I have been reading this article and started to think about the position of LaTeX amongst the current competitors. Currently, I am using it (or XeTeX) even for ‘casual writing’ and sometimes I spend free time by exploring its possibilities and various solutions based on specific packages. When I have to write an article or a report, I consider it to be quite easy (well, in a geeky Open Source way) to use and as I am not an expert on typography (rather an advanced self-learner), the microtypography features via microtype appear to be professional enough. However the article mentioned hereinbefore states otherwise.

Therefore I would like to ask you, the experts, what is your opinion on these points (I am quoting from the same article):

  • Typography rules are not supported well.
  • Microtypography ("all lines on fixed page positions") is not professional.
  • Hyphenation is bad and hard to use.

In short, Latex misses two things: (1) Ease of use; and (2) Quality of typesetting is not as good as it could be.

Does the LaTeX3 project aim to solve any of it?

  • 4
    I'm not sure whether such a wide question fits the format of the site. Asking for how LaTeX handles one of the items in the list per question would be more appropriate.
    – N.N.
    Commented Dec 6, 2011 at 19:42
  • 4
    I'd disagree with several of these claims, in particularly font usage (with Lua/XeTeX), indexing (why is it the worst of all worlds? I can't think of a simpler way) and hyphenation.
    – Caramdir
    Commented Dec 6, 2011 at 19:44
  • 3
    I'm with the other comments here in that this is not one question but lots. Some have been covered here before, some are discursive, and some may have clear answers.
    – Joseph Wright
    Commented Dec 6, 2011 at 19:55
  • 3
    I think the author is in the hate phase of the love/hate oscillation which many of us consistently go through, considering the bulk of material s/he had to write up for that massive ebook.
    – percusse
    Commented Dec 6, 2011 at 19:57
  • 7
    I refuse to take anyone serious who thinks that (La)TeX "Hyphenation is bad and hard to use." He deserves Word. BTW: Why is he using Latex for typesetting? :) Commented Dec 6, 2011 at 22:04

5 Answers 5


I think the author having written a very nice book in LaTeX, is entitled to his views and has some valid points. I agree with some of them but disagree with most of them.

He has valid points on ease of use and the lack of a GUI. However, he misses the point that to incorporate the 100s if not thousands of commands available to a user via the basic TeX engine, LaTeX and the few thousand available packages plus an author's specific macros will provide the most complicated and frustrating GUI possible. It will also not be quicker to find these commands. A GUI works well if everything is available two clicks away, anything deeper and you will battle to find it.

Spell checking is provided for most users via the editor. Grammar checking never worked in Word and is a frustrating experience to have to deal with it.

Figure placement has never been a serious issue for me, neither getting them at the top 20% of the page (see for example https://tex.stackexchange.com/a/35162/963).

I agree with the author that fonts and tables are difficult to use. I don't see an easy way out on tables, but font management could certainly be improved.

I disagree on his comments for hyphenation and index production. I actually view these as part of the strengths of LaTeX. I also disagree on his views of firstly the importance of micro-typography and secondly I have doubts in line with Joseph's comment if he actually was referring to grids or micro-typography as most of us understand it.

He has valid points that LaTeX out of the box does not follow typographical trends. The problem with typography is that it is a very broad term and like fashion it changes. Where typographical rules could be deduced Knuth did incorporate them. He also set the standards for mathematical typography.

If the author can define the typography rules, I am sure that he will end up with a number of styles. How do you define typography for a glossy magazine, a photography book or a dictionary? If these rules can be deduced, they can be programmed fairly easily.

So is there no room for improvement? I think his idea of funding is good. I think from a user point of view multi-column layouts are not what they should be. There is huge room for improvement here. Font management can be improved and perhaps simplify the user interface. The LaTeX3 project needs to give serious thought to a syntax that I am sure will not go well with programmers. A sugary super-set of all commands will be necessary. Thought should be given by the community in developing more templates and classes rather than packages.

Having said all that I need to add, that I have seen and used all the word processing packages from HP specific packages, wordperfect, all versions of word etc. and I have been with (La)TeX since my student days and that is a long time ago. You sacrifice a few things, but (La)TeX also gives you stability. You also don't throw away your knowledge but rather add to it with the years. The only program that now competes with LaTeX is Adobe InDesign. If you are prepared to write your pages using LaTeX, page by page (similarly to a graphic designer) or the way web pages are developed you can achieve the same if not better results.

  • 6
    I completely agree with your opinions. Commented Dec 6, 2011 at 21:24
  • 1
    This is what I was hoping for—a few qualified opinions from a respectable LaTeX user. Thank you. Commented Dec 6, 2011 at 23:02
  • 1
    I found the two following points in the linked blog rather interesting: first, the author complains that thare seems to be no new development on the pdftex sarovar site, only to suggest few lines below to put luatex on second priority. I think it shows some confusion, which is however understandable. The pdftex site never mentions that luatex is basically the successor of pdftex, and that most of new development happens there. Commented Dec 7, 2011 at 17:56
  • 1
    The second point I found interesting was the dismissal of ConTeXt. I found it somewhat hilarious that after suggesting that LaTeX be managed by a company, and that more attention is paid to users, including professional typographers, s/he then goes on and dismisses ConTeXt as "programmer driven", and suggests that it is given 2nd priority. Again, it is somewhat typical, though. While awareness of ConTeXt increased lately among TeX users, and most LaTeX users will know that it exists, few will know what ConTeXt is actually about. Commented Dec 7, 2011 at 18:03
  • I share your disagreement on hyphenation and indices. I also quite disagree on typography. While I trust professional software like InDesign to produce better typography in a special case: glossy magazine, album booklet, etc., I know no means to produce a better typography of a book while keeping it editable. Of course, you could write in anything and then pass the text to a professional designer for manual tweaking. But short of this, LaTeX and microtype do the best I know. Commented Oct 14, 2017 at 13:26

I think the only valid claim is the one about the lines not being in register. This is something you want to achieve if you print on both sides of a sheet of paper, e.g. books. The reason is that when you print on paper that has a low opacity you can see the shadow of the line printed on the other side through the paper. If the lines are in register this shadow is in the same area as the line you are currently reading, this supports a smooth grey value of the line compared to the shadow appearing in the interline lead, this again supports readability.

Due to the way TeX shrinks and stretches the glue between the boxes the lines on every page are in different positions. Matters get even more complicated if you have different font sizes, e.g. subsection titles. Typesetting the main text in register would force you to skip a line in the register as it might bring the text too close to the subsection title.

  • 8
    Grid typesetting can be done: ConTeXt does it, there are packages for LaTeX2e that work and we have some experimental code for a new output routine for LaTeX3 which also covers this.
    – Joseph Wright
    Commented Dec 6, 2011 at 20:34
  • 1
    @uli I have no idea, never used it. By the way, I just remembered that Gonzalo Medina made a package to enhance the functionality of grid. See tex.stackexchange.com/a/35093/586 Commented Dec 6, 2011 at 20:52
  • 3
    @uli Some simple mathematics can prove that the same problem exist for any software. If you have headings or images that are not integer multiples of baseline you will have the same problem. This is almost impossible to get right when you have inline maths. What graphic designers do is divide the page into vertical and horizontal grid and try and fit everything as best as they can, but in many cases they just trick your eye, or they re-size images. Most of them are now moving away from grids just google breaking away with grids or similar.
    – yannisl
    Commented Dec 6, 2011 at 21:36
  • 1
    @YiannisLazarides Sure, if I go through my bookshelf and pick a random fiction book, say a Terry Pratchett, and hold a page against the light, then the lines match up on both sides of the page. If I take however a random non-fiction book I'm hard pressed to find one that has matching lines. It’s just me being curious, if a LaTeX solution exists for the simple single column no-math-at-all fiction book case. ConTeXt is the only one I know off by now, and that’s not strictly a LaTeX solution.
    – uli
    Commented Dec 6, 2011 at 21:47
  • 4
    Regarding grid typesetting, the details manual is an example of a ConTeXt document show grid typesetting with headings, floats, and math.
    – Aditya
    Commented Dec 7, 2011 at 2:30

I think the author is partially right when he talks about mismanagement although I wouldn't use that term. The problem is more that there is no project management at all, there are just individual programmers.

The problem already starts with such basic questions as "what actually is (La)TeX"? There are engines, there are distributions, there are packages and individual tools but they are all taken care of by different people. Some areas of LaTeX are very well taken care of: Packages like koma-script, fontspec or biblatex have extremely productive and responsive maintainers. But just think of what would happen if one of those decided to stop developing for LaTeX? Since this has happened before, many packages and tools are not in good shape. bigfoot, for example, is potentially an extremely useful package – there has never really been a full-working release. A long known problem between xindy and hyperref has never been fixed, polyglossia hasn't been updated for quite some time since it original programmer has stopped working on it (fortunately, the situation has improved here, there now is a new maintainer). polyglossia desperately needs to be adapted for luatex, but no one really seems to have time to work on that (LuaLaTeX itself is a problem: While the work on ConTeXt progresses nicely AFAICS, the respective LaTeX packages have a hard time catching up).

The problem is to some degree probably typical for open source: the individual programmers mainly work on things they need or are interested in themselves. This is completely understandable, but it leads to a quite chaotic situation – especially in the TeX world since there are so many different elements. Depending on your needs you might suffer badly from this situation or hardly notice it at all.

To be honest, I don't know how this situation could be improved: Funding and professional project management would surely be a great thing – I just don't see how it could happen.

  • The Taoist Zhuangzi said, "Good order results spontaneously when things are let alone". Funding yes and minimum steering of the projects. A lot of Knuth's work was via his PhD students, Liang for example worked on hyphenation etc.
    – yannisl
    Commented Dec 7, 2011 at 13:39

I'm make my living on writing (being a lawyer) and use hardly anything but LaTeX. That said, I miss some features:

  • Better support for twocolumn texts, e.g. twocolumn footnotes.
  • Longtable with a function to add all values of a certain column and (!) to calculate a subtotal at the bottom of a page and naturally a carry forward at the beginning of the next page. This should be possible with LuaTeX.
  • margin notes which can go across a pagebreak.

But well, that's it! Folks, either my requirements are so humble or LaTeX simply is great. I'd say the latter, because Word hates me.

  • 2
    The first one is easily satisfied by ConTeXt. The second one should be too, as you noticed. As for the third one, you'd have to ask ConTeXt wizards, but I wouldn't be much surprised if it were possible.
    – mbork
    Commented Dec 7, 2011 at 17:25
  • In any office where their a lot of typically similar documents, reports etc., LaTeX wins hands down. Your modest request for subtotals could easily be answered on this site. It is not as hard as you imagine (but it will need to runs). Out of curiosity, do you produce your letters also with LaTeX? The reason I am asking is that we produce most of our documents with LaTeX, except letters (I am a Projects Director, in a large construction Company.)
    – yannisl
    Commented Dec 7, 2011 at 21:43
  • @YiannisLazarides I'm producing all letters using LaTeX, class scrlttr2 (KOMA-script bundle).It takes up to some hours to code the head of the letter, but with some experience in LaTeX it is not really hard, and even if you hit a difficulty, you just can ask at www.komascript.de So go ahead.
    – Keks Dose
    Commented Dec 8, 2011 at 8:08

I actually disagree that latex's ease of use is poor - I have worked with professional secretaries (as in, decades of experience, have completed both secretarial college and training periods), and seen one reduced to tears by Word's flaky styling and numbering features. The kind of behaviour that they expect is the kind of behaviour that LaTeX provides.

So, my answer is, with XeTeX: minimal gui support for configuration, and selecting font styles, plus training. LaTeX (and friends) does not have documentation that is widely available, and targeted at absolute beginners.

  • In principle I agree with you: Word is much hard to use than Word – the problem is that most people don't perceive it this way. And we wont change this perception by insisting that LaTeX is actually easy to use.
    – Simifilm
    Commented Dec 7, 2011 at 11:55
  • I also think LaTeX's ease of use could be improved in one area: If someone starts with LaTeX he's confronted with a many different packages which often do similar things; even worse: there is a lot of outdated documentation around. Some kind of "core" consisting of essential up-to_date packages with proper documentation could provide a better user experience.
    – Simifilm
    Commented Dec 7, 2011 at 12:01
  • @Simifilm: That is why I suggested that it needs entry-level training materials.
    – Marcin
    Commented Dec 7, 2011 at 12:18

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .