lubuntu-desktop team mailing list archive
Mailing list archive
Re: Benchmarking of Lubuntu themes
On 02/17/2011 07:31 AM, 神癒礁湖 (Rafael Laguna) wrote:
> My apologies about being so "explicitly scientific".
It's *great* to be scientific about testing! But that does usually
involve some work, not just pointing to a 3 year old blog.
> But, don't you think that Phoronix suite is a bit heavey just for
> testing GTK themes? Which isn't really its purpose (I don't really
> need to know the speed of my chipset, just working over it with
> different themes), I'd preffer to do that with a tools specifically
> designed for that, that uses Cairo for rendering controls and other
> widgets, like GTKPerf does.
I'm not convinced that GTKPerf is really the right tool, or that it
measures the kids of operations that users commonly use in Lubuntu, in
the proportion that they generally use them. But I am not a GUI expert
at all! If you think GTKPerf *is* the appropriate tool, that's fine, I
defer to your knowledge in the area of themes... please go ahead and use
*that* to do some repeatable and documented testing of themes in LUbuntu
in 2011, and document both your method and you results, and then we'll
have something useful to work with :)
If you feel the PTS lacks a suitable test for theme performance, and
that theme performance is something it will be useful to measure, then
my suggestion would be that you develop such test, and get it added to
the Phoronix Test Suite, so the whole community benefits from it! As I
am sure you know, there is no need to run the whole PTS, we can select
just one or a few tests that reflect what it is that we want to measure.