droundy: > On Wed, Feb 06, 2008 at 04:42:02PM +0000, Eric Kow wrote: > > So, I haven't had a chance to look into the freezing function yet, but > > I did do an informal performance comparison using the Data.ByteString > > version and the --disable-bytestring version. > > > > In short, the bytestring upgrade helps... a little bit. > > > > Doing a local get of the GHC repository, best of three trials, takes 2 > > minutes and 15 seconds (--disable-bytestring) and 1 minute 57 seconds > > (Data.ByteString). This is a 13% improvement in my calculations > > (18/135s), which is not bad considering how little it cost us to > > implement. > > Sounds great! > > Also worth testing is > > time darcs pull -a && time darcs obliterate --last 100 -a > > which you should repeat a few times (particularly as the first pull is > likely to have no effect). Also you can try this with --last 1000 to get a > more vigorous test. > > > Maybe we can egg on the bytestring people and get them to submit > > patches taking this further down (for example, they could improve our > > between newlines and nth newline stuff). > > That'd be nice.
If you can get darcs working with bytestrings, I'm happy to write bytestring-optimised variants of things you need, to squeeze out more performance. -- Don _______________________________________________ darcs-devel mailing list darcs-devel@darcs.net http://lists.osuosl.org/mailman/listinfo/darcs-devel
