From owner-freebsd-bugs Tue Jun 20 05:55:38 1995 Return-Path: bugs-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id FAA18581 for bugs-outgoing; Tue, 20 Jun 1995 05:55:38 -0700 Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id FAA18575 for ; Tue, 20 Jun 1995 05:55:30 -0700 Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id WAA10482; Tue, 20 Jun 1995 22:52:46 +1000 Date: Tue, 20 Jun 1995 22:52:46 +1000 From: Bruce Evans Message-Id: <199506201252.WAA10482@godzilla.zeta.org.au> To: bde@zeta.org.au, rgrimes@gndrsh.aac.dev.com Subject: Re: bogus release version string Cc: asami@cs.berkeley.edu, bugs@freebsd.org Sender: bugs-owner@freebsd.org Precedence: bulk >> `cvs co ...; date >some-header.h' :-) >> >> I'd prefer for `cvs commit' to print the date step commit the changed >> header (not recursively, and without doubly the number of commits). >cvs co occured once for /usr/src on freefall since the 2.0 tree, it has >never been checked out again. >some-header.h is not how the version string gets built, it is needed >by /sys/conf/newvers.sh. Use #include. >This would not handle the 3 way branch we now have in cvs. Each branch should have a different newvers.sh. The same timestamp will do for them all. This gives more timestamps than kernel versions, but this isn't a problem. >We need to think this throught carefully and come up with somethings >that works well, but does not require excessive version reving of a >commited file. Some committed file has to change for the change to be visible in cvs. Revving the file once per day might be good enough. Bruce