I’ve just pushed the latest release, 2.6.0, cutesy release code name “A veiled threat!”, to Github. As always you can file issues here. And if you’re interested in text / document mining, please join me for @jonathanstray’s Document Mining with Overview tutorial. If you miss the live webinar tomorrow, it will be recorded and you’ll be able to replay it any time.
What’s New?
- Cutesy release code name: “A veiled threat!”
- There is now a minimalist version of the workbench, which I call the “Base.” The base consists of the chosen Linux desktop, the command-line version of R and the GGobi data visualization system. Both R and GGobi are installed from the Linux distribution. This means that in principle any Linux distribution with both R and GGobi can be used, although I am still only supporting Fedora 18, Linux Mint 14 and Ubuntu 12.10.
For the impatient, there is now a three-step process to get started:- Install a supported Linux desktop.
- Download and unpack the install scripts.
- Run the ‘install-base.bash’ script.
- All the R packages installed in R-platform and the options are installed for all users in a site library. See Managing Libraries for the details on how this works.
- Shiny Server is now included in the Node.js option.
- I’ve added more PDF tools to the ScrapingTools option.
Road Map
- Next cutesy release code name: “Why, is one missing?”
- Riak is on hold for a bit. My highest priority right now is getting some test applications done using Shiny Server.
- Those who have been following my data and computational journalism projects over the past few years know that one of my long-term goals is for the tools to be available on Windows desktops / laptops. While I personally prefer Linux for numerous reasons, Windows is much more readily available in newsrooms.
This week, I discovered a project that brings me much closer to that goal. It’s an R package called installr. On a Windows R installation, ‘installr’ will update R to the latest stable version and install numerous other packages, all from the R console. This includes the MikTeX environment, RStudio Desktop, git and quite a few other tools. With ‘installr’ and RStudio, a Windows machine can become a first-class open source computational journalism platform just like a Linux desktop. Here’s the Github URL.
So I am looking at ways to integrate ‘installr’ into a spin-off of the workbench for Windows users. I’m not sure yet whether this will be a separate project or part of the Computational Journalism Publishers Workbench.