On Wed, Oct 29, 2008 at 07:50:32PM +0000, Eric Kow wrote: > On Wed, Oct 29, 2008 at 10:30:12 -0700, Jason Dagit wrote: > > I want to see benchmarks too, but I thought I would justify why we > > expect this to be no slower than the previous code...Everything below > > is stuff that we discussed during the Sprint. > > Well, attached is a second set of comparative timing tests, sorry, only > run once and with no nice output yet. Hopefully you can use a graphical > diff tool to do side by side comparison. > > A nice little summariser script, maybe using the Haskell tabular > library might be handy
Unless you've named them wrong, it looks like salvo-9b has on the an effect of slowing darcs down, when it has any significant effect. That doesn't sound like a good optimization... It'd also be good to test against pre-salvo-8 with or without bytestring. I don't strongly object to requiring bytestring, but given that we *know* that it's not as efficient as it could be by at least 10% or 20% (unless this is a timings error or a misnamed-files issue), it does seem worth tracking down what went wrong while we've still got two similar versions of darcs to test against. I'll run some quick tests on my computer to see how things look here. Could you try doing a darcs check test? Darcs check isn't a common operation, but it's commonly a slow operation, so it's a pretty good canary for telling when certain operations are slowed down (i.e. reading patches and applying them). I notice that annotate is particularly slow with the new bytestring, which suggests that perhaps patch reading (either zlib interactions or maybe parsing) has been slowed down disproportionately). David _______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
