2011/4/9 Victor Eijkhout <vic...@eijkhout.net>:
> The other day I had the brilliant idea of writing some python code to
> generate the lilypond input for exercises. After all, generating the lilypond
> code for all triads, all scales, whatnot, is pretty simple.
>
> So now I have this exercise study that generates close to 2000 scales like
> "\relative c { a b c d e f g a g f e d c b a}". And "finding the ideal number
> of pages" takes an ungodly amount of time. Minutes. Probably over 10.
>
> Which makes me wonder if pagination is quadratic (I really hope it's not
> exponential!) and if so, if it should be.
Not for pagination, but rather for total time, I propose you the
experiment to determine this:
If you are in a command promt and have the 'time' command, or you are
using Frescobaldi which shows elapsed time of rendering, could you
tell us the times in seconds for 32, 64, 128, 256, 512 and 1024
scales?. If your system memory is big enough, you'll hopefully be
able to complete the 1024 stage.
Now, if you have gnuplot you can do a simple graph with these data; if
not, I'll do it for you. With a logarithmic scale on both axes, look
at the type of curve shown.
--
Francisco Vila. Badajoz (Spain)
www.paconet.org , www.csmbadajoz.com
_______________________________________________
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
