through
http://www.cs.wisc.edu/~gthain/bentley/1000.gif
You might want to look at these URLs by hand. Your task is to run an image sharpening program on each of these images, and report the file size of each one before and after. This should be done in a reasonably efficient manner, but in a way that can be reproducible by anyone following in your footsteps.
Programmatically, each file may be downloaded with the curl command like so:
curl http://pages.cs.wisc.edu/~gthain/bentley/0001.gif > 0001.gif
The transform we would like can be performed by the "convert" command which is located in ~gthain/convert on osg-ss-submit like so:
OMP_THREAD_LIMIT=1 convert 1.gif -unsharp 0x1000 1_us.gif
Note that convert is not installed on any of the execute machines.
The final output should be a report of the sizes for each of the original execute files and their unsharpened relatives. There is more than one way to break up this work, so we'd also like to be able to discuss the design decisions you've made along the way.