thanks runtest 1.0rc 1.0 -1
2009/10/19 Dan McGee <[email protected]> > On Sun, Oct 18, 2009 at 10:17 PM, mcu GPS <[email protected]> wrote: > > usage: vercmp <ver1> <ver2> > > > > return values: > > < 0 : if ver1 < ver2 > > 0 : if ver1 == ver2 > > > 0 : if ver1 > ver2 > > [a...@myarchtp mkarchlive] $ vercmp 0.1 0.1_rc1 > > -1 > > [a...@myarchtp mkarchlive] $ vercmp 0.1_rc1 0.1 > > 1 > > [a...@myarchtp mkarchlive] $ vercmp 0.1_rc1 0.1_r1 > > 1 > > [a...@myarchtp mkarchlive] $ vercmp 0.1_r1 0.1_rc1 > > -1 > > > > > > 0.1_rc1 > 0.1 ? > > 0.1_rc1 > 0.1_r1 ? > > > > this is a bug? > > No, it isn't a bug. For every example and corner case you try to fix, > you will break two more. The rules are quite simple- split on dots > ".", and then compare numerically if possible, followed by > alphabetically. We don't deal with shitty versioning schemes because > there is no universal way to do so. > > See pactest/vercmptest.sh for some examples of what vercmp does support. > > -Dan > >
