On Sunday 31 July 2011 21:40:35 Jon Seymour wrote: > On Mon, Aug 1, 2011 at 5:21 AM, Christian Couder > > <christian.couder@xxxxxxxxx> wrote: > > On Sun, Jul 31, 2011 at 1:55 PM, Jon Seymour <jon.seymour@xxxxxxxxx> wrote: > >> + if test -n "$BISECT_UPDATE_REF"; then > >> + eval="$eval echo '$BISECT_UPDATE_REF' > > >> '$GIT_DIR/BISECT_UPDATE_REF';" + fi > > > > I don't like this very much. In fact I realize that the eval thing we > > use is buggy because the result of 'eval "eval"' will be the result of > > the last command in "eval", so we won't detect if one of the first > > command in "eval" failed. > > How about I fix the eval stitching so that it uses && rather than ; to > connect each statement with a final true at the end so there is no > dangling &&? Yeah, please fix it in a separate bug fix patch. > > So perhaps something like: > > > > eval "$eval" && > > + ( test -n "$BISECT_UPDATE_REF" || echo "$BISECT_UPDATE_REF" > > > "$GIT_DIR/BISECT_UPDATE_REF" ) && > > echo "git bisect start$orig_args" >>"$GIT_DIR/BISECT_LOG" || exit > > Probably this would be better :-) > > ( test -z "$BISECT_UPDATE_REF" || echo "$BISECT_UPDATE_REF" > > "$GIT_DIR/BISECT_UPDATE_REF" ) && Yeah :-) Thanks, Christian. PS: sorry I replied to you only previously -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html