From owner-freebsd-gnome@FreeBSD.ORG Thu Oct 20 10:39:49 2011 Return-Path: Delivered-To: freebsd-gnome@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 1B8F51065670 for ; Thu, 20 Oct 2011 10:39:49 +0000 (UTC) (envelope-from rfg@tristatelogic.com) Received: from outgoing.tristatelogic.com (segfault.tristatelogic.com [69.62.255.118]) by mx1.freebsd.org (Postfix) with ESMTP id CF5B88FC0C for ; Thu, 20 Oct 2011 10:39:47 +0000 (UTC) Received: from segfault-nmh-helo.tristatelogic.com (localhost [127.0.0.1]) by segfault.tristatelogic.com (Postfix) with ESMTP id 22F06BDC34 for ; Thu, 20 Oct 2011 03:39:47 -0700 (PDT) To: freebsd-gnome@freebsd.org In-Reply-To: <4E9F9C2B.40808@FreeBSD.org> Date: Thu, 20 Oct 2011 03:39:47 -0700 Message-ID: <4361.1319107187@tristatelogic.com> From: "Ronald F. Guilmette" Subject: Re: Build problem - help wanted X-BeenThere: freebsd-gnome@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: GNOME for FreeBSD -- porting and maintaining List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 20 Oct 2011 10:39:49 -0000 In message <4E9F9C2B.40808@FreeBSD.org>, Doug Barton wrote: >On 10/19/2011 20:13, Ronald F. Guilmette wrote: >> Still, to whoever the maintainer of gtk is... Shouldn't you, ya know, fix up >> the port dependencies? You know, so that gtk depends on having >> gobject-introspection-0.10.x installed? > >I don't know enough about the specifics to answer that question. However >in general the FreeBSD ports in general don't support partial updates. >The assumption is that all of your ports are up to date. We do our best >to support as much backwards compatibility as we can, but you're going >to have less trouble generally if you keep everything up to date. Thank you for your response, but it is clear that you dont' understand how the problem I reported came about. So please allow me to explain. I am just simply trying to bring up a fresh new 8.2-RELEASE system and trying to get a bunch of ports I need to have installed on it installed. To do this, I have run either portinstall or portupgrade on something like 30-40 different ports of things I need. And no, I actually _did not_ watch carefully every line of output as these things were all being built and installed, so I have no idea how the OLDER version of gobject-introspection (0.9.12) even came to be installed on this system AT ALL. I can assure you however that *I* never explicitly ASKED for the old version of that specific thing to be installed. In fact, up until I had to try to solve the gtk build problem, I did not explicitly do ANYTHING that obviously would have led to the build and/or installation of ANY version of the gobject-introspection thingy. In fact, even now I have no idea what the bleedin' thing even is, and almost zero desire to know. I just wanted to build & install ImageMagick, gthumb, an some other tools I use frequently. It would seem that one of those, or perhaps something I installed from pre- built packages off the distribution CD stealthily installed a stale and unusable version of this gobject-introspection thing when I wasn't looking... rather like some naughty child. (Sigh.) In short, it is all well and good for you to admonish me to keep all of my installed packages up-to-date[1], but in this case, the failure of the GTK build, and specifically the rather entirely unhelpful and uninformative way that it died (a) was not in any sense my fault and (b) was, I think, more a symptom of bad engineering in the port that it was a symptom of any person or group of people not having run around and made sure that everything was spic and span and ready for military inspection prior to even trying to build the port. If I was the only person on the planet who had ever run into this gtk build problem, or the only one who had ever been baffled by it, then I'd say that you might have a point, and that maybe I fell down in my responsibilities to do everything letter perfect before even starting the gtk build. But googling around for some of the same cryptic failure messages that I got reveals that I'm far from the only person who has been bitten by this cock-up. So, you know, I'll just say again what I already said... Maybe the gtk port maintainer could do the world a favor and create a proper and complete list of dependenceis for the gtk port. If one engineer (i.e. the gtk maintainer) can spend ten minutes and fix this problem so that fifty or a hundred other people don't stub their toe on it and each waste a hour searching for the proper work-around, then isn't it a more efficient use of human manpower generally to have the one gtk maintainer fix the problem? You wrote: >The assumption is that all of your ports are up to date. What's that they always say about ass-uming? If this port needs 0.10.x of something else to build... well... isn't that why ports have dependency lists? If not for cases exactly like this, then why even have such things as dependences in the ports system? Regards, rfg -=-=-=-=-=-= [1] There is a certain chicken-and-egg dilemma about your casual admonish- ment that I should keep all my ports up-to-date. I was actually TRYING to update a port... the gtk port... when this cryptic build failure occured. How exactly does one ``keep all of one's ports always up-to-date'' if the process of updating ports itself fails?