From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:43521) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WGWDL-0007IP-LV for qemu-devel@nongnu.org; Thu, 20 Feb 2014 11:11:13 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WGWDF-0000Pa-Hy for qemu-devel@nongnu.org; Thu, 20 Feb 2014 11:11:07 -0500 Received: from mx1.redhat.com ([209.132.183.28]:43938) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WGWDF-0000PN-9S for qemu-devel@nongnu.org; Thu, 20 Feb 2014 11:11:01 -0500 Message-ID: <53062910.3030009@redhat.com> Date: Thu, 20 Feb 2014 17:10:56 +0100 From: Paolo Bonzini MIME-Version: 1.0 References: <1392827321-10205-1-git-send-email-roland@kernel.org> <5304E4E4.10703@redhat.com> <5304F3D4.9060501@suse.de> <53052B92.7080305@redhat.com> <530620C7.7060208@redhat.com> <5306270E.4020105@suse.de> In-Reply-To: <5306270E.4020105@suse.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH] scsi-disk: Add support for port WWN and index descriptors in VPD page 83h List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?ISO-8859-1?Q?Andreas_F=E4rber?= Cc: Roland Dreier , qemu-devel@nongnu.org Il 20/02/2014 17:02, Andreas F=E4rber ha scritto: > Not unexpected, it's the older; the - convention was introduced possibl= y > with QOM around start of 2012. Or at least there it's been enforced, an= d > by my understanding of QOM and QMP visibility it then applies to device= s > as well. > > Regarding QMP, I consider it smarter to do the _ -> - matching at > QemuOpts level than somewhere inside QOM. That's fine, because device_add uses QemuOpts internally. > For -cpu we have such compatibility code (although non-QemuOpts) in > target-i386/cpu.c, converting all underscores. Unfortunately that won't > work as long as there are underscores in old properties. Maybe you have > some cool patch idea? Well, no cool idea except "once conversion is done at the QemuOpts=20 level, do a full sweep of s/_/-/". This should be done before we'll be=20 able to create devices with object-add. I think as long as we have a=20 plan, consistency trumps design-by-committee convention. Paolo