Performance of a binaries vs source compliation!
I have been doing some research lately about the performance difference
between binary packages, and those compiled from source. From what I have
learned so far their really is not a good tool to verify the difference.
I keep hearing about how important a source package is over a binary. Some
Gentoo fan boys swear by it. But I want to know about facts.
If I compile my kernel and use binary packages such as we do here with debian.
How much of a performance hit will I take?
Has anyone done a source vs binary package performance test comparison?
How much faster would a XFree86 source compiled be against a binary of the
same package under similar compile flags? Could the average user tell the
difference? How stable would the combination be?
I know that their is always a trade off, between speed and stability, a lot of
people can compile with speed optimizations with less stable results. The
Gentoo forums are full of discussions of unstable choices used in
compilation.
What I would really like to see is a comparison of basic packages done with
speed and stability as a base to see what holds up.
Is their such a beast, or has someone done some basic research to see what the
difference is?
I would like to see a basic Debian package vs Gentoo Emerge package on the
same system comparision.
Anyway hope I haven't gone two much off topic, just curious.
Rthoreau
I like top posting, that is what headers are for!
Reply to: