From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marius Bakke Subject: Re: [PATCH 3/4] gnu: xapian: Use 'modify-phases'. Date: Mon, 17 Oct 2016 19:50:07 +0100 Message-ID: <87zim27qu8.fsf@duckhunt.i-did-not-set--mail-host-address--so-tickle-me> References: <20161017165506.17750-1-mbakke@fastmail.com> <20161017165506.17750-4-mbakke@fastmail.com> <20161017180015.GB31629@jasmine> Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:41832) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bwCzE-0000s0-Aw for guix-devel@gnu.org; Mon, 17 Oct 2016 14:50:13 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bwCzB-0004mT-3l for guix-devel@gnu.org; Mon, 17 Oct 2016 14:50:12 -0400 Received: from out1-smtp.messagingengine.com ([66.111.4.25]:59121) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1bwCzB-0004lq-0T for guix-devel@gnu.org; Mon, 17 Oct 2016 14:50:09 -0400 In-Reply-To: <20161017180015.GB31629@jasmine> List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org Sender: "Guix-devel" To: Leo Famulari Cc: guix-devel@gnu.org Leo Famulari writes: > On Mon, Oct 17, 2016 at 05:55:05PM +0100, Marius Bakke wrote: >> * gnu/packages/search.scm (xapian)[arguments]: Use 'modify-phases'. > >> + `(#:phases >> + (modify-phases %standard-phases >> + (add-after 'unpack 'patch-remotetcp-harness >> + (lambda _ >> + (substitute* "tests/harness/backendmanager_remotetcp.cc" >> + (("/bin/sh") (which "bash")))))))) > > Please change it to use (which "sh") instead. Bash changes its behavior > based on whether it is called as `bash` or `sh` (see Invocation in > bash(1)). I'm aware, but decided to remove it in the subsequent patch since it was "buggy" anyway, and is not needed when remote TCP tests are disabled. It was left to avoid justifying the fix when changing to modify-phases. The phase also does not end on a #t. But I agree, might as well just fix it and leave it even if the tests don't use it.