Re: Rhythmbox Predictive Playback - Weekly Report

["Charlotte Curtis" <c f curtis gmail com>]

Week 8, or thereabouts...

I finally, with the help of a variety of people, came up with a
package that can be installed on *gasp* someone else' computer.  I
should really test it out on my laptop too, but just in case it is
functional, it can be acquired from
http://code.google.com/p/rhythmbox-predictive-playback/downloads/list.
 However, it also requires the latest release of Marsyas, which can be
found at http://sourceforge.net/project/showfiles.php?group_id=84982&package_id=96585&release_id=612499.
 I really wish this project weren't so time-consuming to get started
(takes anywhere from half an hour to several hours to do the initial
analysis, depending on the size of your Rhythmbox library and the
speed of your computer), but if anyone is interested I'd love to hear
feedback.

This week I will be working on testing my plugin on my much slower
laptop, and doing whatever optimization is required.  I will also make
some UI changes - right now I've got a popup window, which I heard is
bad news, and I don't have any notification of the running process.
Finally, I realized that I have a major flaw in how my program works:
I use Python's subprocess.Popen module to call an executable, but I
pass as an argument the list of songs that are currently in the
playlist.  Obviously, the maximum length of this argument is limited.
I haven't yet hit that limit in my testing, but I'm sure it's there.
My options for getting around this seem to be:
a) Rewrite the plugin in C++
b) Change the executable to one that reads in arguments while its
running and then use subprocess.communicate to pass it the values
c) Write the list to a file and read it from the executable

I'd love to be able to do (a), but I'm not quite sure that I have the
time.  (b) would probably be the second-best solution, because (c)
would be an annoying amount of reading/writing.  On the other hand, it
may not be as much of a problem as I imagine, so maybe it's okay
as-is.

I may not have anything to report in three weeks time - On Wednesday
July 30th, I am going on a trip back home (from Toronto to Saltspring
Island, BC), which is pretty exciting as I haven't been home since
Christmas.  I return on August 4th, and then on August 5th I leave for
a conference in Ann Arbor, MI to give a talk on my research from last
semester (creation of a finite element model of a human wrist).  I
won't be returning until August 10th, which is pretty close to the
"suggested" pencils down week, so perhaps I won't be able to heed that
suggestion (especially since I technically will not be allowed to work
on the project, since I'll be in the U.S.).

That's all for now,

Charlotte