Return-Path: Received: from mail1.windriver.com (mail1.windriver.com [147.11.146.13]) by mail.openembedded.org (Postfix) with ESMTP id 113B173232 for ; Thu, 10 Sep 2015 23:48:11 +0000 (UTC) Received: from ALA-HCA.corp.ad.wrs.com (ala-hca.corp.ad.wrs.com [147.11.189.40]) by mail1.windriver.com (8.15.2/8.15.1) with ESMTPS id t8ANmBcX015013 (version=TLSv1 cipher=AES128-SHA bits=128 verify=FAIL) for ; Thu, 10 Sep 2015 16:48:11 -0700 (PDT) Received: from seebs-worktop (172.25.40.228) by ALA-HCA.corp.ad.wrs.com (147.11.189.40) with Microsoft SMTP Server id 14.3.235.1; Thu, 10 Sep 2015 16:48:11 -0700 Date: Thu, 10 Sep 2015 18:48:10 -0500 From: Peter Seebach To: "openembedded-core@lists.openembedded.org" Message-ID: <20150910184810.566531e7@seebs-worktop> X-Mailer: Claws Mail 3.9.3 (GTK+ 2.24.23; x86_64-pc-linux-gnu) MIME-Version: 1.0 Subject: opkg --install and "attempt-only" packages X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.12 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: Thu, 10 Sep 2015 23:48:12 -0000 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit So, I'm not 100% sure yet, but it appears to me that if you have a large list of "attempt-only" packages, and *any* of them can't be found at all, opkg bails out without doing anything about the rest of them. This seems a little surprising, because my assumption would be that the expected behavior for "attempt-only" packages would be to try to install as many as possible. Looking at other cases, it looks like the RpmPM and related code try to handle each package separately to build a list, but OpkgPM doesn't. I note also that RpmPM will allow an empty list of mandatory packages, while OpkgPM only accepts an empty list when attempt_only is True. This affects us because we have a setup in which we put a very large pool of packages in as attempt-only because we know that there will be small variances in package lists which are hard to anticipate, and the net result is that none get installed at all. -s -- Listen, get this. Nobody with a good compiler needs to be justified.