On Thu, Apr 15, 2010 at 16:19:38 +0200, Guillaume Hoffmann wrote: > Is there a recipe somewhere to generate the charts that are on the other > pages?
Max is working on integrating his chart-generation approach into darcs-benchmark using the hs-gchart library. He had some Python code which screen scrapes the wiki stuff (or maybe reads the RST files) but I don't think it understands our new annotations '?' and '~' (for < 5 and < 20 runs respectively). Note that if you're using HEAD darcs-benchmark then your benchmarks are saved in ~/.darcs-benchmark (I think) and that subsequent runs will accumulate the results. This is just us acting on an assumption that more data is better data. Somebody with more time could look into integrating criterion into darcs-benchmark so that we have a better idea how much data is "enough" data. -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9
signature.asc
Description: Digital signature
_______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
