Speaking of the ss randomizing, if a math nerd comes around, please do fix my very lame way to fake a rng with two sticks a piece of yarn
.
For now, I'm looping on all the ss, and using the string length of the last 3 as a variable (with a static seed for two of those var on the first file), and then a very lame operation:
ss_frng_prev * ss_frng_pprev + ss_strlen - ss_frng_pprev * 7
(Where ss_frng_prev is the previous strlen, ss_frng_pprev the strlen even before the previous one, and ss_strlen the string lenght of the current file).
I also would like to avoid using a chunk a /dev/urandom fed through md5sum of something like that, and do it all with shell builtins, in order to avoid forking like crazy. (And, of course, the Kindle's shell doesn't support Bash's $RANDOM -_-").
Well, it's better than nothing, but it's far from perfect, so if you experience weird 'non-randomness', that's why ^^.
@yecam: That's what calibre does by default, btw. (unless I changed it somewhere along the ligne, but I don't think so
) (${kindle}/documents/${author_sort}/${title} or something like that, too lazy to fire up calibre to check :P. But it's configurable anyway
).