Why should I go down the LaTeX lane?

Ok. I consider myself a real new-be when it comes to writing for print. I’ve been reading up a lot on LaTeX lately, more so since the latest beta of Scrivener, with its brand new MMD support and all. I get that LaTeX is a great tool when it comes to setting your type, that it’s free, adoptable cross-platforms etc. – but I’m still not sure that this really is something for me.

I rarely produce texts that are supposed to go directly into print. Why shouldn’t I settle for – say Apple Pages – when it comes to produce my PDFs? Why should I infester my system with almost 400 MBb LaTeX software?

Please, someone convince me. Either way.

I agree. :slight_smile: I implemented MMD export for those who want structural writing systems, but I think for most writers, RTF is most convenient. I am really thinking here about novelists, short story writers, journalists, article writers and so forth. As I understand it, LaTeX really comes into its own for science and maths writers, academics who have special typesetting requirements that would be a nightmare to achieve in Word or other RTF-based systems. More general writers really only want bold, italics and maybe an image here or there. Book writers will want some minimal formatting - such as adding a title page and some headers and footers - which is painless in most word processors (Word, Mellel, Nisus) - so that they can print the whole thing off and send it to a publisher. Short story writers will have similar requirements. Most article writers and journalists will have even less requirements - all they need to do is get their text written and into a format that they can e-mail to their editor. Most editors will want Word or RTF format for this - I doubt LaTeX would be welcome. :slight_smile: Many short story magazines want stories in RTF format, too. It seems to me that LaTeX is all about nice typesetting, but for the majority of writers (excluding academics who need special symbols), typesetting is not really their concern - that is left to the folk who prepare their text for publishing…

Anyway, I know Fletcher and AmberV will disagree, and I am sure they can give you some very good reasons for using LaTeX. :slight_smile:

All the best,

I’m with you on this, guys.

I’ve just installed LaTeX and something went wrong. My permissions got screwed up, everythig went into a spiral of crashes, and now my iMac wouldn’t sleep. I’m afraid I ought to recover my system with a SuperDuper – with a splash of wit completely unusual for me, I backed up my harddrive before downloadingg the Latex installers.

And, for this to work, I need to hunt down and install an array of templates, something called either Pearl or Jade, and learn some blue markups.

I’ve never needed LaTeX, and not sure why I started getting into it. I’ll go back to my cozy and comfortable Scrivener + RTF routine which both I and my editor like :slight_smile:

I do use // for italics though – a habit burned into my little brains by Ulysses.

I write for a number of different places and work the other side of the desk for a few others, and I’ve never seen any submissions guidelines that asked for my files to be delivered in Latex. Word docs lead the list, with RTF and Text files listed as acceptable.

I’ve never claimed LaTeX was for everybody, it’s clearly not.

You know LaTeX might be a good idea when:

The thought of using a product by Micro$oft literally makes you queasy
You are publishing your work yourself, and want it to look like a professional did it, not a kid with a crayon
You understand that “shell command” does not refer to trying tell a peanut what to do
You want to focus on writing content, not on worrying about formatting
You want to be able to create RTF, PDF, plain text, and web pages all from the same source document, without doing it by hand
You’re writing an academic document, particularly math, physics, or computer related. In which case you’re probably already wondering who Bill Gates is and why LaTeX hasn’t taken over the world.
You have this feeling in the back of your mind that clearly text based formats are better, but haven’t quite found the tool that makes them useable for you
You think that BibTeX is the greatest thing to happen to reference tracking since the printed page
You want to print your own books from an online printing service
You think that the separation of content and form as used in XHTML and CSS makes so much sense, that you can’t understand why people didn’ do that sooner

If most of the above statements sound like Greek (and you don’t live in Greece…) then MMD->LaTeX might not be something you need or are ready for.

If you are currently total happy with your Word or RTF approach, and you have someone else to take that RTF and actually make it look nice before having it published, then it is probably not worth the time and effort.

I have taken a couple of Cory Doctorow’s books that are CC licensed and converted them to MultiMarkdown text formatting (took just a few minutes, as the format is really close to what most people do anyway). I am in the midst of exporting a few to LaTeX using different stylesheets and will make a zipfile showing some of the different ways of formatting output with MMD, all from the same source document. What I need is a good academic document to do the same thing with, to show off some of the support for tables and figures and equations.

In this case, a PDF is worth a thousand words. If you compare the input text file, and the output PDF files and decide that the quality makes you drool, it might just be worth it.

The important thing is to stick with what works best for you, and for many people RTF is totally fine. If it works, don’t break it. But if you find yourself limited by the RTF and word formats, and you hate the fact that no two programs display either format the same way, and you want to generate high quality PDF files yourself, then LaTeX is the way to go (in my opinion). It’s much easier (in most circumstances) than trying to do it yourself with Adobe’s tools. And even Pages has a lot of limitations when it comes to PDF - most online printers cannot use the PDF that Pages produces.

I’m not trying to convince anyone that they must switch from RTF to LaTeX. But if you’ve been thinking about doing it, or you know that you need something new but aren’t sure what that is, take a look at the PDF’s when I get them ready.

Fletcher, this is a very good answer. Thanks a lot – makes the whole thing quite clear.

Maybe, some time, the reason will prevail and I’ll convert.

Actually, I completely agree with you, and I even put some words down in the FAQ that essentially say what you just said.

LaTeX itself is a specialist tool that by its nature requires a bit more computer savvy than most Mac applications out there. It sits on the UNIX sub-system, requires awareness of how to get there if you want to make modifications to it, running programs using a command line, and so on. However, where I would argue in its favour is if somebody was faced with the choice of learning InDesign/Quark or LaTeX for creating a book. In this case, LaTeX is a lot easier. A lot. All three of these systems are clearly more advanced than the median writer needs, and for some would require a lifetime to master; and years to get to the point where they are comfortable with it. I would never recommend anyone try to learn Quark so that they could send prettier RTFs to their publisher – that would be insanely silly. Likewise, it is fairly silly to learn LaTeX just so you can make RTFs out of it for your publisher. It is like learning Photoshop so you can send pictures down to the lab for developing. It doesn’t make any sense. Of course, then there are the hobbyists – the people who love using a computer for the sake of it being a computer. I am one of those geeky types, so I’ll muck around in things I have no business mucking around in – because I enjoy it.

That is LaTeX though. It does not speak as much for the concept of structural writing itself, which is much larger than LaTeX. The use of MMD itself as a nexus format for structural writing in general – I am much more of an advocate for that. The ability to get your writings into XML with the snap of a finger is priceless. XML can go – nearly anywhere. It is how the LaTeX file gets created. I create the FAQ using XML. I have a big Scrivener file that I hit export on, and a second later I have something down in BBCode for the forum. Still need that RTF for the publisher? No problem – export to RTF. As long as the source file can be described using text, or the application can read such a format and convert it into its internal format, XML can get your file into that shape. I do strongly believe that this is much more important to the average writer than LaTeX will ever be. It means wrapping your head around some different concepts, and learning to see text in a way that we have been conditioned to believe is “ugly,” but its really not that bad once you break out of that conditioning.

But, even that isn’t for everybody. Fletcher said it best: Don’t fix what isn’t broken. What these tools allow Scrivener to do is enter a market that very few applications claim, and even fewer do it well. That will help it appeal to a minority with special needs. It does not, at any point, mean that all writers should try to force themselves into a workflow that only a minority need.


You are right to separate out the use of LaTeX from the use of MMD. I suppose I focused primarily on LaTeX, and not on MMD.

The Markdown syntax (and by extension, the MultiMarkdown syntax) is not perfect. For anyone who’s interested, I recently learned that some people working with the one laptop per child project had discussions regarding Markdown as a default format.

I would LOVE to see them work on a replacement for Markdown and MMD for structured text files. More than LaTeX or XHTML, this is a fundamental advancement that could allow a great many other benefits.

For example, when you download a book from the Gutenberg Project, the text file is somewhat useless. There is not always a consistent structure, and you have to tweak each file before you can convert it to a format that is more useful for your needs.

For example, after a little tweaking, I have taken files from Gutenberg and formatted as high quality PDF’s (suitable for reprinting in a collection of stories from your favorite authors, or whatever) If the Gutenberg files were in MultiMarkdown Syntax (or something like it), the power and flexibility would be ENORMOUS.

MMD itself, is actually really simple. For the most part, just pretend you’re writing an email.

Whether using this syntax, if you know ahead of time that the only thing you will do with your Scrivener file is create an RTF to send of to a publisher, then it may or may not be worth your time to fiddle with MMD.

Without a doubt, MMD is a niche tool. But for people who need that tool, I have yet to find anything else like it.

I wonder how things will fare when POD publishing becomes common (if it does, as it looks set to)… Publishers will need to use a text format that the POD machines can download quickly and which can be turned into print quality format just as quickly. It is likely that general users will start to adopt such formats/syntaxes so that they can publish their own books or whatever via POD machines… Or maybe I have just been reading too many “Tomorrows World”-style articles. :slight_smile:

I think just such a thing could happen. One possibility would be LaTeX or something like it. Another could be XHTML and CSS. Or DocBook. Or …

The problem right now is that we are in that transition point where some leading edge users are pushing the boundaries, but there is not yet a consensus standard that everyone can get behind. The closest thing thus far seems to be that PDF is sort of a “common final pathway.” Which is why I worked hard on getting MultiMarkdown to be able to create high quality PDF’s, and LaTeX seemed like the best tool at the moment to do this.

But I think the key on the other end of the workflow will be a simple to read and a simple to create plain text format that anyone can create on their own, and then they can use “black box” tools to convert it to PDF. I suspect that most writers don’t care about the inner workings of that black box, as long as it doesn’t get in their way. That’s what I am trying to do with MMD by having some well chosen defaults that suit the need of most users without their having to think about it too much.

As a LaTeX user for the last fifteen years, I must say that I cannot imagine it ever becoming a common documentation format. The problem is that it’s just not very readable in electronic form. Yes, it produces awesome PDFs. Yes, I’ve spent over $300 just to buy a single font from Adobe so that I could typeset my PDFs in the most beautiful form of Garamond that I could find (complete with dropped baselines on 6s, full ligature and kerning support, etc).

But try as I might (and I have tried this experiment a few times), once I’ve converted a document to LaTeX, it might as well have died. I never go back to my LaTeX documents. There is something so inherently distasteful about the syntax, that I unconsciously avoid touching them once they’ve been formatted.

RTF suffers from an almost opposite problem. Because it has no real concept of styles – and because it can get very touchy once I’ve finally made things look pretty – I find myself looking at them, but not touching them for the opposite reason: because to do so is likely to somehow, subtely destroy the formatting. I can spend hours just to make an RTF look semi-reasonable; and then one formatting decision later, I have a horrible mess on my hands that again I run away from in fear.

My pursuit of a good documentation system: where the ASCII is clean, and amenable to quick alteration; and where the PDF is beautiful, and follows exactly to the overall style: have let me again and again back to POD formats.

At first I threw in my lot with Perl’s POD. Then I wrote my own system for Emacs, called Muse. Now I am looking at MMD. They recover a lot of the flaws of RTF and LaTeX, without sacrificing all that much.

But of course, they do sacrifice something. I’ve never found a contextual layout system that really understands poetry. Or building indexes. Or tables. Or supporting math formatting. Usually the system in question has a favored output methodology (XHTML in the case of MMD), and allows you to “inline” raw data in that one format to get around the weaknesses of the markup syntax. (My own Muse system does the same thing with Lisp).

I guess what I’ve been looking for all this time is a marriage of two things: a simplified input syntax, and a tool which manages the greater complexities needed for real document formatting.

This is why Scrivener+MMD is so far looking great. I’m about to start writing a .NET component for its MMD syntax, so that I can write website content in Scrinever, and post it transparently via ASP.NET.

But the potential here is for Scrivener, as an interactive tool, to bridge the gap between simplified markup and complex formatting. Not that Scrivener needs to do the actual formatting – for this, I think LaTeX is far better – but it could certainly do things like allow me to graphically layout a table, and then behind-the-scenes turn that table into acceptable LaTeX/XHTML syntax.

I don’t think MMD should be used for everything, and Scrivener just become an MMD front-end that passes the untouched text on to some Perl scripts. I think Scrivener can do what nothing really does as well right now: offer an environment where the writer focuses on CONTENT (not formatting), and produce an output document which can be passed to a subsystem whose job is to focus on STYLE (not content).

As for using MMD because it allows for a simplified intermediary system that others could standarize on: The fact is that formatting cannot be simplified too far. Once you start wanting real formatting (or even just one exception to a rule), you will break whichever simplified markup system you’ve chosen. And since the world cannot ever fit into a single format, I believe there will always be a need for intelligent tools to bridge the gap between simplicity of input and complexity of output.


Totally agreed. That’s one of many reasons to use MMD or something like it as a “front end” for LaTeX. Despite all the work I have done with LaTeX over the last year, I have barely scratched the surface.

MMD does offer tables and math formatting. I am looking at a method to offer indexing and a glossary. As for poetry, it could be done using a variation of the format used for code (aka verbatim in LaTeX) that could be modified to use the verse formatting quite easily. I am looking into this as well. Prior to Scrivener, I suspect that most MMD users were from the science/computer geek background, whereas now I will have to look to the needs of the writing geek pursuasion as well. :wink:

The problem is that at some point on the complexity curve, you have to do it yourself. MMD is great for something like 90% to 99% of what a writer needs, depending on the type of writing. For most fiction writers, it is probably sufficient as is. For non-fiction, it depends. But when you start getting too esoteric, or too picky, you need to do it yourself. MMD could translate your document into LaTeX and handle that 95% and then you manually add the other 5%.

I believe I mentioned this elsewhere, but I recently converted a friend’s PhD thesis from Word to MMD. I created a custom XSLT for it, and it looked great as a PDF, and I even had it printed as a hardcover book. But there were something like three separate instance where I needed to add some hand-tweaks to the LaTeX code to make it look right. For a 135 page document, that’s an acceptable trade-off for me. MMD handled 135 pages, including a table of contents, and all the cross-links, table formatting, etc. I had to handle 3 lines where there were run-on words that needed manual line breaks and hyphenation. I couldn’t expect an automated system to do much better than that. Oh, and it also handled all of the references, and the bibliography.

Throw in the fact that it’s all free, and it looks even better… :wink:

There’s a trade-off between the learning curve of a system, and how precisely you can format the output. LaTeX has a steep learning curve, but can give you the exact output you want if you are willing to work for it. MMD has a much lower learning curve, but doesn’t offer quite the same degree of control over the output. Actually, it’s designed to limit your control and rely on standard formatting to give the output most readers expect. That said, you can take the XHTML or LaTeX it generates, and hand customize it to provide what you want. And in the end, it is still much easier than doing the whole thing by hand.

MMD is by no means the end-all, be-all of plain text syntaxes, and I look forward to whatever replaces it.


My impression: LaTeX in its most authentic form is first and for all a hobby of an intellectual kind of people: and much more of intellectuals who work in the field of science than of intellectuals who work in the humanities. LaTeX enthusiasts are people who like to be challenged, who don’t panic, who tend to laugh when things take an unexpected turn, and who don’t mind working for many hours in order to get things right.

As far as I know, in the professional publishing world LaTeX is virtually unknown. There everybody works with Quark or with Indesign. It is not my intention to offend anybody, but I don’t have a high esteem of the professional qualities of many people who work as layout designers in the publishing world. Of course there you have a number of excellent people too: exhibition catalogues, for instance, are often done brilliantly, by designers who really know their métier. But as soon as it comes to designing the layout of mainstream books, the results are often disappointing. In recent times I have seen a number of really expensive books put on the market by publishing houses of international renown, which from a technical point of view were less than modest. Many publishing houses simply don’t like (or can’t afford) to invest in quality design.

I really believe there is a market for applications which focus on people who want to design their books by themselves, but for whom LaTeX, Quark and Indesign for various reasons are no options. Mellel, Nisus and Pages could become such an application, but are all still too limited in this respect. And perhaps Scrivener + MMD could become an interesting option too.

Timotheus, you’ve hit the nail on the head. In my opinion, the word processor of the future is going to resemble something a lot more like LyX and a lot less like Word. It will be easy to use; follow the What You See Is What You Mean philosophy; output to some universal XML format; and contain a number of transforms which create ready-to-print PDFs. If you have never used LyX: The way text is displayed in the editor is not how it will print. It uses an internal style system to portray structure. A title is big and bold like you would expect; it is automatically numbered. It will allow the user to create a LaTeX document without even really know what it is. It would, like LyX can do, create a PDF with one menu command.

Where LyX falls short today is that it was designed on Linux, and as such is a bit unorthodox to Windows and Macintosh users. It has a steeper learning curve than most people are willing to take on. It was, in short, designed to be best harnessed by LaTeX users. That is why the future will be something like it.

The problem with today’s word processors is that they get people very easily to an “Okay,” level of quality, and then stop. Even an expert finds it extremely difficult to take it beyond that okay level. What JWiegley had to say about the woes of getting something looking good in RTF are a big part of this problem. Style based word processors have it a lot better than trying to create a book in raw RTF, but they still bear many of its same flaws. Something that had the philosophy of LyX could get these same people to a publishable level of quality, without hampering the ability for professionals to go as far as they want.

So what about the back-end? I think LaTeX is underestimated on that score. You are right in saying that it is rarely used in the design world, but where it has a much bigger hold and respect is in the automation movement and the areas you already pointed out. There are many places where nothing else is accepted. It has just as much a hold on the publishing realm as Word does in the relationship between writer and publisher in the popular market. Where it has a growing hold are the places I’ve mentioned in a previous post. The automation industry is moving towards it. The purpose of this rant is what could be, not what is.

Does what it looks like in a raw form really mean anything? I think not. JWiegley is correct in that, right now it could never be mainstream because it looks only a step above raw RTF – but does that matter at all once a seamless start to finish workflow is in place? Does it matter to the professionals who want to muck around in it? Just look at the Web for your answer. Raw HTML is a mess of readability. XHTML is a large step better, but it is still nothing you were ever want to write in all of the time. Does that matter? Billions of people surf the web and never once think of the way raw HTML looks. Even designers, these days, do not have an in depth knowledge of it – trusting tools like Dreamweaver to do all of the dirty work. It matters neither to the author, nor the consumer, what HTML looks like in its raw form. Only the professionals and hobbyists care. With the right steps taken in the market and industry, LaTeX could be no different! Whether or not it dies once generated is beside the point, because I posit that in its most automated form it will be nothing more than a generator. Going back to LyX, you make a change in your document, hit Cmd-T, and instantly you have a PDF. You never see the LaTeX, you never open it, it doesn’t even exist as far as the user is concerned.

As with HTML, LaTeX could easily become the intermediate generator for visual content. As with HTML, there could be the powerful tools like Dreamweaver which control nearly every aspect of it, and the easy to use tools like NVU that allow practically anyone to make a decent web site at the expense of full spectrum power. The key is that both utilise the same underlying technology. This is the fundamental gap that publishing currently has. The tools use an entirely different underlying technology in nearly every software package – easy to use and professional alike. One could argue that this is because a web page is fundamentally much more simple than all of the needs in publishing. On the surface that is true, but there is nothing in LaTeX that limits you – it actually could serve as a global generator for all publishing, with the right tools.

Something like MMD is a big step toward bridging this gap that LyX attempts to bridge. LyX requires no knowledge of LaTeX to function, but it requires a lot of knowledge about LyX, which in some cases is just as intense as knowing LaTeX – that is its flaw. MMD does not have that flaw. It requires no knowledge of LaTeX, but it is extremely simple to use. Its flaw is different: It does not have the visual component. That should not be underestimated, especially since authors have been spoiled with visual representations of their documents for more than a decade now. Its other flaw, as has been pointed out, is its lack of full spectrum power.

This hypothetical future word processor would be as easy to use as MMD, be visual in the sense that LyX is visual, and require no knowledge of LaTeX. It very likely would not be full spectrum, and that is where it would fall more into the NVU realm than the Dreamweaver realm. There will likely always be a need for this separation. There you have the difference between a next generation LyX, and a next generation InDesign – the key is getting both of them utilising the same underlying technology.

The Scrivener+MMD connexion is very interesting indeed. When I first had the idea of merging the two, I never dreamed that the designer of MMD would get involved, and be open to looking into changes which would make MMD more suitable for publishing books. While it still requires one to follow the pros and cons of MMD very closely, I do believe it is a big step towards this future ideology. There are already visual aspects of Scrivener that turn into structure. Hierarchy can become XML outline; a few of the special features like footnotes and annotations can become meaningful meta-data in the document. This is but a taste of what this hypothetical future word processing environment would be like. Do you have to know what \chapter{Step One} \label{stepone}, looks like when you move a document to the second level of hierarchy in your Scrivener Draft? Of course not. You don’t even need to know that it looks like “## Step One ##” instead of “### Step One ###”.

Of course, my entire argument rests upon whether or not the personal POD revolution will ever take place. For evidence of that, I invite the comparison of our sister art form – the music industry. It is a few steps ahead of the writing industry. The big labels took to sacrificing more and more quality for bottom line, in the face of sinking revenues. This happened over decades, but became obvious to nearly everyone in the past decade. The music market became a glut of second-hand music with a pretty face. It became extremely difficult for a truly talented musician to get their music out there – even artists that had already established themselves in the industry, due to strict contractual obligations written to preserve this shallow-is-better mentality. The book industry is only just now entering that “glut” phase. An article that Keith linked to is one of the first well written opinions on that matter that I have seen, and I am sure there are many more out there. The author noted that walking into a bookstore these days is to see wall to wall of marketing hype for tripe that should have never seen the light of day. Not at all unlike walking into Tower Records. Things not only published for money, but written for money and nothing else. This is precisely the state the music industry is in. There has always been an element of that, ever since the central publishing to super-store system existed, but only lately has it become a major stifle in the state of freely available quality.

Since that time, musicians have taken matters into their own hands. With the combination of the Internet as a marketing engine (via paid and grass roots alike), innovate distribution channels like CD Baby, and cheaply acquired digital home studios, there are now many thousands of artists – genuine musicians – who can get their voice heard and would never have had that chance otherwise. Independent has become the mainstream in both cinema (I specifically avoided comparison with cinema earlier, because it is such a collaborative art form, but it does bear a resemblance to how the music industry has withered) and music. It has taken little longer for the book market to reach the point where independent became necessary. We are only just stepping into that phase. If these other two entertainment vehicles have followed this precise pattern, it would be folly to think that literature will not follow suit given how many of the early warning signs are already extant. I think the next five to ten years are going to be very interesting times for authors. The POD potential, combined with the Internet, combined with a few brilliant ideas like a CD Baby analogue (take a look at lulu.com and creativecommons.org for what this might be like) could very will cause a boom in quality independent publishing.

Ten years ago, independent music was sneered at as being consistently low-quality and rarely worth looking in to (the labels exist for a reason, was the common retort). I am of course speaking in generalities, there have always been sub-cultures of people who venerate the non-mainstream, no matter how good or bad that non-mainstream may be. Self-published books still carry that stigma, but that is changing for precisely the reasons I have put out in the previous paragraphs. People in the know are losing their faith in the literature engine, and the small-press and independents are starting to flourish. The bookstore I frequent, Powells, already has a large section devoted entirely to this phenomena. It could be argued that the explosion of the zine format falls squarely within that. It is only a matter of time before this bleeds into the popular mainstream, and the old mantra: The publishers exist for a reason, will lose its relevance.

Back to the software: One thing that still sets independent books apart from their mainstream alternatives is the quality of the print. Small press books are frequently low quality, and self-published even worse. This is no different than the “garage band” syndrome a decade ago. But as with the rise of cheap home studio equipment enabling everyone near professional quality music releases – there will be a rise in the tools necessary for authors to create publish quality books straight out of their desktop computers. This is probably even more destined to happen, since it is several orders of magnitude easier to design a good book, than to do quality sound engineering! Of course, it could be argued that quality editing is just as much a resource problem, if indeed a problem of an entirely different nature.

To wrap up this very long rant: Back to LaTeX. What it represents is a partial analogue to the cheap home studio. It solves the problem, very cheaply (I hesitate to say free, since time is money in this modern world), of creating publish quality works – but it does not address the problem of quality editing. It solves typography and composition, but not content. Content quality is outside of the scope of this argument, interesting though it may be. It could serve as the universal generator of the future, there is nothing about it that limits it in that sense, except the current lack of tools.

And that is what this thread should really boil down to. Whether or not LaTeX takes a lifetime to master, or is ugly in its raw form, is beside the point. The key to focus on is whether or not it is possible to find a layer of abstraction between it and the user. I do believe that Scrivener+MMD is a giant step toward that. It is not the end, but a big step. It represents an abstracted, easy to use interface that has powerful book management features and exports to a powerful, universal generator format.

Well, I’ve gone on too long, and most of this is not relevant to today – unless you are part of that society which is riding the crest of the wave.

I think we’re all circumambulating around a similar feeling: That present systems are highly useful, but not ideal. We want a tool with both precision and “luft” – the feeling of joy that glues you to the keyboard as you engage in the writing process.

For my own writing needs, MMD happens to be 100% effective. Also, since I do not expect publishers to accept a formatted PDF from me – I assume they prefer to stick the text into their own systems – I have little need for the precision of LaTeX.

But ah, I love the beauty. So there is LaTeX, allowing me to provide killer output when I do have to publish something on my own.

What will be the solution? On pen and paper, I write with a Pelikan M805 on Moleskine notebooks. I’ve noticed that whenever my pen is out of commisson, or I’m out of notebooks, I simply stop writing. But when I have the two, I actually look for reasons to use them. They are that pleasant to use.

On the computer, I have no such analogue – yet. Scrivener so far is the best of all I’ve tried (and damn, I’ve tried a lot; Lyx was extremely promising, but it feels too much like writing in a straight-jacket).

I don’t want to see MMD extended with funky syntaxes to provide indexing and math and the rest. Can you imagine reading an MMD file extended to provide indexing? What would be the point of that?

Here’s what I’m thinking of right now: Defining an absolutely exhaustive XML schema for typesetting documents. Something like LaTeX in XML (I believe a project like this already exists). Then, remove from MMD the ugly extensions that exist for doing things that could be expressed in this XML dialect.

Thus, when we do need those three-exceptions-within-150 pages, we can use tags to break out the XML formatting, which can be properly translated to LaTeX and output to PDF.

Give me a simple, simple syntax for 99% of my writing, and I will happily learn an ultra-arcane subsystem for the last 1%. Let it be the domain of specialists and FAQ boards, I don’t care. But it gets us toward a tool which is not encumbered by its own abilities!

Remember those Kung Fu movies where a humble old man seems sweet and innocent until 50 people try to attack him at once? Watch “Zatoichi”, for a good example. He’s the kind of writing system I want. :slight_smile:


Now we are bleeding two threads together :slight_smile: but I did have one thing to say: I don’t see how a syntax as simple as a caret in front of an indexed word would be at all intrusive, or difficult to input. If you had a cluster-file somewhere taking care of associations, it would literally be as ^simple looking as that, certainly a lot less mind and eye expensive than a chunk of XML wherever you need an index entry? Of course, I’m not Fletcher, so I don’t know what the syntax would look like, but given his syntax decisions so far, I think if he does decided to go with indexing it will be out of the way. Perhaps I misunderstand what you are saying, though.

By the way, you can insert XHTML right now. It gets passed through the engine just fine. So if you want to add a special index set to an XSLT, then it is pretty simple to define your own features. If you don’t mind messing with Perl, you can even make your own syntax. That is one thing I really like about the way Scrivener+MMD is set up. You can use the defaults, but if you really need some specialist XSLT or something, adding it is no problem. It goes a long way toward the 1% Kung Fu moment.

Oh yes, and I am curious about the typography XML project you mentioned. Do you happen to know where I can read a little on it?

Absolutely agreed. Which is why I am working on the glossary first, because what I have so far is this:

Sample paragraph discussing a whatsit[1].

(The (U) indicates that this should be filed under U, rather than W for some reason. Useful when your term is a symbol)

This syntax works just like footnotes, defaults to footnote behavior in XHTML, and is not really any more intrusive. MMD is responsible for converting that into the XHTML equivalent that basically adds some spans around the term and the “file under”. The XSLT is then responsible for converting it into something else (e.g. LaTeX).

I am open to alternative tags (the “Glossary:” part that you think might be better).

I too am not sure about the indexing part, though. I realize that just because I add some syntax, doesn’t mean everyone has to use it. But it does decrease the readability of a plain MMD file.


  1. Glossary: Whatsit (U)
    A subclassification of thingamabob. ↩︎

I suppose this could work, even for the typical case of requiring multiple forms of phrasing:

And then you go about ^refilling the pen.

Oh, and I looked up that XML project. It seems that people can output LaTeX to XML, but I found only “wishlist” entries for XML/Latex. Perhaps it’s a project worth attempting, using a simple stylesheet to render it back to true LaTeX. I’ll look into it.

Also, got a fair ways into coding my .NET MMD parser today (using Mono). I found a Markdown.NET parser, but it’s old and ugly and is basically a bunch of regexps outputting XHTML. My own parser tears the document down and represents it internally as a rich hierarchy, so I can later decide whether to go to XHTML, or something better. I want an intermediate format which is a superset of all systems to be used, at an expected cost of readability, but designed such that only the rarest of users need delve into those aspects of it – and yet they remain for the delving. Perhaps it will have to become a branch off of MMD in the end, to incorporate this move away from XHTML tyranny.


Many thanks for this discussion. It really helped making things a little clearer for me. I decided to trash the MacTeX image file instead of installing it. Although it would have been great fun to get to know LaTeX a little better, I don’t think it would have been worth the effort right now. I should really use what little time I have to actually write, not learn new software.

This doesn’t mean that I don’t favour structural writing. Actually I must say I still prefer working with plain text, rather than RTF. I find it much more natural to use markers instead of applying bold and italics during the creative process. This is the one thing that Ulysses actually does better than Scrivener, which of course isn’t much of a surprise. Scrivener excels in all other aspects, though.

MMD is great, but I do have some personal issues with it, mostly when it comes to paragraph syntax. Yeah, I know. :wink:

As someone who knows LaTeX well, I think you made the right decision. It has a monster of a learning curve, and what you’d get from knowing it – beyond what automated tools can do for you, like Scrivener – would hardly be worth the time spent writing that you’d have lost.