Disc rot ate one of my Maxell rewritable compact discs. This is the second disc to fail me from the bunch of 10 they came in. They’re pretty old in terms of IT time, but I’ve had floppies that lasted much longer.
The year of the ‘to each his own’ desktop
A comment on Slashdot pretty well summarizes what is also my view of all this fuss about the year of the Linux desktop:
Users who don’t want to learn how the machine works already have two major systems designed specifically for them: Windows and OSX. To me it makes perfect sense that Linux would be Open Source because Microsoft and Apple both recognize that the real money is gained by appealing to the general public and the general public is nearly technophobic.
That being said, I would add that there’s nothing keeping a bold newcomer (such as Canonical) from breaking into the general market with a Linux desktop either, provided that their version is ‘good’ (i.e. dumbed-down IMO) enough. I’m just not too worried whether such development takes place or not.
A Linux desktop for the masses would obviously help in improving hardware support, which is why I do still see it as a Good Thing. But for me it’s not the be all, end all outcome that other, perhaps more competitive people than myself see it as.
Make some noise
I adapted Geek Thoughts‘ “Making Noise with ImageMagick” in creating video noise — a video simulating what you used to get when you turned on a television with no signal, back in the analog days. To combine the noise frames created by ImageMagick I used Netpbm, which I had to install separately (that is, it wasn’t installed by default in Lenny).
Here’s the Bash multi-liner I used to generate 48 frames worth of noise:
(for i in `seq 1 48`; \ do convert -size 640x480 xc: +noise Random \ -blur 1x1 -fx intensity -normalize ppm:-; done) \ | ppmtompeg kohina.txt
And here’s kohina.txt
which specifies how ppmtompeg packs the frames. Note that I just threw this together without too much thought, so it definitely needs tweaking depending on what the video is used for and so what (quality, compression, speed) matters. For example, I only used I-frames due to lack of imagination.
PATTERN IIIIIIII OUTPUT kohina.mpeg INPUT_DIR stdin BASE_FILE_FORMAT PPM INPUT_CONVERT * GOP_SIZE 8 SLICES_PER_FRAME 30 PIXEL HALF RANGE 2 PSEARCH_ALG LOGARITHMIC BSEARCH_ALG SIMPLE IQSCALE 1 PQSCALE 1 BQSCALE 1 REFERENCE_FRAME ORIGINAL
And here’s what it looks like, recompressed with Theora (using Oggconvert): kohina.ogg
(2,8 MB). Generating an accompanyng soundtrack is trivial with Audacity.
SSH inside a while read loop
Got bit by the fact that SSH inherits stdin from its parent, which rendered my fancy remote-command-in-a-loop useless. Passing -n to ssh made it work the way I’d expected.
WebDownloader for X settings
The app which calls itself “WebDownloader for X” comes in a package called d4x and saves its settings in ~/.ntrc_2
. How’s that for consistency!