From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from lists.gentoo.org (pigeon.gentoo.org [208.92.234.80]) by finch.gentoo.org (Postfix) with ESMTP id C176B13828B for ; Fri, 27 May 2016 18:24:04 +0000 (UTC) Received: from pigeon.gentoo.org (localhost [127.0.0.1]) by pigeon.gentoo.org (Postfix) with SMTP id 0C7F814225; Fri, 27 May 2016 18:23:47 +0000 (UTC) Received: from smtp.gentoo.org (smtp.gentoo.org [140.211.166.183]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by pigeon.gentoo.org (Postfix) with ESMTPS id 85315141E3 for ; Fri, 27 May 2016 18:23:45 +0000 (UTC) Received: from [192.168.2.63] (85.253.84.66.cable.starman.ee [85.253.84.66]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) (Authenticated sender: leio) by smtp.gentoo.org (Postfix) with ESMTPSA id D3CA9340CB4 for ; Fri, 27 May 2016 18:23:43 +0000 (UTC) Message-ID: <1464373418.13834.79.camel@gentoo.org> Subject: Re: [gentoo-dev] [RFC] gtk/gtk2/gtk3 USE flag situation From: Mart Raudsepp To: gentoo-dev@lists.gentoo.org Date: Fri, 27 May 2016 21:23:38 +0300 In-Reply-To: <1e38b9eb-268b-30ca-d29e-060ffdb1db97@gentoo.org> References: <1464358866.13834.59.camel@gentoo.org> <1e38b9eb-268b-30ca-d29e-060ffdb1db97@gentoo.org> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.18.3 Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Gentoo Linux mail X-BeenThere: gentoo-dev@lists.gentoo.org Reply-to: gentoo-dev@lists.gentoo.org Mime-Version: 1.0 Content-Transfer-Encoding: 8bit X-Archives-Salt: f7d89ecd-74f7-41a2-9c19-1fd7071c0ea4 X-Archives-Hash: 560dc33412421b038a69202902b2e9e0 Ühel kenal päeval, R, 27.05.2016 kell 13:14, kirjutas Anthony G. Basile: > On 5/27/16 12:59 PM, rindeal wrote: > > On 27 May 2016 at 18:54, landis blackwell > m> wrote: > > > I stopped reading after you reminded me it was 2016 > > > > Good to know, thanks for stopping by. > > > > Yeah the "its year" meme has been making its rounds of the > internet. > > anyhow, my 2017 question is about avahi.  right now i have USE=gtk > and > gtk3, where gtk really means gtk2.  i'm not going to change that > because > it fits QA's specs.  but i could remove it altogether and just drop > gtk2 > support for the next release.  good idea?  bad idea?  i guess i'm > asking > whats the status of gtk2 in gentoo seeing as its dead upstream. I don't see a strong reason to have gtk2 support if there are no gtk2 consumers remaining in the tree. I somehow doubt that's the case for avahi, though? If gtk2 support is removed though, then per gnome policy gtk3 component should come with USE=gtk and per QA policy USE=gtk3. The QA policy is not finalized and completely contradicts our side of things, hence discussions are needed, but did not conclude. This is a thread to continue this discussion, I suppose. I wouldn't change anything till then really, especially for libraries, where it's mostly handled by USE depends anyway. One thing is USE flag naming for gtk2 and gtk3, another thing is if apps should provide support on building against either. I strongly disagree with using the same flag names for "provide gtk2 linking higher level library" and "Build this GUI application against gtk version 2". QA proposed policy has this regression. Also for libraries it is a "either or both" situation, for apps it's a "one or the other" situation. This gets awful very quick if any application maintainer decides to express it with a REQUIRED_USE="^^ ( gtk2 gtk3 )", combined with the libraries in tree that then would have REQUIRED_USE="|| ( gtk2 gtk3 )" + libraries in tree where gtk component is optional, and if available both gtk2 and gtk3 versions are available with REQUIRED_USE="gtk? ( || ( gtk2 gtk3 ))". This would be clean with only libraries using gtk2/gtk3 as is GNOME teams current policy. Then the natural thing would be to have a different USE flag to mean to prefer gtk2 for user facing applications when possible instead of gtk3 (USE="force-gtk2" being suggested here for that) With this approach we can cleanly handle any upcoming gtk4 as well, while naturally moving users who don't care into using the latest version. I also strongly believe that USE flags should be first and foremost about features, not an expression of the name of an external dependency. Gilles did a huge start of work on mapping gtk* use flag usages in a spreadsheet, but given the volume, got distracted before finished. I pointed this out to the QA team to look at, but have not heard anything back. The point of the exercise was, that it turns out half of the tree seem to mis-use USE=gtk in some way, and QA should really concentrate on helping with that first. Then the situation would be much clearer, as there wouldn't be all that many USE=gtk's remaining. Anyhow, this is where the discussions stalled last time around. I think we should first figure out about the USE flag usages (naming + library only or not). We can bikeshed if gtk2 version for applications should be provided more liberally in addition to gtk3 as a one or the other choice later on. I don't feel too strongly about making that hard if it doesn't step on our library USE flag namespace. Mart