From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from 7of9.schinagl.nl ([88.159.158.68]:52721 "EHLO 7of9.schinagl.nl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751717Ab2EMR04 (ORCPT ); Sun, 13 May 2012 13:26:56 -0400 Message-ID: <4FAFEED8.7080903@schinagl.nl> Date: Sun, 13 May 2012 19:26:48 +0200 From: Oliver Schinagl MIME-Version: 1.0 To: Antti Palosaari CC: linux-media Subject: Re: AF9035 experimental header changes References: <4FAFD21D.9000801@schinagl.nl> <4FAFD888.8080801@iki.fi> In-Reply-To: <4FAFD888.8080801@iki.fi> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-media-owner@vger.kernel.org List-ID: On 13-05-12 17:51, Antti Palosaari wrote: > On 13.05.2012 18:24, Oliver Schinagl wrote: >> Hi antti, >> >> I've just updated my local branch of your experimental branch and got >> some conflicts because you moved the header inclusions from the C file >> to the header file. Why is that? I thought it was really bad practice to >> have includes in header files. >> >> http://git.linuxtv.org/anttip/media_tree.git/commitdiff/7d28d8226cffd1ad6e555b36f6f9855d8bba8645 >> > > Because "struct state" was inside af9013.h and I added "struct > af9033_config" to state. Due to that af9033_config visibility I was > forced to move af9013.h include to the af9015.h and moved all the > others too. > I see, I just learnedthat having includes in headers is just something really bad, possibly causing include loops etc etc. Its just something that's generally not needed and should be avoided. ideally anyway :) > What is problem you has met? Well it's just movement of those headers caused merges to conflict and fail :) > > DVB USB driver header file is something like private place for data > other than code as it is not included by any other driver. > > regards > Antti Oliver