From owner-freebsd-ports@FreeBSD.ORG Wed Nov 29 15:56:17 2006 Return-Path: X-Original-To: ports@freebsd.org Delivered-To: freebsd-ports@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id A7DFC16A4A0; Wed, 29 Nov 2006 15:56:17 +0000 (UTC) (envelope-from mi+kde@aldan.algebra.com) Received: from aldan.algebra.com (aldan.algebra.com [216.254.65.224]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9AF2143DFA; Wed, 29 Nov 2006 15:52:23 +0000 (GMT) (envelope-from mi+kde@aldan.algebra.com) Received: from aldan.algebra.com (aldan [127.0.0.1]) by aldan.algebra.com (8.13.8/8.13.7) with ESMTP id kATFqO9d001267 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 29 Nov 2006 10:52:24 -0500 (EST) (envelope-from mi+kde@aldan.algebra.com) Received: from localhost (localhost [[UNIX: localhost]]) by aldan.algebra.com (8.13.8/8.13.7/Submit) id kATFqNTj001266; Wed, 29 Nov 2006 10:52:23 -0500 (EST) (envelope-from mi+kde@aldan.algebra.com) From: Mikhail Teterin To: Lowell Gilbert Date: Wed, 29 Nov 2006 10:52:22 -0500 User-Agent: KMail/1.9.5 References: <20061127222709.aa35ab22.nork@FreeBSD.org> <200611290957.26508@aldan> <44irgyjlyf.fsf@be-well.ilk.org> In-Reply-To: <44irgyjlyf.fsf@be-well.ilk.org> X-Face: %UW#n0|w>ydeGt/b@1-.UFP=K^~-:0f#O:D7whJ5G_<5143Bb3kOIs9XpX+"V+~$adGP:J|SLieM31VIhqXeLBli" Cc: ports@freebsd.org, Norikatsu Shigemura Subject: Re: ports/devel/icu: PTHREAD_LIBS clean X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 29 Nov 2006 15:56:17 -0000 On Wednesday 29 November 2006 10:45, Lowell Gilbert wrote: = ICU is different in that not only can it be called from multiple = threads, but those threads can (in theory -- as far as I know, you're = right that nobody uses this capability right now) be interacting. = š = > You are (slightly) pessimizing your installation by building a = > thread-aware ICU without having a need for it... = = On the other hand, that could change at any time. š = = Think of all the other programmers in the world as independent = threads, that could use the API in new ways... š;-) I know. And the ICU developers think, threads should be enabled by default. The reason they aren't right now -- in the FreeBSD port -- is because of certain bugs in thread implementation(s) on FreeBSD/ia64, which lead to crashes in ICU. Since nothing is using ICU in a multi-threaded fashion _right now_, the current version of the port builds without threads by default. Norikatsu confirms, he is not calling into ICU from multiple threads either, hence my suggestion, he does not need to change the option. When the ia64's thread-implementation improves, the next update to the port will change the default... -mi