At 06:47 AM 12/2/2002, Jeff Trawick wrote: >[EMAIL PROTECTED] writes: > >> wsanchez 2002/11/29 03:05:59 >> >> Modified: . Tag: APACHE_2_0_BRANCH CHANGES acinclude.m4 >> buildconf configure.in >> build Tag: APACHE_2_0_BRANCH binbuild.sh >> modules/aaa Tag: APACHE_2_0_BRANCH config.m4 >> modules/filters Tag: APACHE_2_0_BRANCH config.m4 >> modules/mappers Tag: APACHE_2_0_BRANCH config9.m4 >> server Tag: APACHE_2_0_BRANCH .cvsignore Makefile.in >> Added: build Tag: APACHE_2_0_BRANCH PrintPath config.guess >> config.sub >> Log: >> Back-port changes to allow build without an in-tree apr and apr-util. > >Why does the stable tree need this?
Because apache-2.0 is a long-lived version, and this will help builders for many, many months to come... >Why can't we wait more than a few hours to merge such extensive >changes into the stable tree? We can and should... >At the moment it looks like we're at an impasse on whether or not to >review before merging back to stable, but the very least people can do >when it is not a trivial change, particularly for build issues, is >wait a while to to see what blows up. Agreed. It's good to discover problems on the sandbox before we move them into stable. A couple (or three) days is probably good to let most folks' autobuilds chew on cvs HEAD and discover new problems. So I like the philosophy of Fred's patches for 2.0 as well, but it would be good to iron them out in the sandbox before they are backported. Bill
