paperclips and pipe cleaners
Sunday, November 12 2006
The Ubuntu installation on my laptop Linux machine had become a junky mix of Ubuntu and several flavors of Debian. Some things, particularly the Java environment, had become irreparably broken. I'd been unable to install a working copy of the Arduino IDE, which I'd hope to do, since I have hopes of this laptop being the controller for all my various microcontroller-based household electronic projects. So today I just blew everything away and started from scratch with a fresh new Debian installation. It felt good, like a crap that leaves your asshole pleased for hours.
Unfortunately, though, I couldn't get my old SMC wireless card to work for lack of that mojo they put into Ubuntu (and with a name like Ubuntu, it has to be mojo that they put in). This turns out to be a smaller issue than I'd originally thought; with the Arduino connected to it the laptop is going to have to be wired down anyway, so being forced to use an ethernet cable won't ruin my plans.
Once I had the laptop working acceptably, I returned to the problem of logging solar thermal data coming up the USB cable from the Arduino-based solar sufficiency controller in the basement. When I'd last left this issue, I was redirecting the output of the Unix cu command-line terminal program into a Perl script that was datestamping each line of data and then sending it on to a log file. This was working fine when started from a tty session, though it did tend to take over that particular session. Starting the process as the laptop booted, however, proved to be impossible. The script would fire and then take over the computer, refusing to allow whatever followed in the boot process to get its day in the sun. But there was a solution to this problem. In Unix there's always a solution, although sometimes you have to fashion it out of the software equivalents of paperclips and pipe cleaners. The solution came in the form of a handy Unix program called screen, which can launch virtual tty sessions.
Once I was logging data reliably, it was time to figure out how to display and interpret it. I didn't really make any progress on that front today, thought I did write a PHP script to parse the data in the solar data log file and write it into fields of a freshly-installed MySQL database. Once there it's pretty much helpless against my years of data mining and presentation experience. But this was enough for one day.
For linking purposes this article's URL is:feedback
previous | next