Slashdot Mirror
Ask Slashdot: Do You Use Markdown and Pandoc?
BartlebyScrivener writes "I am a author, screenwriter, law prof, and a hobbyist programmer. I love MacVim and write almost everything in it: Exams, novels, even screenplays now that Fountain is available. I use LaTeX and WordPress and so on, but several years ago I discovered Markdown and the wonderful Pandoc. I searched Slashdot expecting to find lively discussions of both Markdown and Pandoc, but found nothing. Do Slashdotters look down their noses at these tools and do their work in HTML and LaTeX? I can't imagine computer geeks using Word instead of their favorite text editors. If not Markdown and Pandoc, what tools do Slashdotters use when they create documents that probably need to be distributed in more than one format: HTML, PDF, EPUB or perhaps even docx?" And then there's DocBook, LyX, and a host of other markup languages. What do you use, in what context?
Markdown, Pandoc, and Vim/MacVim are my primary working tools for writing documentation.
I use landslide. Somewhat fancier html5 slide output.
...is good enough for me. I will not get any more explicit than that.
reStructuredText. It is a lot more powerful than Markdown while still maintaining beauty. It is a bit more strict with formatting though, as is always the tradeoff for more power.
I use the Rmarkdown flavor of markdown.
Perhaps there's no conversation because there's no particular controversy, and thus, nothing to discuss?
is the only tool real geeks need to write multi format documents.. and bbcode of course. ;)
I use markdown all the time. Mostly moc on Mac to edit it.
Markdown is gaining popularity again thanks to the environment and community around GitHub. That said, I afraid that most people still would prefer wysiwyg systems, as it easier to use than 'feel like a programmer' when using weird codes such as HTML, MarkDown, bbcodes, MediaWiki etc.
shamelessly putting forward a young project of mine but look, here I'm using markdown (piped to anything, pandoc works fine) inside HTML documents and in multiple blocks throught the document.
webnomad is a minimal implementation to use <markdown> </markdown>inside a plain html and bootstrap styled website. I hope this inspires someone. I'd advocate for markup tags inside HTML6 eheh
I use the BEST editor there is - $EDITOR.
I save files as 7bit ASCII, hard-wrap at column 72, and if a file extension is required, i use .txt.
1% APY, No fees, Online Bank https://captl1.co/2uIErYq Don't let your $$$ sit in a no-interest acct.
I can't imagine computer geeks using Word instead of their favorite text editors.
Indeed, everyone has his favorites. People have argued text editors to death. Find what works for you and use it (and stop flaming us). Also, programmers write software, not documentation. All document conversions are left to the secretary (and she may do it in any way she pleases).
I use HTML when I just want a document that works everywhere. When I want to impress or actually care about typography, I use InDesign and render to PDF—nothing beats its optical margins and paragraph-optimized justification.
Markdown strikes me as something really great for people who don't know HTML. Otherwise, it doesn't really save you a significant amount of typing and it is significantly less powerful.
I think you have a bunch of axis going on here at the same time.
1) Open source tools vs. closed source
2) General use vs. specialized
3) WYSIWYM vs. WYSIWYG editing.
Let's start with (3). I generally like WYSIWYM more than WYSIWYG environments. That being said WYSIWYG is very useful for content where consistency doesn't matter as much. While there are WYSIWYM systems for presentations but they only really work for data publishing better than WYSIWYG. I think a fairer comparison on this axis might be LyX vs. Word or FrameMarker vs. Word. Going up something like higher end composition engines vs. something like InDesign.
Then there is general use vs specialized. LaTeX is optimized for text with equations which is specialized. Fountain is for screenplays. That's not the same as a product like Word which is all purpose.
Finally open source vs. closed source is a complicating factor. When we consider Word do we consider all the myriad additional cost extensions, for example SharePoint or just the core product? For OpenSource do we consider the entire platform and how these components work together? More importantly closed vs. open goes beyond editing to broader computing issues.
Because it has C-x M-c M-butterfly.
I'm using pandoc sometimes, but in my experience it is a bit like programming in BASIC or using some RAD tool in programming - 90% is easy and achieving the remaining 10% of functionality is almost impossible without reverse engineering. That being said, you only get problems when tweaking slides or when you have a lot of formulas. For screenwriting and non-math-extensive documents it should be fine and sure beats any word processor. In summary, let's just say pandoc is better than MacFarlane's philosophy. *ducksforcover*
I use asciidoc as a quick and easy way to take minutes of meetings. I usualy mail the participants a quite decent-looking pdf version of the minutes immediately after the end of the meeting. I also use gitit as a personal wiki. I must admit that I sometimes confuse the syntax of the markdown used by gitit with the quite similar asciidoc. I could switch to pandoc for my minutes, of course, but the pdf produced by asciidoc looks a lot nicer than pandoc's
I've been using Sphinx to write user's and developer's manuals, as well as a host of other things. It supports a number of output formats and uses reStructuredText as the input format. reStructuredText can be a bit much, but using snippets helps relieve that pain tremendously. It's a really fantastic tool.
That's classic!
Not even joking. It's habit. I use it for my creative and technical writing and find myself doing this:
###to buy
----
* toilet paper
Yeah I do. I use macvim, pandoc, markdown, latex, and apvlv, and ... every day. Not sure what I would discuss about it. Not sure that I am lively either.
If you're writing a novel, a tool like Scrivener is a lot better than a text editor of any particular sort. It's designed for writers and makes it easy to do things like keep track of and organize all your notes, which if you're writing a novel is going to be far more important than whatever command is used to change the font.
-- "So they told me that using the download page to download something was not something they anticipated." - Bill Gates
I only use the old reliables - WordStar and WordPefect. Markup the way nature intended.
I searched Slashdot expecting to find lively discussions of both Markdown and Pandoc, but found nothing.
Lively?!
Have you seen the comments today?! It's like being in a bar with a bunch of angry drunks arguing about football, politics and religion - all at the same time.
And you're bound to get at least one "Well, I programmed emacs in Lisp an Scheme and I have a ''do my fucking work for me'' app because I'm fucking smarter than all of you morons. And I just sit home and drink beer and collect a paycheck. "
And I'll get flamed for calling it an "app".
Yeah, yeah, yeah, I'm a fucking moron - there, I saved you some typing.
Your welcome.
Pretty much the only thing we use at this point. Everything is MD + Pandoc!
User org and export to whatever format you need. It is the swiss army knife of text mode. You can write blogs (octopress-mode), documents (exporting to latex), and programming (babel-org). And on top of that you can do todos and agenda. Check http://org-mode.org
pandoc your-file.tex -o your-file.odt
this is all I know about pandoc. I am new to latex, but learning it and want to use it for journal papers, books and presentations.
I use latex2html to convert tex to html. I've used vim, nano, emacs but could adopt them. My favourite text editor is gedit.
We are working on compiling an (e)book with texts from several sources, all collected within etherpads for doing proofreading and translation with multiple editors. One suggested markdown for typography. To get further from there I stumpled upon pandoc. After getting rid of some special-character-issues (thus installing some latex-extensions for that), it does a pretty good job for exporting pdf or odt (haven't checked out epub yet). Seems like I will use odt for final design & typography and export the epub and print-pdf from there. I get the txt-exports from etherpad-URLs with kind of a shell-script. There wil be a another shellscript for renaming the files to get them in a particular order (so there will be csv-file with "pad-id/name/slug" and "number-for-sorting"). In the end it will be a book by the german blog netzpolitik.org about the NSA.
John Gruber is an intellectually bankrupt opportunist and markdown still makes me laugh. Apparently HTML is just so hard as a tagging language...
And you've failed to convince me that either one is necessary.
I use vim and lyx when I edit latex.
Now get off my lawn, and stop betteridging headlines.
I am more of an AsciiDoc-fan myself. AsciiDoc syntax is straightforward and maps to DocBook,so it's perfect for writing books with all its specifics.
I use LibreOffice, because I hardly do html and need to keep compatible with the crap they use at the customer and the employer when it comes to office formats.
In general, most commercial text editors are lousy at both editing and layout, but they contain all the "features" that people who don't understand either find handy to mangle their documents with. LibreOffice isn't that different from the others in that respect, but at least it's free and it runs on all the systems I use for daily work, regardless of their OS.
When I have to work on code or markup languages, I generally use VIM, since it's also on practically all the systems I use for this sort of work.
Since I have to often rely on systems with nothing but generic stuff installed, I don't want to have to bother with setting up the "perfect environment" all the time, so I adapt by using what's easily available.
I was promised a flying car. Where is my flying car?
I use Markmin. It is more flexible than markdown. For example it supports LaTeX formula & references, HTML5 media tags, QR codes, colors, and classes.
The converters (markmin2html, markmin2latex, markmin2pdf) are single files without dependencies.
http://www.web2py.com/init/static/markmin.html
Sounds like you really want a confirmation that your preferred choices are shared by others. Well, not shared by me.
Personally I use troff for letters and such (PostScript for the printer or PDF otherwise, editing not required), with a custom macro set that does letterhead and so on, I use nvi (no vim, no emacs, TYVM) and I cooked up my own vehicle (in awk) to ease writing html and css. I'm not expecting anyone to use the latter (it's not released) and troff, well, that particular lawn is nice and empty. Note that these are very much my preferences, and that's all there is to it. I don't need to share anything whatsoever, except the end results that aren't ment to be editable by others.
NIH? With the awk thing, sure. Same thing with markdown (which I don't like very much, though it's not the worst in use. for say a wiki I'd prefer CREOLE) and all the many other attempts. Upshot for me is that if I need to change the output I can do so relatively easily. It's my vehicle, I know how to deal with it. If I would need interop, well, plain text does amazingly well, especially if I stir in a little scripting.
Otherwise, it heavily depends on just what you're doing, who you're working with, how much they are married to broken systems from convicted monopolist vendors, how much you can afford to tell them to take a hike, and so on, and so forth. There really isn't a good, "universal" answer, and while many fields could be improved immeasurably, perhaps such an "universal" answer is impossible, if it even makes sense to strive for this. Better conversability would be useful, though. While I'm taking swipes, we've seen the mess with a certain widely touted solve-it-all from the guys who tried to standardise the web, for example, though quite often the real horrors get to hide behind the facade it so nicely bloats the actual content with. Small favours, right up until they up and munge or eat your data. Yes, I've seen that happen.
Curiously, I heard $author relate how O'Reilly dealt with his troff source for a book: Convert it to docbook, then run that through their own scripts to generate troff again.
Context is important, one could surmise.
I take notes using the syntax of my employers wiki which makes it easy to cut and paste from my $EDITOR="joe". My custom syntax highlight file might not be perfect, but it certainly makes the text more colourful: trac.jsf
I use Emacs for writing; most of it is in XML (usually DocBook, sometimes TEI, occasionally XHTML). I use XSLT to transform it to LaTeX if I want publication-quality PDF, but more often the document is the input to other people's toolchains which want XML first. I occasionally transform to other formats (HTML, Word, or some wiki formats which are largely MarkDown-ish).
I author in XML because most of what I write involves quite detailed and very specific structure, and DocBook and TEI provide appropriate levels of markup for this. I made a conscious decision to go this way a very long time ago, when it was all still SGML, and I have never regretted it.
Most people don't have that level of specificity to adhere to. All the formats you mention have their areas of application, even Word, but there is a growing undercurrent towards using HTML5 as the default format, driven partly by the fact that Ebooks use it. The publishing industry is very interested, as they hate and detest Word, and only use it because its change tracking is useful and it has usable style-editing, which OpenOffice and friends don't have (ie they have no style margin like Word). It was very clear at the XML SummerSchool last month that there is growing support for HTML5 in editing tools, and some new advances in editorial control (eg systems like Xopus, FidusWriter, and Poetica) mean there may even be a way to escape from Word :-)
--
Disclaimer: editor interfaces are my thesis topic; I have no connection with any of the above except the XML SummerSchool.
...that any Slashdot posting opened up a new world to me. This is one of those - way too rare - postings. I use to write flat text, and hand that to secretaries or colleagues for formatting it into whatever they want or need: html, word, slides, whatever. Software documentation, to me, is generated javadoc - so basically html generated out of flat-text code comments. Of course I knew and know about the existence of LaTex et al.. Yet, as a software architect, I guess I have been, for decades, plain lazy. I simply write flat text in emacs or notepad++. Man, you opened up a universe to me !
Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
It depends what you're trying to do but for documents DITA is fairly robust. It's a steep learning curve to get into xslt, xslfo, and xpath if you're not used to them.
We use markdown in git for all our documentation.
Documentation about internal processes is a bunch of markdown files, with a haskell program which uses pandoc's libraries to add an #include statement, so the documents can be split up nicely.
We also use ikiwiki for project wikis, so it's also markdown in git, albeit a different dialect.
We also have other tools for more structured documentation, but these are yaml files with text portions being markdown, in a git repository.
All of this is designed so we can use git workflows for documentation work, rather than just the coding work, since it allows patch review, merging and powerful history tracking.
Mou (http://mouapp.com/) + pandoc on mac for me. Really nice environment. Mou is a lovely app to use and pandoc is sublime; a great example of the unix philosophy of simple but powerful tools that can be composed.
For multi-part docs I combine them with Sphinx (http://sphinx-doc.org/). I write in markdown, use pandoc to convert to rst then sphinx to generate pdf/html as required. I know I could write in rst and simplify the chain, but prefer markdown to write in - particularly with good authoring tools like Mou about.
Going back to Word/LibeOffice feels cumbersome now.
As a web developer, I understand the need to allow users to create content. However, I consider full blown web-based WYSIWYGs (such as TinyMCE and CK Editor) to be terrible tools. Yes, they take care of most of the dirty details, but they also have the capacity produce bloated garbage markup. I've always found bbCode annoying. I've used a couple of Wiki syntaxes as well (MediaWiki, Jira), and find them only slightly less annoying than bbCode. Because I know HTML, I prefer to just write HTML when I know the final format will be HTML.
Markdown syntax is clean, succinct, and can be extended when needed. The vast majority of non-savvy users only need to do basic formatting (bold, italics, headlines, lists) and Markdown covers that very well, in a way that the user can't do much damage to the prescribed styling of the content.
I've very briefly experimented with LaTEX, not enough to have actually used it for anything. I was not aware of PanDoc, but it looks very interesting.
I like the emacs org mode (also available for vim, and even your mobile), though primarily for structured outlines/note-taking, its markup allows document formatting and it can export to a range of formats including PDF.
Notes, documents, TODOs, all in one format.
I use Latex for almost all my texts: small, large, letters, presentations, what-have-you.
Also I copy texts from web sites to Latex and print it out for reading.
For texts that are only for me I use: plain text.
I don't understand why some (most) people are scared of Latex. In Linux you just install per package manage Latex+Kile and then you are one mouse click away from a nice Pdf document.
Small example from http://neurotheory.columbia.edu/~ken/cargo_cult.html
Gives you a very nice document to print out and read at your leisure.
\documentclass[10pt,abstract=no,toc=flat]{scrreprt}
\usepackage[hscale=0.69,vscale=0.79,heightrounded,includehead]{geometry}
\begin{document}
\begin{multicols}{2}[\section{Cargo Cult Science}]
During the Middle Ages there were all kinds of crazy ideas, such as that a piece
of of rhinoceros horn would increase potency. Then a method was discovered for
separating the ideas--which was to try one to see if it worked, and if it didn't
work, to eliminate it. This method became organized, of course, into science.
\end{multicols}
\end{document}
http://www.mueller-public.de - My site http://www.anr-institute.com/ - Advanced Natural Research Institute
I use it for blog postings, with the tables and footnotes plugins. I also use it when converting simple text files into e-books with Calibre, which has a built-in markdown converter. I have nothing against html, but typing an open and close tag for everything gets to be tedious when you just want to write something simple, especially for tables.
I've had to edit other people's book-length Word documents before. These tend to be a mess, because many people don't know how to use styles, so the formatting is a big mess. I copied the text in Markdown, formatted the headings as H1 and H2, wrote some simple html for the embedded images, converted the document to HTML, and imported it back into Word. It sounds like a lot of steps, but I was able to do this faster than clicking on 200 pages of the Word document to fix all the inconsistencies. It goes without saying that the person got the document back, and then messed up the formatting again, because they didn't think Word styles were a real thing.
As I recall, Cory Doctorow at a book signing mentioned he used Markdown or something similar in his writings, and then kept the documents under a version control system to be able to see the changes. This is something Markdown should excel at, much better than a wysiwyg editor.
It is "AN author", not "A author".
Troll is not a replacement for I disagree.
Yes, definitely. I've found Markdown an excellent tool for writing. Pandoc (combined with Apple's TextUtil) has proven useful for converting older documents, to a certain degree, but I find I only use Markdown for serious writing. Am currently in the process of building a toolchain that converts Markdown writing into a variety of format, blogging about it here:
http://rubyredbricks.com/blog/2013/10/09/ruby-pub-part-1-toolchains/
and here:
http://rubyredbricks.com/blog/2013/10/02/ruby-and-hpub/
... if I knew where to look on /. for a guide to it. I don't even know how to quote a post.
It works. Has good basic functionality. Allows good review markup. And importantly everyone knows how to use it. Once the document is written I print it as a PDF and link to that PDF on a wiki for sharing. Burn me now?
Avoiding "Computationally expensive" stuff isn't the reason for using Latex.
When I use Latex, it's because I need to make a complex document with footnotes etc. and I want to use a real (i.e. programmable) text editor (GNU Emacs in my case).
LibreOffice has its uses, but you can't grep .odt files.
(Actually, instead of "LaTeX quality wysiwyg", which sounds like you want software other than Latex, but which gives equal quality, you probably meant to say "wysiwyg LaTeX". If that's the case, I agree. But since I'd written my comment before I spotted this, I'm posting it anyway.)
Expert in software patents or patent law? Contribute to the ESP wiki!
LyX rules. All the advantages of LaTeX, without having to know LaTeX. The simplicity of normal word processing. So no "compile errors" when printing, for example.
Articles, 30-page reports and even a book - lyx handles it all well. At least none of the problem word users around me complain about. No "screwups" when documents go above a certain size. No sudden changes/resets of styles. Cross references are always right. Even in special cases, like having a file with a chapter included into two different reports - so that the chapter and page numbers for the stuff turn out different in each of the two documents.
LyX+beamer for presentations. Works well, and I can make animations using Tikz too. Well, those animations are perhaps not for beginners.
LyX makes any kind of output that LaTeX can make. Usually PDF, as anybody is capable of receiving PDF. But also html when needed.
If you use one tool for everything, you're lost in the amateur world.
Professionals in every profession on this planet have a vast array of tools and pick the right tool for the right job. It's one of the things that makes you a professional.
I personally use WYSIWYG editors for general text editing like letters, articles and some books. Stuff like Pages or iBooks Author (since I'm on a Mac).
For longer books and fiction, Scrivener is my tool of choice and since it can export epub, it was my tool of choice for the book I published last year. As an author's tool, it's fantastic and not at least knowing it (even if you find it's not to your taste) is like being a programmer and not knowing what an IDE is.
Speaking of which, I abhor IDEs for programming because all I've ever tested are sluggish and more confusing than worthwhile, and they never work the way I want, and I pity everyone who uses Eclipse. So for programming, HTML and such, Sublime Text 2 is my choice, though I've used TextMate for years and it was fantastic. For Subl, there is better plugin support, however, and its auto-completion and coding tools do almost everything that IDEs do, at least for the stuff I code (which is mostly web-programming, so I don't need a compiler and debugger built-in).
There's also LaTeX and LyX, which I haven't touched in a long time but I still love and I'm sure I'll find my way back to it eventually. There's also a bundle of other text-based tools I use for specific parts, like graphviz (which really shines if you use the GUI tool to get an instant-preview). Basically, to end where I started: I always try to look at the job and find the best tool for it. Swiss army knifes are cute if you're trecking and can't bring much stuff, but when you're on your computer and can bring whatever tools you want, bring the best ones, not the cute one.
Assorted stuff I do sometimes: Lemuria.org
Confluence by preference
Word or openoffice for docs
Vi or Notepadd++ for text
Depends on what you're doing. You can't use grep, specifically, but you can search by regular expression in LibreOffice.
That's one way to slice it, but not the only way. I think publishing has an unhealthy relationship with LaTeX. Markup languages have come a long ways since 1980. Why are we so stuck on this one? Another language that is (1) more human readable (2) easier to machine parse (3) renders to equal or better quality (4) is wysiwyg friendly, should be quite possible.
I won't join Slashcott. OTOH, If Beta goes live, I just won't be back until it's fixed. Sorry Dice.
At one point I found myself composing some email (started with notes for a conference, then code-heavy email) in Github's editor, rendering, and then pasting into an email. And then I realized that I'm a programmer and can make stuff to do stuff. So I made Markdown Here (or the project page on Github). It's a Chrome/Firefox/Safari/Opera/Thunderbird/Postbox extension that allows you to write email in your normal email editor, but in Markdown, and then render it before sending. It also supports syntax highlighting and some TeX math support. And it works in Google Groups posts, Evernote (web interface) notes, Blogger posts, and so on -- anywhere that uses contentEditable or designMode for the editor (including TinyMCE-based and CKEditor-based ones).
So, yeah, I use Markdown for email. There are about 15,000 Markdown Here users, so I guess they're part of the answer too.
The idea that a text be complete is stupid!
We write text, then assemble article, then bind book, then collection series.....
No document should have a fixed type and be complete!
Please stop the school of structure documents.
Notepad for text.
OpenOffice for markup & publishing.
Since I do write a lot of legal stuff (I thought I'd packed that racket in, but no, it won't let me...), I want/need something that loads in 0 seconds or close as dammit, doesn't have ANY bells and whistles to distract me, that can keep up with my typing speed (even OOo has problems sometimes, what with the app competing for CPU cycles with everything else, most importantly the keyboard buffer!), occasionally (particularly during meetings) I'll kick up a DOS session and use EDIT.
Political debates have me rolling my eyes so much I think I got optical whiplash. I should sue. - Foamy The Squirrel
I write fiction, so my focus needs to be on the content rather than presentation, and I envision events in the story, not how the literal text appears. I had to use a non-WYSIWYG word processor called AppleWorks 2.0 on an Apple IIc long ago as a kid, and dealing with the codes was quite distracting; upgrading to WYSIWYG MultiScribe aka BeagleWrite, where I could pay attention just to the story, was a massive improvement.
Also, when it comes to fiction, the publisher is likely to control the appearance, and then many users will use Calibre to reformat it yet again to match how *they* prefer texts to look. Bye-bye hours of tinkering with margins.
If I was writing an article that involved mathematic equations or similar things where positioning is crucial, then it might make sense to use markup/markdown, as making sure everything is aligned properly and all that. It's just not a terribly reasonable approach to use if you're trying to immerse yourself in a novel.
Now mostly at Usenet:comp.misc & SoylentNews.org (it's made of people!)
/usr/bin/vi, what else? Not vim, not elvis. No color, no auto-format, no hype.
Precisely. I write mostly in TeX (real TeX, not LaTeX) but use LibreOffice for some documents. One of the major reasons that I continue to use TeX is that I can edit it as I please, which generally means using GNU Emacs. LibreOffice and its kin are so much inferior as editors that I feel crippled when I have to use them. Indeed, when writing something that is mostly text, even if it is ultimately going to be formatted in LibreOffice, I almost always write it as plain text in Emacs, then import the plain text file into LibreOffice. TeX is also easier to edit using other Unix text tools and to generate programmatically. Perhaps surprisingly, I have also found it much easier to produce documents with a lot of images in TeX (using the psfig macros and dvips/dvipdf) than in LibreOffice. LibreOffice tends to get sluggish and to give me a hard time getting the images scaled the way I want them.
Inventing new tag names isn't the right way. Use the script element with the right mimetype: <script type="text/markdown"> and some javascript to render it.
If you want wysiwyg markdown: on Windows, check out texts.io. On linux, Uberwiter. On Mac... there are probably lots of options besides texts.
If you write fiction, or any book-length texts (a Ph. D. thesis or academic papers, frex), you owe it to yourself to check out Scrivener. It's available for OS X, Windows, and in beta for Linux. Closed software and charged for, but worth it.
If I'm writing a paper, I use LaTeX. Yes, the macros are a pain, but I find it takes less time than writing a paper in, say, Word (page breaks, sections, image placement, etc. need only be written once in LaTeX, but in Word they need to be revisited at each draft); and LaTeX's output quality meets my standards (while Word's, say, doesn't).
If I'm writing documentation, I use Markdown. It's simple and it has links. The output quality is far lower, because I expect readers to prefer reading plain HTML anyway -- a PDF would be inconvenient for them, even if it would be prettier.
Generalizing those notions:
1. The best tool for the job is the one that lets you produce and edit content as quickly as possible, while meeting your requirements. In other words: if monospace, left-justified text files are satisfactory, you should probably be using plaintext. When starting a document, first pick a set of features, then choose the tool that has those features and gives you the fastest workflow.
2. Distributing in multiple formats shouldn't be a concern: you can convert pretty much any open format to any other. Your _master_ copy needs to encode all the features you use.
Pandoc has certainly implemented a lot of what is needed for scientific writing, but one of the main things holding it back now is the lack of ability to attach labels to figures, tables etc. and then autonumber these items (ideally, with different numbering schemes), e.g. .... the data displayed in Table {mytable} and Figure {myfigure} show that ..
LaTeX, of course, does this already, and can be embedded in the Pandoc text, but I'd prefer to restrict the necessary LaTeX-isms to mathematical equations - such things tend not to work very well when converting Pandoc to anything other than LaTeX.
As someone that writes a great deal both for online and offline distribution, I use markdown *extensively*.
It's fabulous for the grunt work of formatting: headers, italics, links. The rest can be done by tossing in HTML, XML, or whatever other markup code is needed. It's fabulously lightweight and fast and unobtrusive.
In fact, for all those "I wrote my dissertation in LaTeX, *sniff* *sniff*" people here on Slashdot, how about this:
I wrote my dissertation on an iPad 2, in Daedalus, in markdown, embedding HTML or other kinds of markup as necessary, then formatting it all in a final pass through a couple different parser/formatters. Sometimes the right tool for the job is the one that you have to think about the least—the one that stays out of your way—and for me, that's markdown+Daedalus.
(Yes, I'm prepared for the onslaught of accusers, ridiculers, and doubters here—prepared to ignore them.)
STOP . AMERICA . NOW
I use markdown + Deadalus for long-form (i.e. dozens of pages) content that will go to print.
I use markdown + Mou for short-form content that will go online.
If you write for online distribution at all and you don't know about Mou, you should definitely check it out. Similar statement for those that write books but don't know about Daedalus.
STOP . AMERICA . NOW
...jEdit or gVim for entry, pandoc for conversion, Firefox for viewing in HTML. I take notes in markdown during tech meetings. It"s scary-fast to create good looking notes.
How do I use Libre Office's regular expression searching to find out which of these 200 documents have the word 'slashdot' in them? Oh did I tell you they are in a tree of folders?
That's a one-liner in the shell with plain text files. A slightly longer one-liner will fire up emacs (or other editor) with all the matching files opened ready for editing. find, grep, xargs...
Write once in XML using domain-specific vocabulary and markup; convert to HTML/PDF/plaintext using XSL.
Simples.
Definitely AsciiDoc. It's awesome. As easy to use as Markdown, more powerful and scalable all the way to serious publishing. O'Reilly books now use it as a source format too. Converts to DocBook amongst other formats, so you get all the benefits of that whole toolchain too.
all these people who use latex to write documents, must be same type of people who go home and birch themselves.
As a computer Science graduate, then post graduate, then wed developer and now support geek I have NEVER ever known or heard of anyone using actually using latex. Not one single person in my degree class used it, nor anyone I have ever worked with.
I tried using it a bunch of times because I'd heard it was the way to go, but then I realised how much time I'd have to invest learning the language before I could even begin to start writing it was just crazy.
Any Word processor, even M$ Word and all it's idiot hidden obtuse formatting crap is a better solution then learning a completely new programming language just to write a document or report and good luck sending your carefully formatted Latex document around the office or via email to "regular people". They'll do what everyone else does and try to open it in Microsoft Word and then ask you to resend it again in Word format for them.
I know three professional writers. 1 technical author and 2 published novelists. Guess what they use? The same thing that everyone else does: A standard word processing product ala M$ Word. (or Scrivener)
I also have a good friend working in for a UK publishing company. She uses Word 90% of the time and Adobe Indesign for typesetting and printer friendly layout.
There may have been a time when Tex/Latex was worth investing 100s of hours in learning how to use, but that time has long since gone.
I use what 99% of the rest of the world uses: A Word Processor.
Microsoft Word (if you learn to how to use Styles correctly) Will let you create pretty decent documents and very complex layouts and enable you to export to .doc, .docx, .rtf, .txt, pdf and I'm sure (though haven't checked) there'll be an add-in to export to .mobi/epub (although I know that if you want to self publish on the major ebook websites - you can submit your manuscript in Word format and have their publishing backend tools do the conversion for you automatically)
Personally, as a Linux only user, I don't use Microsoft Office(and always preferred Wordperfect's "ShowCodes" feature to the obscure self-formatting decisions that Word seems to employ anyway). LIbreOffice Writer will do the job, but still misses out on several features available to Microsoft Word. .odt, doc and epub (v2 and possibly some of v3 IIRC)
Calligra Author will save output in
Kingsoft WPS Office Writer: is the closest Linux alternative to Microsoft Word. Works very, still in development and missing VBA/Macros and font embedding (the latter also missing from LIbreoffice). WPS doesn't support as many formats as LibreOffice, but in my opinion is far superior for general document/report/authoring than Libreoffice, especially for long documents and/or complex layouts.
There's always Google Docs Writer that you can use and then make use of the myriad of free online conversion tools to create whatever document format you want.
Lastly, Scrivener is available natively for all platforms, Linux, Windows and Mac and is just about the most perfect and excellent writing tool in existence (it's not free for Windows/Mac, free while in Linux beta though) but by far the best writing tool of them all.(IMO)
Markdown sounds like a good idea, as does ReStructuredText but when you just want to write, you don't want to have to go and learn yet another markup language (even a simple one). You just want to get on an write.
Mediawiki to update the ShapeOko wiki
LaTeX for page formatting / layout / pdf manipulation
LyX for general writing (but export to LaTeX to customize things)
InDesign at work.
I did use Scribus when updating the documentation of the ShapeOko hobby-level CNC milling machine though, and was disappointed that there wasn't an easy / obvious way to get the text from the wiki into Scribus --- couldn't there be a command in Scribus to open a web page as the basis of a document or a page?
I have added some ``engines'' which support pandoc to TeXshop though, but haven't gone beyond that.
Sphinx of black quartz, judge my vow.
oh! hey! thanks. that's a good tip. I'll still keep the same approach in webnomad for non-js compat, but well I'm convinced we don't need new tag names.
Another language that is (1) more human readable (2) easier to machine parse (3) renders to equal or better quality (4) is wysiwyg friendly, should be quite possible.
3 and 4 can be contradictory, as being wysiwyg friendly adds the requirement of looking nice when text is added or removed. Text jumping around when a single character is added is not a problem when using a non-wysiwyg editor, but it would be a big problem in the user friendliness of a wysiwyg editor. In other words, the set of algorithms applicable in wysiwyg editors is a subset of the algorithms applicable in non-wysiwyg editors.
I work in a Formula 1 team and I implemented a script that generates automatic reports for the CFD simulations using those tools.
You actually could use grep, but since an odt (and ods) file is basically a zip archive, you'd first have to unzip/extract the content.xml file from the archive first and then grep that file for the text you're after.
For those doing scientific programming, the IPython notebook is a joyful place for interactive exploration and can be appropriate for document creation. Notebook cells can have code, images, or text, and text can mix Markdown and LaTeX (rendered in the cell via MathJax). Notebooks can be converted to HTML or PDF (via LaTeX), using the nbconvert utility (which depends on pandoc). For serious document production, this is not even remotely a replacement for LaTeX, but it can be a great place for interactive work.
The wiki community had their chance to standardize on a markup language but instead we have an atrocious assortment of wildly disparate syntaxes with Mediawiki unfortunately leading the charge... Now we need another one?
Ok; I'm old. I grew up using LaTeX for everything. Even generated invoices automatically from an ascii file of client/project/hours direct to printer using a Makefile. Now I don't 'write' anymore. I edit with vim and if I need to publish to html, I wrap the file with ... I simply don't care about formatting anymore. Content is where it's at.
The original question doesn't really ask about LaTeX, but everyone here is talking about that. So, is markdown used around or not?
I do use Markdown, it's convenient for quick formatting. I am not going to use LaTeX just for a few simple bold/italics needs...
LibreOffice has its uses, but you can't grep .odt files.
If you rename an ODT file as a ZIP file, you can extract the contents, which are mostly plain text XML files. Those XML files, on the other hand, may contain complex data (raster images, and such) encoded as Base64 objects.
In fact, I started using LyX back in... 1997 or so?
Not only it is used and looks like a WYSIWYG editor, but actually frees your mind from actually caring how it will render on a page of a given size. Just write what you mean (they call it WYSIWYM — M for Mean), and when previewed/printed it will be beautiful. Why? Because it is LaTeX doing it.
I write my personal stuff in reStructuredText, which is like Markdown but with more processing possibilities and more extensible.
When it comes to professional stuff, depending on the size of the project, I use reStructuredText or DocBook.
DocBook is a must when complex stuff (like cross-references, recursive file inclusion...) is involved, and XML processing enables many fancy features, and output to many formats.
reStructuredText documents can be converted to DocBook, so it's possible to use a master DocBook document and rst fragments.
rst is definitely a plus for straight reading, collaborative edition and SCM.
DocBook is not as painful as binary files, but XML edition, diffs and merges is not as trivial as text chunks.
Then I use LaTeX for stuff with too many formulas, or stuff that really needs to look pretty.
But I dislike that there is no real separation between structure and looks of the document.
Also, LaTeX is harder to process and slower to compile.
DocBook or reStructuredText toolchains can process chunks of math code in LaTeX syntax, which is a nice tradeoff.
In all cases, writing non-trivial documents is easier when a good build system is used. ...) between documents, when using this kind of text-based tools.
What I like most is the ability to share external resources (chunks of text, generated data, images, generated images,
I started it when I was writing lots of text on an Alphasmart stand-alone keyboard, almost the best tool for first-draft writing. But really bad for editing. So I'd upload raw text to a file, for further processing on a real conputer. I wrote a program to do preliminary cleanup, making it more like HTML. Emacs after that.
Yes, that was over a decade ago. The program is written in Modula 3.
I discovered that the notation I was using on the alphasmart was more convenient than raw HTML, so I continued to use the ad-hoc notation even after I migrated to a laptop, and slowly changed it and the program according to taste. The program now generates .fodt files, which I do not edit.
My notation has no log-term syntactic nesting constraints (as HTML does with its and tags) so it is a natural for use in a revision management system. (Merges preserving tree structure are notoriously hard to do correctly; i.e., yielding valid tree structure after).
I'm considering changing to markdown. A project I'm involved i has chosen Asciidoc instead.
I consider ease of use with a revision management system (I use monotone) to be crucial.
The main feature I've found to make this easy is for the markup to use separators instead of brackets whenever possible. Thus use a mark to separate paragraphs rather than two to enclose them. Maybe there are a few things that can't be anaged this way, but for the most part the big things can be.
-- hendrik
But the last time I cheked, none of the markdown-like systems offered that in a reasonable way. Semantic markup means that I do not simply make some text bold or italic, but I give the reason why it should be somehow formatted differently than the surrounding text. Reasons could be that it should be emphasized, it introduces a new term, or it is a foreign phrase (such as "a priori"). All this is usually set in italics, but sometimes it makes sense to change this for certain categories. Also, I guess, alternative output devices (such as audio or Braille devices) could benefit from this, although I am not aware of any practical details.
copy con filename.txt that's only on my fast pc, the slower ones I flip switches.
Reading the Ars Technica review it seems that LG desperately wanted it to be a copy of the Samsung Galaxy S4.
Personally I don't like Samsung (or probably this LG phone) because they change too much from the stock Android version. But I am happy with my Sony Xperia Z1. Very happy in fact, even though there's still some software problems (got update last night though, hopefully it's even better now).
/ The Arrow
"How lovely you are. So lovely in my straightjacket..." - Nny
Although I have never used Python and I program in C++ 100% of time, if I were to rule the world I would issue an ultimatum to all programming languages to conform to Python syntax within 5 years after which all backward compatibility will be dropped! Python offers a terse human readable syntax. Now for a verbose machine readable syntax which is still very human friendly, I would choose XML. So again, Latex gets five years to confom to Python (terse) or XML (verbose) or both flavors. The idea is that we can keep on complicating things in the pretext of simplifying them. All we need is Python and XML syntax, syntax highlighint text editors (Vim for shell and Geany for GUI), PNG format for images and bzip2 to package all the stuff in a single file ready to be dispatched to a printer or a projector via a rendering application.
Really? The last time I used cards was due to a class requirement. I was taking a class that had a programming assignment and the professor actually wanted us to turn in card decks. So I sent a copy of the program to my virtual punch and sent a message to the operator asking him to send the file in my punch to a physical card punch. His response: SERIOUSLY? And this was back in my FORT-G and FORT-H days. I've never heard of anyone coming anywhere near an 029 keypunch in the F77 era. In fact, I think it was the quarter after I took that class when they finally ripped out the keypunches and the vending machines where students got blank cards.
CUR ALLOC 20195.....5804M
I think the grandparent post was talking about doing a search of files within a directory (or tree) outside the editing utility. What lunatic is going to manually load each file in a directory and repeat a search? (``Well, that string's not in this document... let's close this file and try the next one....'')
CUR ALLOC 20195.....5804M
You don't know that... and please pay attention:
"Depends on what you're doing."
You don't think I thought of that? I don't know what he wants to do, but I assumed the possibility that he might not know that LibreOffice search supports RE searches, and that it might possibly be useful to him. I don't know that it will be, but I also don't know that it won't be. It depends on what he's doing.
I won't join Slashcott. OTOH, If Beta goes live, I just won't be back until it's fixed. Sorry Dice.
My mistake.