* Charles Wilson wrote on Sun, Feb 14, 2010 at 03:36:57PM CET: > Charles Wilson wrote: > > Charles Wilson wrote: > >> Charles Wilson wrote: > >>> This one, I think is OK for pre-2.2.8 -- what do you guys think? > >>> OK to push? > >> In response to review comments over here: > >> "Re: [PATCH] Enable runtime cwrapper debugging; add tests" > >> http://lists.gnu.org/archive/html/libtool-patches/2009-12/msg00014.html > >> > >> I've created a followon patch to this one which allows the cwrapper > >> tests to pass on platforms which don't use the cwrapper but instaed use > >> the shell wrapper (e.g. linux). > > > > ping... > > > > ping again.
The option parsing in the original patch is overkill -- no need to re-quote things if all you're going to do is remove a couple of entries from "$@", that can be done with set x "$@" shift shift type handling. The reference to _AC_INIT_PREPARE is not needed. Did you consider that the program we're wrapping might have argument structure like --some-option some-arbitrary-argument-to-this-option and that the arbitrary argument might reasonably start with --lt-? Just sayin. The followon patch adds even more bloat for $LINENO which I don't understand what you're guarding against, and who this is trying to help. I know you deserve a better review, but I've been AFK and the 72 hours are almost over. Cheers, Ralf