The Agony Of Word

· geek · mumblings · software ·

I really dislike Microsoft Word. It’s a necessary evil in academia, where .doc or .docx is the editable format of choice, but I avoid it wherever I can. It’s true that things have got slightly better since I upgraded to Office for Mac 2011. Office at least looks like a proper OS X application now, and some of the annoyances and downright bugs have been removed. But it makes things unnecessarily complicated and has a nasty habit of biting you just when you are most in need of having a fully-functioning word processor.

I have recently started to use plain text files written in markdown format wherever I can, using Pandoc (the wonderful “Swiss army knife” of text converters) to convert it to whatever format I need at the end. This has turned into a much smoother process since it has integrated citeproc-hs. Almost everything I write requires citations and a reference list at some point, so citeproc integration has made this much easier. All you need is a bibliography file in one of the supported formats (it accepts quite a good range, including BibTeX and EndNote format), and then you just put placeholders in the text which get converted to in-text citations. You can even specify a citation style via a .csl file if you like. It works very smoothly. Pandoc itself has some nice extensions to the markdown format so that you can specify the title, authors and date of a work, create footnotes easily, create tables and so on. If you’re comfortable in a text editor, and you are used to doing mighty battle with Word, it’s a breath of fresh air. The best bit is that you don’t even have to worry about what format you want the file to be in at the end: you just write and then decide later.

I had a paper to write for a journal, with a colleague as my co-author. Since I was going to be writing the first draft, I decided to do it in Pandoc, then convert the file to Word format (via .odt format) for my colleague at the end. Despite the fact that it was quite a difficult paper to write, and the fact that I was under severe time pressure because we had a deadline, I loved the process of writing in markdown in a text editor. For once, I was concentrating on what I wanted to say, rather than fiddling with heading styles and so on. I could write in MacVim, which has become second-nature to me now, and in which I am very comfortable. Every now and again I would preview it to see how it was looking, but mostly I just worked with the text.

At the end, I converted it, and the conversion went pretty smoothly. Then the pain began. After enjoying myself so much in Pandoc World, being plunged back into Word Purgatory was a shock to the system. I had been using the markdown2pdf command in Pandoc to convert directly to PDF (using LaTeX or XeTeX as an intermediary), and the result was gorgeously typeset. When I took a look at the same paper in Word, the full horror of Word’s so-called typesetting hit me: what had happened to my beautiful paper? I know that the journal doesn’t care about that (it’s what they expect), but I found it deeply dispiriting. The frantic hours I spent, just before the deadline, trying to wrangle the formatting into the required specification were absolutely maddening. The page numbers were the last straw.

Of course, it had all come down to the wire. I absolutely had to submit the paper online by midnight and at 11pm, I was trying to add page numbers in the footer. What could be easier, you might think? After all, it’s a common enough task. I found the command to add page numbers, and ended up with a bizarrely-formatted footer, in which the page number had a black border to the left of it. I didn’t want a black border, I just wanted a page number, plain and simple. I tried to somehow get rid of the border, and tried just about every editing control in the entire package, with no luck. OK, well, there must be other formats for page numbers, surely? I found the styles for footnotes and had my eyes seared by a variety of ghastly designs. Who has ever wanted their page number to be sitting in a blue coloured block, with (for good measure) and additional blue hairline above and below? And that was the least offensive of the styles. Somewhere deep within Word’s recesses, there may be an option to have a plain, unadorned number in the footer, but I did not manage to find it. Eventually, at 11:40pm, I managed to somehow coerce the page number into a sensible format. Unfortunately I don’t even remember how I did it, so I’m likely to face the same pain again next time.

I really wish I didn’t have to deal with Word. My lucky colleagues in Computer Sciences, Mathematics and Physics use LaTeX all the time, and that’s the expected format for sending work to journals. Unfortunately that’s not the case in Biosciences, where all the journals expect Word format. Even if some journals accepted LaTeX, most of my colleagues wouldn’t be able to use LaTeX, though I might be able to persuade them to use Pandoc. In dark times like the evening I have just described, I’m almost tempted to re-train as a computer scientist, just so that I don’t have to use Word again. Extreme, I know, but Word drives me to it.