From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mga14.intel.com ([143.182.124.37]) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1Ubcea-0002or-Dq for openembedded-core@lists.openembedded.org; Sun, 12 May 2013 22:14:02 +0200 Received: from azsmga001.ch.intel.com ([10.2.17.19]) by azsmga102.ch.intel.com with ESMTP; 12 May 2013 12:55:46 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.87,657,1363158000"; d="scan'208";a="301361136" Received: from unknown (HELO helios.localnet) ([10.255.12.83]) by azsmga001.ch.intel.com with ESMTP; 12 May 2013 12:55:45 -0700 From: Paul Eggleton To: openembedded-core@lists.openembedded.org Date: Sun, 12 May 2013 20:55:43 +0100 Message-ID: <3529451.hQA8zCVn47@helios> Organization: Intel Corporation User-Agent: KMail/4.10.2 (Linux/3.8.0-19-generic; KDE/4.10.2; i686; ; ) In-Reply-To: <1368384890.16243.79.camel@pb-ThinkPad-R50e> References: <1368189198-1294-1-git-send-email-jukka.rissanen@linux.intel.com> <518FB7F9.2020102@linux.intel.com> <1368384890.16243.79.camel@pb-ThinkPad-R50e> MIME-Version: 1.0 Cc: Otavio Salvador Subject: Re: [PATCH 0/5] Enable VPN support in ConnMan X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 12 May 2013 20:14:02 -0000 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="us-ascii" On Sunday 12 May 2013 19:54:50 Phil Blundell wrote: > On Sun, 2013-05-12 at 08:40 -0700, Saul Wold wrote: > > On 05/12/2013 06:27 AM, Otavio Salvador wrote: > > > I think so, it'd be good to have it in oe-core and allow use of vpn :) > > > > I would like to see what the full dependency set looks like for these, > > clearly there is the vpnc, openvpn, l2tp and pptp recipes, but what else > > and what licenses are they under. > > I don't think we necessarily want openvpn, l2tpd and suchlike in > oe-core. None of those things seem very "core" to me (in an embedded > context) and testing them seems like it would be a bit of a challenge. I agree, these don't belong in OE-Core. We already have them in meta- networking. > Equally, we certainly don't want to have dependencies in oe-core > pointing to packages in meta-oe or any other layer, since this would > make it impossible to test oe-core in isolation. So I would be inclined > to say that the right way to deal with this is for those connman bits to > go in a .bbappend which lives in the same layer as the recipes in > question. That doesn't work well for software layers - it is not a good thing for various recipes to get rebuilt just because you add meta-oe to your configuration for example. The protocol we've established is to add PACKAGECONFIG options to enable the dependencies but have them disabled by default; these can be enabled as desired in distro or local configuration when the layer satisfying the dependency is also enabled. Cheers, Paul -- Paul Eggleton Intel Open Source Technology Centre