gtkperf - I'm maintaining - feature requests/new tests?

[gQuigs]

I've started adding a couple fixes to gtkperf (including code from the pts test to let it run from the cli).
So far the only big test I've changed is I've made "add text" much faster and also forced it to actually update the screen. Previously, on some machines it takes a full hour during which the screen never changes.

Current changes are here https://gitorious.org/gtkperf-fixes/...commits/master

I would eventually hope to get it included in PTS.

Thanks for your suggestions,
Bryan

[kozaki]

Hi gQuigs, thank you for maintaining gtkperf and for your openness to pts user wishes
My first need would be to make the various gtkperf tests results understandable for the non ?ber geek.
i.e. have been looking into OpenBenchmarking database for tests that'd let me compare the 2D capabilities of present AMD and Intel APUs (for heavy GUI tasks without any 3D). And it seems I lack a Master in computational science to make any decent use of it ;-)

Sorry if this is off-topic

[gQuigs]

There is really much less to it than you might think.. (at least gtkperf). Each test does a specific usually simple operation inside GTK.

For example the spin increment test, this is pretty much the core of it:
gtk_spin_button_spin (GTK_SPIN_BUTTON (appdata->spinbutton_entrys),
GTK_SPIN_STEP_FORWARD, 1);

or increment by 1, whenever idle.

http://gtkperf.sourceforge.net - goes into more of the reasons for creating it in the first place. Check out the example tests there as well.