On second (third?) thought... On Thu, Jul 3, 2014 at 5:56 PM, Brian Norris <[email protected]> wrote: > On Thu, Jul 3, 2014 at 5:49 PM, Jeremy Kerr <[email protected]> wrote: > ... >>> @@ -39,7 +45,8 @@ set_patch_state() >>> update_patches() >>> { >>> local cnt; cnt=0 >>> - for rev in $(git rev-list --no-merges --reverse ${1}..${2}); do >>> + for rev in $(git rev-parse --not ${EXCLUDE} | >>> + git rev-list --stdin --no-merges --reverse ${1}..${2}); do > > I think I can probably combine the rev-parse and rev-list to avoid the > extra pipe: > > git rev-list --no-merges --reverse ${1}..${2} --not ${EXCLUDE} > > I'll test and resend, if that looks OK.
Actually, I think the rev-parse serves as a little extra parameter checking. Unless someone can pin a good reason one way or the other, I'll stick with my original patch as I sent it. And I've been testing that version for a few weeks now. Brian _______________________________________________ Patchwork mailing list [email protected] https://lists.ozlabs.org/listinfo/patchwork
