Hi Phil, The point I am making is that you need to reboot the machine between version runs, to equalize matters such as disk block caching. If you run 2.19.16 and then 2.19.18 the latter may be using cached data from the former, such as shared DLL’s already loaded and so on.
Now that you have my attention on this topic, on my Linux system I build lilypond from source. When compiling with the standard optimisation flag -02, my current score in progress compiles in 7.0s +/- 0.1 s. Out of interest, I recompiled with -03 (yes I know everybody is scared of this) and I get very reliably 6.9s +/- 0.1s consistently, over many runs. I would conclude the extra optimisation wins as extra 0.1 second. Hence, this is observable, but practically speaking, not enough to matter. I also observer that lilypond is single threaded and can only take advantage of one CPU, so there is nothing to be gained from have more cores. And re caching algorithms, when I first run this score I get ~12-15s, which on subsequent runs drops to 7s. This is normal UNIX behaviour. Even Microsoft publicly admitted Vista was a mistake, technically and user interface wise, and repudiated it. Any reason to be still on it? Andrew On 18 April 2015 at 08:52:44, Trevor Daniels (t.dani...@treda.co.uk) wrote: Phil Holmes wrote Thursday, April 16, 2015 3:43 PM Well, I remain mystified, but I can confirm the speedup on my Windows Vista laptop with a 4-page score: With 2.19.16 this took 25.2 22.6 22.9 secs With 2.19.18 this took 12.6 12.5 11.8 secs
_______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
