The topmost answer in the very popular question What are good learning resources for a LaTeX beginner? recommends the LaTeX Wikibook. I found it helpful as well when writing my thesis, because it provides hands-on practical advice (and because it is always among the first hits on a web search about LaTeX).

However, in the comments it is implied that there are errors and/or bad advice in the Wikibook. It is pointed out that the commenters could easily edit the book, but the discussion ends there. A beginner reading this could be detracted from the Wikibook by these comments (I guess that's their purpose after all) even after these potential flaws were fixed (if there are any).

To underline this, the Wikibooks guidelines tell you to be bold when editing books. My understanding is that replacing Bad Advice™ with better advice in a book would be desirable, especially because the book are meant to be instructional resources.

Thus I'd like to pick up the thread started in the comments of the above-mentioned answer. Since we (as a community) recommend the Wikibook (it is the topmost answer in the "Resources for beginners" question, after all), and the possibility to improve it exists, we should strive to improve the book wherever we see potential for improvement, for the benefit of LaTeX beginners.

Personally, I don't feel knowledgeable enough to discern bad LaTeX advice, but there are people here who can easily tell what's A Bad Idea™ from years of experience. And if those people among us pointed out flaws in the book, we could discuss and eventually fix them.

However, I'm not sure how one would organize such a quest:

  • A community wiki question on the meta site or on the main site, where people can point out flaws which they found, which can then be discussed in the comments and subsequently fixed?
  • Some sort of chat event, like the "Answer the Unanswered" events? (disclaimer: I've never taken part in one).
  • Determine that this is not our job and "the Wikibook community" should fix it? (wait for the problems to "fix themselves", so to say)?
  • Determine that, actually, there are no flaws in the book, and this post is superfluous?
  • 9
    this post is not superfluous. perhaps i'm inept, but an attempt some time ago to correct a particularly egregious error failed in that i was unable to get registered so that the attempt could even be recognized. someday, when i have time to burn, i will try again. i think i'm qualified -- 35 years as editor of tugboat. or maybe i will just read and comment on sections, and turn the corrections over to someone who is permitted to make them. in the latter case, it would be helpful if people who are so permitted would identify themselves. Commented Nov 11, 2015 at 14:13
  • Based on my Git experience, a wiki like learning is not a good idea for learning TeX. LaTeX Wiki is convenient but hardly ever useful for slighltly sophisticated action.
    – percusse
    Commented Nov 11, 2015 at 18:20
  • 1
    @barbarabeeton As far as I understand it, anyone can edit any page there, without registration. Just click the edit button at the top, or the one above a specific paragraph. However, unregistered changes must be reviewed (similar to here, I guess) before they are visible.
    – Fritz
    Commented Nov 11, 2015 at 19:02
  • 5
    @Fritz -- that's also what i understood, but after more than an hour spent making fixes, what i proposed wasn't even posted for review. i don't have any problems with review; if something i submit is reject in review, and a cogent reason is given, i will accept it. i do have a problem with inability to get something accepted into the review queue. but as i said already, i'll try again, when i have time to burn. Commented Nov 11, 2015 at 19:19
  • I also think this sounds like a good idea: even if we can't implement all of the suggested edits, it would still be very useful to have a list of errors or addenda for each page on Wikibooks LaTeX, as a reference, and it sounds like a CW post would be particularly good for that. Commented Nov 11, 2015 at 19:54
  • 2
    Yes, the wikibok is bad. It was discussed the least days in chat and some changes have already been made. But TeX.SX is not the place to really discuss wikibook-matters. Each page on the wikibook has a discussion page where (posible) changes are discussed.
    – Johannes_B
    Commented Nov 12, 2015 at 8:30
  • What we need is a bigger awareness (this post is a start for this community) and get people over there. There is a huge amount of work to be done to make the wikibook acceptable and this needs quite a bit of man power.
    – Johannes_B
    Commented Nov 12, 2015 at 8:31
  • 2
    Let's assume the english wikibook was the first, all others mainly translated that. Right now i counted 19 different languages the wikibook is available in. Keeping the changes in mind, there is not one wikibook that needs fixing, but 19 different wikibooks. The russian or vietnamese versions might even be a good source of information, i don't know, i don't know those languages. You see, this is not a simple let's have a session and fix it kind of deal, we need expoerts from all over the world. Or, do we if every wikibook is standalone?
    – Johannes_B
    Commented Nov 12, 2015 at 9:19
  • I have nothing useful to say, except that I agree that wikibook needs more attention. One example from personal experience. The book had this neat table of font sizes, except that sizes were all wrong. Eventually, I edited the book to say that they wrong, and should not be trusted. After another long while, I got around to fixing the table, which is now correct. But it was wrong for at least several years. I have no clue where the bogus numbers came from.
    – Mark
    Commented Nov 12, 2015 at 22:27

4 Answers 4


The Wikibook only will improve through people with a good LaTeX knowledge actively helping and working on the wikibook. Everybody can join and make changes to the pages there. However, instead of just registering and working on the pages or even just working on the pages without registering it is better to first to try communicate with the people there.

Each wikibook page has an associated discussion page where one can join the current discussion, make suggestions what to improve (and maybe also why).

I registered a few days ago and started working on the German wikibook (and the associated dictionary) after I saw what they presented for chemical typesetting. I started on improving that section but I also corrected errors elsewhere, joined discussions on other topics, made suggestions, waited for feedback and finally applied my changes. My impression was that the people are very happy about new people willing to improve things and are glad to help with any question you might have. So far the experience has been quite fun! :)

If enough of us do the same then in the end the wikibook will improve even if it may be a slow process. (I'll stick with the German version for the time being, though…)

  • Greetings Clemens! Nice to see you not only on Wikibooks but on stackexchange as well ;-) Commented Nov 15, 2015 at 18:45
  • 2
    @tampis and on golatex.de, on latex-community.org, on texwelt.de, … :)
    – cgnieder
    Commented Nov 16, 2015 at 12:01

While it would obviously be wonderful if the wikibook was improved, it seems to me that the real problem here is that it is the top-voted resource. Even if the wikibook is improved this second, we have (collectively) been directing people to a poor resource for however long. That's a problem here.

Note that I'm aware that it is not necessarily a TeX SE problem specifically. It may be that the SE voting mechanism is not, in general, very good at selecting better answers. Or that it is not very good at selecting better answers to certain kinds of questions. I'm not saying this is so - perhaps people just vote badly here or perhaps there's some weird interaction between the SE voting system and TeX users. But when I say it is a problem here, I just mean that it is a TeX SE or SE or... problem, as opposed to a wikibook problem. It is 'our' problem in a way that the state of the wikibook itself is not.

Again, I'm not saying it wouldn't be great if it were improved. It would also be great if all those templates were updated, corrected and made sane. Not just good in general. But good for this site. Even so, those are not 'our' problem in the sense that a highly-voted recommendation to use those templates would be.

  • This answer has got 8 upvotes, yet, the troublesome WikiBook answer has got only two downvotes. I understand that people don't like the minus ones in this community, but I appeal to them to reconsider in this case.
    – yo'
    Commented Nov 13, 2015 at 22:52
  • I just checked, and there are now 5 downvotes (including mine) on the posting in question.
    – Mico
    Commented Nov 14, 2015 at 17:18
  • 1
    @Mico I can't down-vote twice. The system won't let me! Maybe we could persuade Paulo to down-vote for the first time?
    – cfr
    Commented Nov 15, 2015 at 0:28
  • 3
    If there are multiple answers to a question, we should upvote those we think are the best answers, not downvote the others.
    – Sverre
    Commented Nov 15, 2015 at 14:33
  • @Sverre My point was that this answer had more votes than other answers. I wasn't making any claim about how a total voting score, relative to those for other answers, should be constituted. However, upvoting better answers is not incompatible with downvoting poor ones. So if the community collectively did either or both in sufficient numbers, this answer would not be the 'best' answer to the question.
    – cfr
    Commented Nov 15, 2015 at 15:05
  • @yo' Not sure if you saw Sverre's comment.
    – cfr
    Commented Nov 15, 2015 at 15:05
  • 1
    @Sverre If there is an answer that I believe is wrong, I do downvote it. Of course, I first comment on it; I did so here a long time ago, but it is still the most popular answer. So why shouldn't I downvote it?
    – yo'
    Commented Nov 15, 2015 at 18:39
  • @yo' Because the answer isn't wrong. The OP is asking for an opinion, and the person gave the wiki book as his opinion. It's under no circumstances a "wrong" answer. A wrong answer is when the OP is asking "how can I do X", and then the answer provides a suggestion which doesn't produce X. So I'll reiterate what I said above. IMO, we shouldn't downvote some answers only because we think other answers are better. Do we ever do that otherwise for "how do I do X in TeX?" questions? Also then, one answer is better than the others. We indicate that by giving it more upvotes (and accepting it).
    – Sverre
    Commented Nov 16, 2015 at 14:58
  • @Sverre We obviously have a different opinion on what is wrong and what is not. Any further discussion seems pointless.
    – yo'
    Commented Nov 16, 2015 at 15:16

A wiki is essentially a reference source. As being such it is a very bad place to be a reference material for the beginners and intermediate users of LaTeX. Surely you can place some information about how to use \date or \author but the complexity increases substantially if you just want to push that title down a bit and make a line break in the institution name.

To do this simply, people do a lot of ugly-as-hell hacks which rub people here the wrong way. Simply because we know better. But it just works why should anyone bother about what we think?

The biggest obstacle in TeX to convey is that there are gazillion-many ways to do the same thing since only the expanded tokens matter. How it got to that point is not relevant and that is the major obstacle in making people learn TeX. Coincidentally this is the same reason why old and wrong habits don't die. Because they kind of work for some years and break down out of the blue sky.

Unlike other programming languages the inplace replacement nature of TeX makes it almost impossible to treat the subject in a Wiki fashion. Also it serves as a horrible source of confusion since internet supplies an authoritative tolerance to anything that bears the prefix wiki.

Instead of wasting the time and effort of the community, if there will be a collaborative effort, I would put my money on an intermediate level TeX treatment of an online book (Distilled TeXbook if you like). That would for example, eliminate lots and lots of discussions, edit wars and so on that you see even on math pages.

It can be much better with a Read The Docs type of format such that the wikibookness can be kept (if that is absolutely necessary) but the flow would be much better with a GitHub-powered pull-merge intake and we can have indices, table of contents and much more. Even better, a LaTeX3 solution can be incorporated to all this at any point. A LaTeX3 based introduction can finally make a clear distinction between the document level and programming level formats. So much so that new generation can give up on the old style and start from scratch without the need of learning every LaTeX version before they can perform the simplest hack.

I have the lesser-likened opinion that this much of backwards compatibility is not useful for any purpose other than bragging about it. Things have changed and so have people. The engines are still around and if you are dying to replicate the 30-years old document you still can. That doesn't mean we have to get stuck with a system because somebody wants to compile a 30-years old document.

Irrelevant opinion: In fact, if TeX can't get out of this PDF-output lock, it will be one of those archaic things because interactivity is killing everything on its way.

  • i got confused mid-answer and couldn't follow anymore. Would you mind expanding your thought process a little bit?
    – Johannes_B
    Commented Nov 13, 2015 at 19:10
  • @Johannes_B which part?
    – percusse
    Commented Nov 13, 2015 at 19:18
  • starting with the collaborative effort.
    – Johannes_B
    Commented Nov 13, 2015 at 20:00
  • @Johannes_B Is it better now?
    – percusse
    Commented Nov 13, 2015 at 20:08
  • 2
    Much better. Sadly, i agree with your point and think that (improving) the wikibook is kind of a waste of time. :-(
    – Johannes_B
    Commented Nov 13, 2015 at 20:16
  • I don't get the point I'm afraid: do you say we shouldn't try to improve the wikibook because there are better formats/ways for learning LaTeX? I believe it's true that there are much better ways of learning LaTeX but that doesn't make the wikibook go away which bothers me, especially with your observation »internet supplies an authoritative tolerance to anything that bears the prefix wiki« in mind…
    – cgnieder
    Commented Nov 13, 2015 at 21:20
  • 1
    @clemens Yes, I think if this is going to get traction by the community here, there are much more efficient ways to utilize that manpower instead of fixing something that is almost doomed to fail.
    – percusse
    Commented Nov 13, 2015 at 22:40
  • But now it is out there, just like the wonderful Thesis template from Sunil Patel. Both are heavily in use but in my opinion, don't deserve to be. We gotta make changes and a start might be to patch the most ridiculous parts to prevent a user from breaking the neck.
    – Johannes_B
    Commented Nov 14, 2015 at 8:13
  • @Johannes_B Again, it is only you who feel like that about the templates. See paragraph 2.
    – percusse
    Commented Nov 14, 2015 at 9:26
  • 4
    @percusse but… but… xkcd.com/386 :p
    – cgnieder
    Commented Nov 14, 2015 at 9:39
  • I have tried to apply some patches for the most heavily used template. This is going on since June. Look at the latest commit the official maintainer made and how he changed my code: github.com/johannesbottcher/mastersdoctoralthesislatex/commit/…
    – Johannes_B
    Commented Nov 14, 2015 at 9:49
  • @Johannes_B You are fighting in a lost war. Let them be. If you are up to it make a superior template and people will flock to yours in a nick of time. Good thing about internet is that content doesn't matter for the masses, they go to places where it is crowded. And also think from the author's perspective a bit: Someone is constantly complaining about some esoteric TeX detail while many people are happily using my theme ;) See what I mean?
    – percusse
    Commented Nov 14, 2015 at 10:40
  • 1
    @clemens How to make a nerd go berserk --> i.imgur.com/hqFCN.jpg
    – percusse
    Commented Nov 14, 2015 at 10:43

I work on a textbook in Wikibooks since many years and I am one of the admins in the German Wikibook community. Thus I can only underline the answer by Clemens: We are really happy that he (and other authors of this site) started to improve the German Wikibook about LaTeX in the last weeks and I want to encourage you all to do the same!

I want to supplement his answer by additional notes:

  • Wikibooks and tex.stackexchange.com share the same license (namely CC-BY-SA 3.0). Thus you can copy and paste any answer from tex.stackexchange.com into Wikibooks. Take care that you add an URL + the answer's author name into the summary of the edit.
  • A good place to report errors are the talk pages which are associated to each article. See Help:Using talk pages for a tutorial how to use them.
  • Wikibooks uses wikitext which is similar to Markdown (Markdown is used on this site). Thus it shall be easy for you to learn the syntax of Wikibooks. See Help:Wiki markup for a good summary.
  • For LaTeX code you should use <syntaxhighlight lang="latex">..</syntaxhighlight> See Extension:SyntaxHighlight for more details.
  • When you need help in editing you can make a question on Wikibooks:Wikibooks:Reading room/Assistance (Wikibooks:Ich brauche Hilfe for the German project about LaTeX)

An additional note for the German LaTeX project:

  • Before Clemens, Johannes and others started to make improvements to this project, the project was not actively maintained. Thus you can make changes there directly and I would also encourage you to do so.

You must log in to answer this question.

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