On Mon, 22 Jan 2007, Gene Heskett wrote:
On Monday 22 January 2007 15:44, Todd Zullinger wrote:
Gene Heskett wrote:
I think the problem for others is that it's hard to tell exactly what
you've done and what you've changed in between these builds. Your
mails haven't been exactly clear, concise, step-by-step descriptions
of how to reproduce a massive compile time decrease by removing
SELinux from your kernel config.
I'm sure you can appreciate that to properly test and verify something
it is best to make the absolute minimal amount of changes in the
system between tests so that other factors can be ruled out.
The idea is to provide proof and the ability for others to verify.
There should be no reason for a timing test to come down to a matter
of belief. This isn't rocket science, but it's not church either. :)
As I said in a previous posting, the only change between those two builds
was ripping out the Livna versions of mplayer and mplayerplug-in and
installing the Dries versions. And disabling selinux. Now which do you
think might have had the major effect?
Without seeing (and diff'ing) the two configs and being able to duplicate
the tests ourselves, who knows?
Steve Friedman
