Hackerific

last update:

I’ve taken over maintainership of the TextMate Matlab bundle as of last week, after what seems to have been quite a long period without a maintainer. Hopefully, I can fix a few of the glitches (hopefully I already have), and I’ve got plans to add various bits of useful functionality. My current todo list includes: Updating the snippets to reflect TextMate conventions (work in progress). Adding more snippets for useful functions to help speed up writing code.

I’ve recently started developing a Gnuplot bundle for TextMate, and I’ve decided to host it on GitHub. Like a lot of people, I’ve recently started using Git instead of SVK, my previous choice of distributed VCS and SVN, so hosting on GitHub seemed a natural choice; plus how could I possibly ignore their cool network graphs?! Gnuplot is an incredibly versatile and complex piece of graphing software which runs on a lot of different platforms, and can output even more different formats.

Recently, I’ve been generating gazillions of sequences of pngs that I’ve saved from MATLAB. The filenames look like: interesting_image_<01--nn>.png What I really want is movies, but I really don’t like using MATLAB’s flakey movie generation code, it produces low quality output with ginormous files, and is really slow. This is where and Quicktime Pro comes in. By allowing you to open sequences of images, and then save them as movies, it’s half way to a nice solution, but it still involves too many clicks, and tends to over-compress my nice clean pngs.

A few days ago, on one of my regular del.icio.us/popular procrastination research trawls, I discovered DTerm, a HUD style, context-sensitive, drop down command line thingemy. It’s pretty cool, with useful features like ‘insert selected items’, and ‘copy results’ and after a couple of updates I can see it being totally great. Here’s a quick screenshot: I’d recommend taking at look at their site for a more complete screencast with some useful ideas.

I used to be a fairly heavily Octave user, but when I started my PhD I had to use MATLAB for some specific toolboxes. I found this quite annoying, because I think MATLAB is big, and slow, and clunky (and leaks memory). So I’m quite happy to report that I think that this might be about to change. On December 21 2007, Octave 3 was released. Now, I’m not sure if this is relevant to this post, but it did prompt me to upgrade, and since I was playing with some MEX (MATLAB external interface) files, I decided to see if I could make them work with Octave.