Le Sam 3 Juin 2006 12:40, Matthew Chapman a écrit : > Pierre, > > The existing write_error() function prints an error, removes the > output file and exits with an error code. So I believe that my patch > is equivalent to what you are suggesting (except you have some > unreachable code :)).
ooh, sorry, I misunderstood the aim of your patch. I missed the fact that copy_stat was able to remove the output file, which seems so absurd that I never imagined it did that. your patch is obviously the good one, now that I've looked it in detail. sorry for the stupid comment. Bdale: will you have time to deal with it ? I can NMU it if you want. Cheers, -- ·O· Pierre Habouzit ··O [EMAIL PROTECTED] OOO http://www.madism.org
pgpkQmyEbynhz.pgp
Description: PGP signature
