On Sat, 7 Aug 2010 11:06:55 -0300 Gustavo Sverzut Barbieri <barbi...@profusion.mobi> said:
> On Fri, Aug 6, 2010 at 9:19 PM, Enlightenment SVN > <no-re...@enlightenment.org> wrote: > > Log: > > stick in some output giving: > > > > 1. date + time you ran (or finished) the test > > 2. version of evas used > > 3. profile requested > > 4. actual window resolution you got > > 5. loop count used > > 6. engine used > > 7. fullscreen flag requested > > since you're at it, why not change benchmark from runs to time? While > some runs go very fast on my PC I doubt it does not contain lots of > errors, some are very, very slow on embedded that people often doesn't > run them. If we keep 30s for each test, we'll know beforehand how long > it will take and effects should be better. because each frame is not identical. things move. that means each frame is a different cost to render. that means if you change the number of frames rendered - you changed the nature of the workload, and thus are no longer comparing the exact same work done. benchmark is no longer accurate. it may now be completely bollocks. -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) ras...@rasterman.com ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
