LaTeX or doing everything in TeX macros.
(BTW the page you linked at https://www.tug.org/whatis.html is also pretty good and note that it says a lot about typography and the finer details of typesetting, and less about ease-of-use: search the page for "support" and "TeX is not the tool for you".)
> often seemed to default to bitmap fonts
Actually, TeX uses only font metrics, and the output of TeX is merely a set of instructions about where to place what characters from what font. (A compact binary encoding thereof, in the form of a DVI file.) A separate program is needed for actually using the font and rendering the output. (These days everyone uses pdfTeX or some other program that combines the two and uses vector fonts, but that's the original design.)
In any case, for the original task that TeX was designed for (completely specifying the appearance of a book, in a future-proof way however technology may change), it is absolutely crucial to combine it with a program for bitmap fonts (and that's Metafont -- it produces only bitmap fonts). Think about it: the rasterization (converting to bitmaps) has to happen somewhere: if you leave it up to some other technology (the one used by the printer or whatever), then you're back to the same problem that led to the creation of TeX, where in a few decades fonts may be produced by some completely different method and you may no longer be able to produce a book whose appearance is identical (except for the typos you fixed or minor rewrites or whatever) to the earlier one.
Requiring someone to read a 500 page document to produce a typeset work does indeed indicate a conflict - I would spend some time finding a URL for the Knuth quote but HN really doesn't like people discussing TeX or Knuth in a negative way, and this hasn't been a very productive conversation so far.
There are a real reasons academics and quasi-academics (aka me) use LaTeX instead of something like libreoffice. Equations are basically impossible without it. I can't write down a complicated relation in LaTeX as fast as I can think of it, but it's a lot closer to that speed than laboriously constructing a relation in the equation editor of any word processor. If you do something like physics, machine learning or topology, trying to write a paper in libreoffice or whatever they call Word now is sort of like trying to write a novel on an iphone. The author never mentions equations. Yeah, it's more difficult for a n00b to construct, say, a letter to the editor in LaTeX than it is in Libreoffice -that's not what it's for! Though of course once you have paid the toll of learning the thing, it's pretty quick and in fact lives up to its promise of not getting in the way of writing the content.
Beyond that, my equations will be accessible for decades and will be completely platform independent. If you have an academic career, this is probably reasonably important.
There's other reasons besides "equations" and "long term stability" (BibTex is probably top third reason) -but those are the big ones.
Edit add: beamer mode is also a good reason to use it once you know it. Take your paper; turn it into a presentation -never leave the same application.
Also, I haven't used it but newer versions recognize handwritten equations and formulae and automatically transcribe them to the equation format.
On the flip side, LaTeX definitely takes longer to write, but it's less frustrating than having to deal with Word's (well intentioned) autoformatting.
Depending on proprietary stuff which changes all the time, or even complex FOSS software with single point of failure developers (I dunno, TeXmacs or something) is a huge risk which went away when I bought a cheap Leslie Lamport book on LaTeX. Doesn't really matter if a new feature in proprietary application-X allows me to also use it with LaTeX. I can just use LaTeX!
I wrote my dissertation in 2004, can still trivially use the LaTeX in it, and I sometimes do. Libreoffice, Word ... pretty sure I'd have a harder time of it. Heck Libreoffice was probably called StarOffice back then!
I can attest to that from first-hand experience. Having done a thesis in LibreOffice once, I wouldn't want to repeat the experience.
That's not an absolute negative against LibreOffice, though. I still use it often and like doing so. But a large document with many illustrations, many text-internal references, code and a non-trivial bibliography? Never again.
And if I never saw another beamer presentation again I'd be a happy man.
Also curious if your field involves presenting a lot of mathematics. If I were a test-tube biologist or a lawyer, I doubt I'd use it.
Mostly, for beamer, I find that it encourages three things:
1) Very uniform formatting that keeps everything in say, a conference, as sort of a blur.
2) Lots of "I'm going to skip this slide..."
3) Encouragement of the "A presentation is a paper you read to people" style.
None of these are unique to beamer, but seems to promote it (in my anecdotal experience).
In business world, beamer is the quirky boffin presentation rather than the workaday powerpoint in corporate colors/fonts.
In my undergrad Honors Physics program, I was required to attend two different workshops on giving presentations (on top of the university requirement of Speech & Comms class) and a technical writing class. Suffice to say, my beamer presentations are not identical, nor so they include "extra" slides. It doesn't take much to drop slides, so it should be easy to comment out a number of slides for a tailored presentation. That's just a lazy presenter.
My personal experience with LaTeX is that it is actually rather annoying to do anything with images, so you seek to explain stuff in text (or equations) instead of trying to develop good diagrams. For presentations, this is exactly the wrong mindset to be in. A good presentation software should nudge you, if not outright force you, to making diagrams and pictoral representation over bulleted text slides. So I suspect beamer really is contributing to the problem of bad slides by making it even more time-consuming to develop good presentations.
One of the things I love about TeX is I can write it in my editor _without_ WYSIWYG. The tooling in Emacs to write TeX is amazing. I can never go back to a graphical writing tool.
Word processors are one of the worst environments I can imagine for actually writing. Lots of people surely write in LaTeX in part because they can use a reasonable environment for writing (i.e. a decent text editor).
Also, if you're in a decent text editor, writing LaTeX can be pretty easy (though for lots of things Org mode or markdown etc. may be more appropriate), and you can set up macros to auto-fill in things like the preamble etc.
I essentially write for a living, and I would not have survived had I not adopted LaTeX.
The thing is that time invested in learning superior alternatives is already spent, and there's no point taking the risks just to try for no reason.
Just like I will not create another website in FrontPage.
[1] https://www.unicode.org/notes/tn28/UTN28-PlainTextMath.pdf
It doesn't look all that bad.
When my wife wrote her doctorate she did none of those things, and I ended up spending the best part of two days laboriously going through the entire document to make it conform to the university's style guide for thesis being submitted. Thankfully she uses Scrivener for writing now which as near as possible enforces doing things the right way the first time.
I would submit that the same is true for LaTeX.
I've done plenty of work where the customer insists that the final deliverable be in Word format, because that's what they're comfortable with. You can do proper formatting in Word once you learn the tools, and it's not even all that hard, although it can be somewhat unforgiving of screwups. The one thing I wouldn't attempt is to convert a TeX document to Word. In theory that shouldn't be entirely possible, but in practice there are about a million ways it could blow up in your face.
But if you have even one collaborator who goes in and hammers something by applying manual formatting, the whole thing can be ruined. Actual images can end up showing up along with its caption in the Table of Figures. Entries can go missing from tables. Lone paragraphs can randomly end up with a different style.
Word can successfully manage large, complex documents, but it requires discipline. (as does LaTeX)
And I've yet to find a guide which instructs on how to set it up properly, how to share styles across documents etc. Do you know of one?
20 years of writing HTML and CSS has spoiled me; many times I'm using Word and thinking "I could do this with one rule in CSS" (often some crazy positioning nonsense with images/tables). If only!
I will admit, however, that LaTeX run on default settings looks good. Whoever designed that bit had taste.
With that said, I wonder if the people who actually did write a thesis/dissertation in LaTeX had to do all that customization themselves.
For example, I'm in NJ and although I didn't go there, I could imagine Rutgers OIT telling people, 'If you want to use TeX, just download our package, it will install everything you need and will output the correct format for submission'
I am pleased to report that the makefile for my thesis still works, and generates close to the original doc, 20 something years later. I can barely open a 10+ year old office doc thanks to incompatibilities within the product.
Back in those days, word processors would not do complicated multi-line equations very well; type setting for print, display, etc. This was a problem for me.
To each their own though. I wound up writing some user manuals for products I developed in the late 90's in LaTeX as well. I've not used it much since.
E. Allen Emerson once came knocking at the unix cave door, complaining that his older papers, written in LaTeX, were formatting differently.
Upon examination, the only difference, literally, was that we had replaced the ancient 300dpi printers with less ancient 600dpi printers. And yes, CMR is a spindley little font.
I've heard good things about Ulysses, too, but haven't tried it. I think that whole class of application is inherently better for editing large documents.
I stopped using it when it went to a subscription, though. I can't accuse a company that lets you write in Markdown, basically plain text, as rent-seeking, but I dislike my workflow being held hostage like that. Haven't really found a suitable, multi-document replacement yet.
Scrivener has much the same document model, but it's pay-once. That by itself made it much more attractive to me.
Don't get me wrong, I like Scrivener a lot, but in lots of scenarios it just doesn't work as a drop-in replacement for a more general word processor.
I have seriously considered offering to pay people with whom I'm forced to exchange documents (especially academic law review articles which tend to have thousands of edits) to learn markdown and git just so I don't have to eyestrain my way through that track changes screen
I write in Word as well and merging edits from collaborators with my in-the-meantime-updated document is really anoying and error-prone. I found https://www.simuldocs.com/, but haven't tried it yet.
You can also write each sentence on its own line - output will still put them in the same paragraph. But it's not as nice a reading experience of the markdown.
As long as you limit your lines' lengths git is just as good for text as it is for code. Of course you're correct that using it for paragraphs without line breaks wouldn't be that great, but my question would be why would you ever do that?
Having said that, an awful lot of tech nerds (arguably like myself) seem to seriously underestimate what word processors can do.
Back when I did write fiction in a word processor, and even now when I have to send fiction to editors who expect things in Word format because that is the lingua franca of the non-technical publishing world, I don't have to spend an awful lot of time formatting manuscripts. Word processors have templates, and hierarchical styles. I wouldn't need a macro to auto-fill in the preamble because I'd have a template that included that.
But if I did want a macro, I could have it! A lot of word processors do have macro languages! Many of them also have search-and-replace capability which is basically regular expressions with an easier syntax. For all the crap Word gets, much of it well-deserved, it's actually pretty good at editing prose. It's almost like they've been developing it for over thirty years or something.
If you're in a position -- which I am -- to use a plain text editor for all your prose writing, and you prefer it, that's terrific. And you could totally make the argument that if your final copy is going to be LaTeX, you should start in plain text, whether or not you're using LaTeX markup from the beginning. But the notion that Word is some kind of underpowered little toy no real author should ever possibly use is just, well, wrong.
I wouldn't characterise it as underpowered so much as wrongly-powered. Just the ability to use macros etc. is not the same as being a real text editor.
I actually started out using word processors, and really wish I had discovered proper text editors sooner. For me, word processors are just miserable working environments.
I find Word a miserable working environment. On the other hand, I find Pages a delight to work in. It's one of those "it just works" programs for me, and I find documenting editing so easy and intuitive.
It reminds me of those who swear by WordPerfect and wouldn't dream of using Word unless forced to; it's not the class of application they care about it; it's the specific set of tools, their presentation, and how the app matches their work habits.
(Same reason that I just can't use Google Docs, no matter how many workplaces make its use a soft requirement)
There should be a mode in word processors that disallows on-the-fly formatting and forces consistent use of styles.
I wrote a 200 page technical book in Word. After a couple days of customizing Word (few macros, hotkeys, customizing the styles to my liking), everything went well and I was able to do everything I wanted w/o trouble.
Word does have a learning curve (e.g. put all pictures inside a text box, it makes layout 10x easier), but once learned it is quite nice for a lot of purposes.
Woah, good tip! As I write more specifications and reports for management I'm using Word more, and there's a dearth of good advice online for getting decent results (it's drowned in Google by the "how to type a document" kind of guides).
Many times I've wished to write HTML and have it formatted like Word, as my styles go haywire.
Talking of styles, I'm surprised there's not a Syle interchange format for Style Sets - Microsoft's online resources don't have any to download. I was looking for a Latex-equivalent default template.
The text box thing is huge. I wouldn't be able to work in Word without knowing that.
Edit: Link! https://templates.office.com/en-US/Blank-and-General
“LaTeX, never has so much bad math looked so good...”
Tex is such a powerful tool. It makes me sad to see someone dismiss it likely because of unpleasant first learning curve, which is also why a lot of people dismiss vim or emacs.
When I was a freshman in college, an upper classmen friend advised me to learn TeX, and helped me through writing several documents. That advice & help was absolutely priceless.
Today I work in a typical tech company environment, and my team writes most of our technical documentation, sales engineering presentations, in-house tech presentations, and even recently some supporting documents for patent applications, all in LaTeX, all stored in version control in one of our repositories.
We can do real code review on technical documentation changes, have the power of many latex packages, beamer, etc., and even build versioned PDFs via our continuous integration tooling.
In some cases we also combine latex source with Mako templates and have actual software that can “push button” update all sorts of charts for new data, client-specific requests, etc.
It is a very powerful tool.
Thus you need to use some flavor of Markdown that covers what you need, which may not also cover another use you have. It still doesn't address output, so you need to eventually process the Markdown into something readable or typesettable, like LaTeX, which you'll need to understand anyway if you want to get the output you desire or fit into a style set given to you by a journal.
Since LaTeX is the lingua franca of math and CS, you're better off just writing in LaTeX and then your advisor/professor/reviewer/editor can see or edit your intent as well as your output.
Unfortunately, if you're in Biology, they're all over the place with their styles, and Word is the default. Perhaps then you will benefit from doing things in your own Markdown flavor and translating it to Word in the end using a little script you've written.
He is not arguing that latex is useless. Not even that isn't supremely powerful. His argument, in essence, is that a word processor is superior for writing and editing what you wrote than latex is, because there is nothing but text on the screen to get in your way. So, editing is easier.
He directs this argument in direct opposition to those who tout that the _main_ benefit of latex is that it lets you forget about formatting. The author is NOT contesting the typesetting benefits of latex. Nor is he saying "Only use a word processor, version control be damned". He is perfectly happy with markdown, for instance.
He is merely saying don't write in latex just because you think it's better at getting out of your way.
I did not presume the author made either of these points. However, the scope of latex’s capabilities does undermine the author’s point about editing in a traditional word processor — that is an argument someone unfamiliar with latex’s capabilities might make, and so referencing the fact that latex is a more powerful tool than this is a valuable perspective for refuting the author’s hasty dismissal.
> “His argument, in essence, is that a word processor is superior for writing and editing what you wrote than latex is, because there is nothing but text on the screen to get in your way.”
I actually dispute what you claim here, because the author spends a lot of time talking about the claimed benefit of TeX that it decouples presentation from content, analyzing it linguistically from written statements of that benefit rather than pragmatically (for example, like separating written text out entirely and using \include or other features to inject text into formatting code).
I think you misunderstood the article because one of the weakest points about it is that it starts by discussing this idea that TeX has a benefit from decoupling the visual presentation, but then the article conflates that topic with the topic of what is on the screen while you’re editing the source. Those two things are not actually very related, and you can edit content in TeX with as much or as little interlaced formatting or style commands as you please (in fact, that’s the whole point).
Not to mention that it facilitates actual code review as a means of collaborative editing of either content or presentation logic or both. Something quite hard to achieve with wysiwyg editors.
Your comment reads like a classic knee-jerk HN contrarian response to me. You’re in a rush to criticize.
For example, you say,
> “Your entire comment is about your workflow, without much in the way of engaging with what the author wrote.”
as if this invalidates any points or makes the original post less embarrassing. But that was my whole intention. I was trying to explain how TeX is useful, because readers of the post might mistakenly think that editing in a traditional wysiwyg editor will be easier based on this post’s mistaken complaints.
A useful reply to a post like that is to talk about how the tool can be used productively, so that people may feel it’s worth trying to edit latex source directly (since that works really well in practical cases, like my workflow which I wrote about).
You seem to have some strange internal standard that the only type of valid critical response has to break down the OP directly and isolate responses to specific points.
Not so. And in fact, this post isn’t good enough to warrant that. It’s a lot more efficient to just say, “nope, you’re wrong, here are real world workflows where, had you believed the advice of the post, you’d be in a much worse situation.”
Further, while even after self-reflecting on the tone of my post (i.e. the choice of “embarrassing” to hopefully connote how short-sighted the OP’s perspective really is), I feel confident that my post was fair, not overly harsh, and useful. Meanwhile, the tone of your post feels like you’ve just got an ax to grind for some reason and want to be negative.
You’re certainly entitled to that, I don’t mind. But your post definitely entrenches me further into the position that my post was pretty fair and that it’s reasonable to characterize the OP that way.
My last effort was a journal paper submission, and the journal in question supplied the (80MB!!!) template which was so crammed full of macros, custom styles and miscellaneous instructions that my AV program quarantined it. When I eventually managed to use it, I literally transferred the content from Notepad++ (my go-to for the act of actually writing - take note, OP) and images etc. from file and spent DAYS trying to get the format right. A hateful experience.
Word is great for casual use. For academic or professional use beyond the barest of basic typesetting, absolutely not.
I realize people may want to customize things later, but making the initial install as painless as possible seems highly desirable - esp. for students.
https://www.overleaf.com
Although things like [Skim][1] exist, Basictex does most of what I need, so I've never really tried wiring it into my existing install.
Pipe dream - If the people who work on macOS's Preview application were smart, they would steal it (or [Marked's][2]) ability to have a custom processor for displaying output. In that case, I could point it at my TeX install and just run my Makefile from there.
[1]: https://skim-app.sourceforge.io
[2]: http://marked2app.com/help/Custom_Processor.html
For the same reasons, I even switched to making all my presentations in latex even though I believe Google slides and Powerpoint are feature-wise unarguably superior.
This is dumb. One can argue that LaTeX is not a convenient markup language enough to write in. But stating that a markup language in itself is a bad thing for writing, while Markdown was invented in 2004 and wildly popular at the time this article was written (2016), means that the author himself is not completely rational when defending his preference, using XML and HTML as strawmen.
His argument is restricted to those looking to use Latex for questionable benefits that they do not fully understand.
I read this in 2017, and..early in 2018 spent months getting good at LaTeX/TikZ (using TexShop), and getting some idea about TeX. It's been amazing, well worth the time invested already. For the first time, I have organised notes on everything, including my programming. And it looks lovely, it's a pleasure to look over, makes me want to do more.
Ugh, just...no.
(de gustibus et coloribus non disputandum)
Even so, it does stumble on more complicated equations (the kind of equations that pure mathematicians write -- the symbol set in LaTeX is just much bigger, and being a typesetting language, it gives the user much more control over element positioning). Equation Editor also doesn't support more esoteric math objects like say Feynman diagrams, but in LaTeX there's usually a package for it. Also, aesthetically, Equation Editor output, though functional, looks a bit off compared to true typeset math. LaTeX-heads are sensitive to aesthetics that way.
In my time, those who outgrew Equation Editor moved to third-party software like MathType, which is much closer to LaTeX in functionality. I'm not sure if people still do that.
With references? The reference management in Word is atrocious and does not scale.
This pair of problems has driven me away from writing any but the simplest of mathematics in Word.
I've also found the symbol set in Word to be too small, but maybe with the universal adoption of Unicode this has been fixed.
I write primarily in Emacs and use markup. I don't try to convince anyone else to do the same. If MS Word (or TextEdit) work for you, great! I tend to write large documents for which I find those tools to be clumsy, but if your milage varies, no problem.
I also like being able to write my papers in vscode, with all the shortcuts and vim settings that I’ve gotten used to from coding. If Vim is second nature to you, it’s frustrating to have to switch over to ms word. Vscode has a good latex plugin.
That said, it did take some getting used to. Anything hard and potentially useful can become a “fetish” whereby we fool ourselves it’s the usefulness that motivates us but really it’s the hardness. I recall a time when I tried to use matplotlib to format all my plots, with labels and such. Horrible experience but I was really proud of these ugly graphs I’d make entirely from code. Then one day i just started saving them and opening them in illustrator for all the bells and whistles. Haven’t looked back.
This is often underestimated. It is possible to write complicated equations more quickly in LaTeX than in handwriting. I have done it.
However, to achieve this, there are provisos: 1) Macros already defined for commonly used structures; 2) Lots of repeated elements that can be copy-pasted; 3) Fast typer with a good text editor. 4) Keywords from amsmath, etc. committed to memory.
The only thing I've found missing from the Word equation editor is: 1) macros, and 2) the output just looks less nice than in PDF form. To be fair I've never exported but I still think Computer Modern has better glyphs.
* XITS https://github.com/alif-type/xits (port of STIX, kind of looks like Times New Romans which is not my preferred style but still better than Cambria Math) * https://github.com/firamath/firamath for Sans Math font
In generally I think Computer Modern is okay, I am not enthusiastic about it.
The cynical: While most conferences in CS advertise both a LaTeX and Word template for submissions, if you submit an article in Word, it seems quite likely it will be desk-rejected for being quackery, possibly without even being read. Is this fair? No. Has the scientific/mathematical community lost out on any great ideas due to this prejudice? Probably not. Think of this as the ultimate, final form of the clear plastic binder from Calvin and Hobbes, except Calvin was right all along and you probably can get away with saying that bats are bugs if you write it in LaTeX. This says more about the process of science than it does LaTeX.
The practical: I can think of three pragmatic reasons.
It seems that nowadays, good science is collaborative. Single author papers are rare. How is it that multi-author papers are written? Writing in a text-based markup language allows co-authors to easily send each other changes and keep a versioned history without having to worry about "track changes" and using file names as tags, and worrying about cross-compatability between ancient version of Office on Mac and Windows and that one person who only has Linux and so opens things in Libre.
In my experience, people do want to get into the weeds with formatting in LaTeX for many reasons, good and bad. For example, how can we shave off two lines of text so we're within the page limit (yes, even though the proceedings will all be published electronically we will get desk rejected for being one page over the page limit where the only thing on the last page are two lines from the bibliography, c'est la vie). That is a bad reason. But good reasons could include improving the look and feel of a document, or controlling where figures are placed, or drawing figures using TikZ. Additionally, having the sum total of the formatting in plain text has advantages. Do you think that undo really un-does everything in Word? Can't you remember some times where you do something, but then un-do and then do "the same thing" again and a different result occurs? LaTeX has many foibles but that is not one of them. Though, of course, sometimes I will wish you luck in understanding exactly why what you wrote in LaTeX works the way it does now. But for example, it is possible to define new mathematical symbols, operators, etc. in LaTeX in a way that is not possible in word processors.
The third pragmatic reason is that everyone else already uses it and some day or another your text will need to be turned into LaTeX for publication. I used to keep a lab notebook in Markdown and sometimes I would feel the desire to copy and paste some text from my lab notebook into a paper (if what I wrote was particularly lucid or well written, or was at least a good place to start from). This almost never ended well. If you have a draft of the paper in Word you might easily spend many days converting it from Word into LaTeX. If it's sufficiently complicated you might never get it converted. This could be a barrier to getting collaborators or submitting at all.
I think that everyone that works with LaTeX, knows, deep down, it is absurd, and the arguments the author runs into are not really arguments anyone would stand by if pressed. I think if anyone is going to write something entirely in prose, that they will never care about the versioned history of or want collaborators on, and don't care if the finished product looks good or bad, you can use a word processor, or at least, this is my criteria for whether or not I use a word processor. Am I drafting an outline of how we're going to write the proposal? That happens in Google docs. Am I actually writing the proposal? That happens in LaTeX, because I want control over the typesetting to fit in the page limit, I want it to look good, and I want collaborators to be able to edit the document and send me plaintext patches rather than docx with track changes maybe or maybe not enabled.
I've never had to submit source files for review. Maybe it's different in your field? In my field (robotics) we submit PDFs for review, and only submit source for publication.
Brilliant. Almost everyone in economics presents using LaTeX (Beamer) for exactly this reason. The outcome is that almost everyone's slides are composed primarily of tiny writing, and are therefore awful.
If you cannot stand latex but want to produce print-quality PDFs, Apache FOP (with Docbook) probably is your best bet. Otherwise your kind of bound to introduce a heavily manual processing step between writing and formatting for print.
Some of the Stackexchange sites make widespread use of MathJax for this purpose: https://math.meta.stackexchange.com/questions/5020/mathjax-b...
(Word does now store equations as XML internally, but it's pretty horrible - 5 kilobytes for the quadratic formula, for example)
Edit: I've just remembered that when I was at university in the late 90s, there was a small community of non-maths LaTeX users ... who were doing things like New Testament Greek. There may be other specialised use cases where having fine control over WYG is more important than WYSIWYG.
Now I've gone looking for it, I can find an example : http://jcsu.jesus.cam.ac.uk/~mma29/essays/dissertation/ - I think the intent of using LaTeX here was really for Bibtex, the bibliographic system, which is an extremely important part of all true academic papers.
ConTeXt is a substantially different beast to LaTeX. It's not LaTeX, although there are some LaTeX-inspired packages and tools to help migrate. A lot of ConTeXt is now written in Lua, and the versions in TeXlive are now sufficiently up-to-date to work from the manuals.
ConTeXt takes the view that it will start with sensible defaults and allow nearly all to be customized. Modules tend to play nicely together in an orthogonal way, but a lot of documents can be written without importing any modules.
The one thing I wouldn't say about ConTeXt is it's easier – you probably won't find the answers on stackoverflow like for LaTeX. It's a much more complicated system and you need the reference manual. But if you know how you want the page laid out and don't want to have to hack away at a style file, then it's a very powerful tool for serious typesetting.
That is true, but it is quite stable and it works well. The last released version is from 2015. The last LaTeX release was 2e in 1994. LaTeX 3 is apparently still being developed by a small number of maintainers. [0]
> (I think) the source repos were not available publicly
From the homepage: "There is no offical repository there the development is going. The source code is released irregularly and placed at the http://www.pragma-ade.com website."
There's a git mirror of the releases.[1]
> Also it is difficult to install
It's in TeXLive and works fine. There was a time of intensive development a few years ago, and TeXLive decided to include older, more stable versions.
> and hard to find documentation for.
The main documentation is the reference manual, linked to from the home page [2]. That said, it's a real shame there isn't more web-friendly documentation. The Wiki has some commands documented, but others are either missing or out of date.
[0] https://www.latex-project.org/latex3/
[1] https://github.com/contextgarden/context-mirror
[2] https://wiki.contextgarden.net/Main_Page
What is different about ConTeXt is the prevailing attitude towards using plain TeX from it. LaTeX really seems to consider plain TeX quite unsafe. I suspect this is because it is hard to build a resilient declarative system on highly weird and procedural TeX. ConTeXt, on the other hand, kind of encourages you to use TeX directly. So that led me to learning more about plain TeX, which now seems much less scary to me than when I was using LaTeX.
Agreed. I think what I was trying to say is that LaTeX will have more answers on Stack Overflow. However, once you've found the command (and you can get an awful long way with the ten listed at the beginning of the reference manual) then the options are well documented.
> What is different about ConTeXt is the prevailing attitude towards using plain TeX from it. LaTeX really seems to consider plain TeX quite unsafe.
This is very interesting. I've found the same difference but in the opposite way. I've had to resort to using plain TeX in LaTeX quite a bit, usually buried in an environment or a command, to achieve what I want. And yes, it can be fragile, but sometimes it is the only way without using KOMA-Script.
I've found that ConTeXt has usually already considered what I want to do as a use case, because its scope is quite a bit more broad than LaTeX.
I wish I had sources for my other comments handy. If I have time I will try to dig some up.
http://wiki.c2.com/?BlubParadox
(I am joking of course; even though I would not say that LaTeX is the 'best' solution for all use cases, I can at least state with confidence that it satisfies some properties that traditional word processing tools do not have.)
Maybe I this is my misunderstanding, is something the wrong with 'composability' of constructs in the LaTeX language? A concrete examples eludes me presently, it is like the latex functions/forms cannot be nested arbitrarily like normal programming language blocks/constructs.
My overall sense is that the underlying reason for this is that supporting math in typesetting is an extremely difficult problem, much more so than people realize. Most typefaces don't come anywhere near being adequate for math, and those that are functionally ok are often really not very good aesthetically speaking. The equation typesetting can also be really difficult, in terms of "under the hood" implementation issues as well as the specification of standards. etc. etc. etc.
As a result, it's not so much LaTeX is good, as much as it is that it's been around for so long, and the domain is so much of a slog, that it's hard to develop something better.
I think a lot of people have a limited vision of what things could be like, and tend to base impressions on limited experiences with alternatives. The proprietary nature of Word does create problems in sharing written documents sometimes; the "black box" nature of WYSIWYG programs in general can be infuriating when you need to be careful with typesetting equations. On the other hand, constantly having every typesetting detail in your face with most LaTeX implementations can be intrusive and leads to "forest for tree" problems sometimes.
There are lots of beautiful texts typeset in LaTeX. But there's also a lot of LaTeX documents out there containing correctly typeset equations that are nonetheless really unpleasant to read or look at, for visual design reasons.
> Seriously, anyone who believes that making people type this…
Just an introductory argument against it, not going into that. I'm looking for meat. Hmm... meat :)
> I know a number of academic authors who seem to spend considerable amounts of time doing that. I shan’t say that this is worse, but is it really better?
I did that too, but just because I wanted to be geeky. Was it useful? No. Would I have done this in a word processor? No. But the fact that I get to enjoy LaTeX by being geeky gives me a good feeling to return to it and write in it. Sounds like I have a fetish? I do! And not only one ;-)
> When it comes to stopping people from creating documents in purple 28pt Comic Sans, teaching them all to use LaTeX is a lot less efficient than stating that you will refuse to read anything that doesn’t match the style guide.
But LaTeX does have better alignment and justification algorithms. You won't see the difference, unless you're starting to compare it with a word processor [1].
> LaTeX does less to prevent authors from getting on with writing documents than TeX does. But if neither of the two existed, and you had to come up with something, right now, in 2016 – would it really be a markup language?
It depends, I see that people use Markdown for these reasons. And for my resume writing this is definitely true. The template also looks quite nice and I wouldn't know how to design a beautiful resume.
> we have LaTeX evangelism and the false implication that word processors don’t facilitate structured writing at all.
I wrote big documents in Word, Google Docs and LaTeX. I prefer LaTeX (my Gdocs fetish is with short documents or any psychology paper I'd need to write collaboratively).
> In sum, the case for writing in LaTeX is more than a little weak.
My reasons:
1. Automatic reference list generation (never saw psych. students use it in Word). I still don't know APA but I never got minus points for my bibliography.
2. The ability to comment. This has been huge for me (as in 10% to 20% better grades in writing huge) Yes, you can do this in Word and yes it also distracts me, when I do this on Overleaf it doesn't distract me. If I want to read, I read the PDF, if I want to write or edit and get meta information, I type.
3. Super quick restructuring thanks to include. There have been many times where I needed to quickly move +30 pages to a completely different location. It just takes one line.
4. Crash resistant: one time there was this psychology student who had her Word document crashed and unsaved. I had to use the command line to copy some temporary saved file that Word didn't recognize (apparently...) and tweak it in order to restore it. It took me an hour. LaTeX is a text file.
5. Tikz: I needed to create a graph and wanted to do it programmatically. Tikz is love, Tikz is life.
6. Knitr: I wanted to be ultra precise and focus on purely reproducable research, with Knitr you can put R formulas into LaTeX and calculate stuff. I handed in my source and PDF, my psychology teacher thought I was a wizard.
7. Different thinking cycle: by hitting compile and waiting for a bit you get some idle time to mind wander. For me it takes the pressure off and sometimes makes me anxious to see the result.
8. Fetish effect: you start to care about typography and the like (see [1]). Sometimes it isn't the programming language itself but the community behind it that makes it so powerful (Hello JavaScript! NodeOS is wonderfully hilarious! ;-) ).
9. Wizard effect: psychology students would leave me alone because they had no clue what I was doing. It also made me look smart :P Is this an argument? Hmm... I have a fetish, you pick out the good ones ;-)
10. Placing images neatly immediately: oh wait, nope this is a downside. Unless you...
> To cut to the chase, LaTeX documents are very hard to read until typeset,
Not my experience on Overleaf. I compile early and often just like in programming since it is a programming language.
> LaTeX is, as already noted, a markup language.
It is also Turing Complete and I have used it to do bibliographic database lookups, replacement and insertions. I used to program a bibliography per medium item and it would automatically figure out what medium item it was (e.g. journal, online blog post, online video, book, etc.).
> disruptive of the text for human readers (including editors and the original author): looking at the screenshot, we see the text mixed up with a lot of symbols that are not part of the text
I'd recommend you to use something like Overleaf and see if that's still the case (disclaimer: not affiliated with them, I actually used ShareLatex before that, I only went to them because I had to since they merged -- ShareLatex is open source by the way, so no internet connection needed).
> By the way, one of the words in the screenshot is not right. It’s a typing error that I deliberately inserted. I know where it is because I put it there, but looking for it is hurting my eyes.
Again, use something like Overleaf. They have spell-check.
> But that isn’t how most people want to write.
Most people also don't want to program. I won't recommend LaTeX to my girlfriend either, but to fellow programmers, why not? Maybe they like the program/debug cycle more.
> Never mind boilerplate code like \documentclass{article} or \begin{document} at the start of your document – running into something along the lines of \parencite[706]{lena_peterson_2008} in the middle of a paragraph and having to mentally parse it into ‘(Lena and Peterson 2008, p. 706)’ interrupts your train of thought and makes it harder to do what you really need to be doing: reading your punctuated words back to yourself to make sure that they ‘sound’ like what you really wanted to say and don’t have any mistakes in them.
I disagree. I never wanted to read my citations full out and I'm happy I can read them by aliases.
> This is what LaTeX is good for: not helping people to compose text, but helping them to make it look nice. If that is important to you, go ahead and give it a look.
LaTeX is good for long pieces of text. I never need to scroll millions of pages since everything is in a folder structure that I can immediately click on. Making things look nice is very important and even a slight competitive advantage, it makes you look more credible.
> I think you’ll probably agree that the LaTeX version looks better than the word processor export version. Whether it looks sufficiently better to justify the additional effort is a judgement call that you’ll have to make for yourself.
I think that students can invest the time. It took me about 40 to 80 hours to get a good flow with LaTeX. After that I didn't have any issues anymore. Also, whenever I got issues it was a fun procrastination activity (students like that from time to time).
I suspect the author never really wrote anything reasonably complicated.
How about:
1) can't have proper a version control with those binary documents word processors generate.
2) no easy way to split huge papers/thesis/book into manageble pieces
3) conversion between formats is a nightmare, with formatting falling apart all the time.
4) typing formulas in Word (and Libre/OpenOffice) is a torture.
5) diffs
6) ...
that's just off the top of my head.
7) collaboration is almost impossible
Having a nice diff-based change history and being able to ignore formatting issues when writing is a blessing.
And even worse (maybe): no way of fixing errors later on; arguably, the error messages of LaTeX are obscure for new users, but at least you can always reset your document to a working state.
The graphical representation shown on screen is somewhat approximate e.g. fonts and sizes are approximate, and line wrapping is totally unrelated to the final output. But it's good enough to understand what's what, and most importantly it's free of all the markup clutter, and that's what the author objects to. It's especially good at this for equations, but it makes a different even just for regular text. Plus it has autocomplete for LaTeX commands (I realise that there are LaTeX editors that do this too).
A lot of people mistake LyX as being a beginner's tool to avoid understanding LaTeX. I would say it is almost exactly the opposite: you still need to understand LaTeX to use LyX effectively. Rather than the markup being right in front of you it's hidden by a layer of indirection, which of course is the whole point, but it makes it harder to understand what's going on when something breaks. It breaks less often (e.g. you never get mismatched close braces) but the hard errors still crop up (e.g. incompatible packages giving really obscure errors about macro expansions).
I've written numerous mathematical documents of various lengths in LyX, and prior to that a few documents in LaTeX including my MSc thesis. When I first discovered LyX I was partway through writing up notes from a lecture course I was attending. In LaTeX it took me about three hours to write up notes for each hour lecture; in LyX it took me one hour per hour of lecture. A factor of three speed improvement! I can hardly imagine how long my PhD thesis would've taken without it.
[1] http://www.lyx.org/
In fact LyX also supports instant previews of equations. That might sound a bit odd since it already displays equations graphically, but it's nice to see a pixel-perfect display of them. It also supports wrapping instant preview of arbitrary bits of the documents, which is really handy when you have inserted explicit LaTeX e.g. a Tikz diagram.
My domain is pretty specific - typesetting chemistry papers, but for that it is absolutely invaluable. There's no tool on the market which can do it with anywhere approaching the ease. Even MSWord, which is now good enough that I can type simple maths about as fast as I can in Latex (although frankly even there I prefer Latex to futzing with a gui), doesn't have nearly the tools I need.
I'll bet this is true in all sorts of smaller domains, too. Typeset anything unusual enough not to have good specialist tools and the freedom of Latex typesetting will make up for the learning curve in weeks if not days.
Creative writers for instance don't really care about typesetting; they want to keep their manuscript organized and to be able to change it easily. Here's en example[1] of a writer showing his personal workflow for writing (in this case using emacs) and you'll see what I mean
[1] https://www.youtube.com/watch?v=FtieBc3KptU
Second, he specifically does encourage using latex to typeset. Just not to type in the first place! Use Markdown, or a word processor, says the author, and then latex do what it is good at: typeset. But don't write long paragraphs in latex with attention-breaking code scattered around the text, since that places obstacles in the path of editing.
But I don't know how you expect me to write a chemistry paper without writing the equations? Just leave them blank? Typesetting is instrumental to simply composing the paper, it's not something I can do after. As I said in my initial comment, nothing else will actually do the task as fast as Latex, so it'd be pretty perverse to type in something else. Maybe it works for your domain, but not mine.
Nevertheless, it still doesn't have a bunch of the symbols I need, which was my point about smaller domains.
I'm glad it works for you, but given the advantage seems to me marginal and it offers no options for extensibility, I'm still unconvinced that we're at or even close to a stage where Latex has no place in writing documents.
The main reason i write stuff in either plain text or LaTex is, that i do not want to spend even a second on figuring out the height oft a heading.
Especially if you are 2 pages in and have to remember random numbers for text size in listings or whatever.
Every document looks like crap when i try to do it manually and on top it wastes my time fidgeting with widths and heights.
I leave the design to people who now what they are doing (not me) and have the computer do the tedious formatting stuff.
Also you can put those Tex docs in source control.
And if you don't like how the document looks, choose a different theme and recompile.
Setting up a LaTeX environment in Windows can be difficult if you do not want to install one of the very big suites (although being honest, who is worried about 2GB when you need more space for something like printer "drivers"?). However, in most Linux distributions, you just choose LaTeX in your package manager and it works. And this is the case for most free and open source software. I do not know about Macs.
One can add just the packages you need and it can be installed in user directories for people who do not have admin privleges.