Irreal

The ever awesome Sacha Chua has an excellent Baby Steps Guide to Managing Your Tasks with Org. It’s still a work in progress but even in its unfinished state it’s a really useful guide for getting started with Org mode.

One of problems with Org is that it can be overwhelming. Org can do so much and has so many facets that it’s easy for beginners to bog down trying to understand everything. Chua’s guide takes the approach of learning to do just one thing with Org: manage your TODO lists. By the time you’re finished mastering that, you will have learned enough about Org and how it works that it will be easy to move on to the next set of Org features.

Chua starts off by having you create a list of tasks and shows how you can manipulate that list by moving tasks up and down in the file. When you’re done she says to change the TODO to a DONE. That’s kind of a bore, though, so she shows you how to toggle it with a simple key sequence. But maybe you want more states than just TODO and DONE. You might want WAITING or SOMEDAY, for example. Chua shows you how to add these and assign speed keys to them so you can switch states easily.

Next she suggests that you break that list of tasks up into projects. That makes it a bit easier to organize your tasks and track progress on individual projects. Finally, she shows you how to schedule tasks for a certain date.

If you’ve been wanting to get started with Org, Chua’s tutorial is an almost painless way of getting started and learning to use Org for a useful task at the same time. Definitely recommended.

Some of the most enjoyable programming I do is metaprogramming: writing code that writes code. Lisp macros are all about that, of course, but the idea is much more general. Often times the code that the code writes is just a simple shell script. Several years ago when my main workstation was a Linux machine, the ripperX program was broken for one reason or another so I wrote a program to read the CD directory and generate a shell script to rip the tracks. It was nice because I had complete control over the format of the names assigned to the tracks.

I just came across a nifty example of this sort of thing over at Sacha Chua’s blog. Chua was working on cutting the video stream from the recent Emacs Conference 2015 into individual talks¹. The idea is to note the start time and duration of each talk and use that in a call to ffmpeg. You can write those ffmpeg calls by hand, of course, but that’s boring and error prone.

Instead, Chua captured the talk title, start time, and end time in an Org table and used the table spread sheet functionality to calculate the duration. Then she used a code block in the Org file to output the calls to ffmpeg automatically. A nice example of metaprogramming using Org mode. It’s a great post and well worth a read. Perhaps it will give you some ideas for your own work. In any event, we’re all the beneficiaries of Chua’s work because now we can watch those conference talks. See Chua’s post for the details.

Ben Maughan has posted the fourth episode in his Org Mode Basics series. This time he talks about formatting the text in Org buffers and adding executable source code blocks to your Org file.

The Org markup language is very similar (in spirit) to Markdown but is a bit richer. If you’re not already familiar with Org markup, take a look at Maughan’s post to see the basic markup and what the results look like. As Maughan says, you can also include blocks of LaTeX right in your document if you need to. Small pieces of LaTeX can merely be added to the file without any special markup other than the LaTeX.

Although Maughan hasn’t mentioned it yet, you can also add HTML for those files you plan on exporting to HTML. Maughan will be discussing exporting Org files in a subsequent post.

As I said before, if you’re new to Org Mode, these posts are an excellent way to get up to speed. They’re short, easily digestible pieces that add to what’s come before. Even if you’re a long time Org user they’re worth reading; you may learn something new.

Christophe Pouzat has posted the slides from his talk Seven Years of Reproducible Research: From R / Sweave to Org. As the title suggests, the talk covers Pouzat’s evolution from using R and Sweave for reproducible research to using Org mode and Babel.

It’s clear that Org/Babel is the better solution for most researchers. Instead of being restricted to just R and LaTeX, Org/Babel lets you use a simple markup language that can be exported to many different formats. Instead of being restricted to R, you can use whatever language is most appropriate and even mix several languages in a single Org file.

If you’re a researcher interested in reproducible research—and if you’re not, you should be—Org/Babel offers an easy way to get started and practice reproducible research on a continuing basis. Of course, you will have to learn Emacs but, the wailing from hipsters notwithstanding, a basic fluency with Emacs requires an investment of an hour or so. Once you start using it, you’ll find yourself learning more and more of its features. The time savings you’ll earn from Org/Babel for writing up your research—even if you don’t use reproducible research—will more than repay the modest investment in effort required to get started.

Spend three or four minutes reading Pouzat’s slides and see if you don’t agree. It really will make your life better.