Re: PRAGMA page_size;

So, tldr: sqlite gets the default right usually. Thanks for digging into this :) Daniel (FWIW: the bacground here is a 1.6 merge which was spending most of its time reading an entries file) Philip Martin wrote on Thu, Feb 16, 2012 at 11:01:33 +: > Julian Foad writes: > > > I (Julian Foad)

Re: PRAGMA page_size;

Julian Foad writes: > I (Julian Foad) wrote: > >> This thread seems to have a useful discussion: >> , which >> seems to be saying that a larger default seemed to be helpful for >> Mozilla on Windows. > > Ah, I suppose the most interesting th

Re: PRAGMA page_size;

I (Julian Foad) wrote: > This thread seems to have a useful discussion: > , which seems to > be saying that a larger default seemed to be helpful for Mozilla on Windows. Ah, I suppose the most interesting thing right now is to experiment o

Re: PRAGMA page_size;

Daniel Shahafwrote: > Apparently changing the PAGE_SIZE pragma has a significant effect on the > wc speed: > > 65536 [...] 29 s >  4096 [...] 20 s >  1024 [...] 19 / 16 s > > Now, obviously just using the 65k (the maximum) for the pragma isn't > always a good idea; but should we make this configu

PRAGMA page_size;

http://www.sqlite.org/pragma.html#pragma_page_size Apparently changing the PAGE_SIZE pragma has a significant effect on the wc speed: % rm -rf d && time ( export _SVN_PRAGMA='PRAGMA page_size = 65536;'; echo $svn; $svn co -q file:///tmp/svn/r1 d && $svn pl -R d >/de