On Tue, 15 Aug 2006 19:51:14 +0300, Eero Tamminen wrote: > > http://cairographics.org/~cworth/gtk2.6_vs_2.10/ > > It would be nice if the details tables would have a row containing > also the number shown in the summary table (to see what the "1.05" > number in the summary actually was) and a note about whether > the summary number is median, average etc. I've added the average (it's a mean of the per-widget ratios, by the way) to the sub-tables, so hopefully that makes things more clear. I also added some details to the front page about CPU, X server version, torturer version, and the various library versions. I'm also now putting results into dated sub-directories for easier comparison of future runs, (though in the process I accidentally wiped out the results from the initial run). I have posted results from a subsequent run in which I doubled the number of iterations that the torturer runs. The new results can be seen here: http://cairographics.org/~cworth/gtk2.6_vs_2.10/20060815-1/ Before I deleted the old results, I did compare the numbers a bit and it seemed to me that there's a lot of noise in the "Expose" column. I do get the feeling that the "Resize" column is more reliable, (since it is measuring tests that run for much more significant amounts of time). And if that "Resize" column is reliable and an accurate measure of GTK+ performance in some sense, then it shows that GTK+ 2.10 is less than 10% slower than GTK+ 2.6, (at least on my system---which is an x86 box and obviously doesn't take any of the floating-point hit that people have seen on embedded platforms). I'll run a few more tests today and post those so we can get a feel for how stable these numbers are. -Carl
Attachment:
pgpbHtiwHwh4Q.pgp
Description: PGP signature