From owner-freebsd-ports@FreeBSD.ORG Mon Feb 20 22:31:01 2012 Return-Path: Delivered-To: freebsd-ports@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 651B61065674; Mon, 20 Feb 2012 22:31:01 +0000 (UTC) (envelope-from max@mxcrypt.com) Received: from mail-vw0-f54.google.com (mail-vw0-f54.google.com [209.85.212.54]) by mx1.freebsd.org (Postfix) with ESMTP id E2A618FC16; Mon, 20 Feb 2012 22:31:00 +0000 (UTC) Received: by vbbfa15 with SMTP id fa15so5533432vbb.13 for ; Mon, 20 Feb 2012 14:31:00 -0800 (PST) Received: by 10.52.70.165 with SMTP id n5mr8649817vdu.55.1329777060268; Mon, 20 Feb 2012 14:31:00 -0800 (PST) MIME-Version: 1.0 Received: by 10.220.155.198 with HTTP; Mon, 20 Feb 2012 14:30:30 -0800 (PST) In-Reply-To: References: <2705.46.129.107.107.1328924592.squirrel@mymail.acsalaska.net> From: Maxim Khitrov Date: Mon, 20 Feb 2012 17:30:30 -0500 Message-ID: To: Alex Dupre Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable X-Gm-Message-State: ALoCoQkS8sNtIoBl9f1hU1xpKqgoQVUKBxl/rIlsIXdrjZbfG8H0AGTT/CZG0AGu6+Oy+vG29Qk+ Cc: freebsd-ports@freebsd.org Subject: Re: Requested changes to lang/php5 port 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: Mon, 20 Feb 2012 22:31:01 -0000 On Fri, Feb 10, 2012 at 9:38 PM, Maxim Khitrov wrote: > On Fri, Feb 10, 2012 at 8:43 PM, =C2=A0 wrote: >>> On Fri, Feb 10, 2012 at 12:41 PM, Maxim Khitrov wrote= : >>>> Hello, >>>> >>>> I'd like to request two minor changes to the lang/php5 port: >>>> >>>> 1. Change 'CONFIGURE_ARGS=3D' line to 'CONFIGURE_ARGS+=3D' to allow >>>> customizations to the arguments. One reason for this is that some >>>> extensions, like security/php5-hash, have to be compiled into the >>>> core. Otherwise, 'session.hash_function' option cannot use hash >>>> algorithms offered by that extension. See the following bug report for >>>> more info: >>>> >>>> https://bugs.php.net/bug.php?id=3D53789 >>>> >>>> 2. Unless there is a very good reason for keeping them, I think >>>> '--enable-libxml' and especially '--enable-mysqlnd' should be removed >>>> from the default set of arguments. Make these two part of the options >>>> dialog, if you want, but there are plenty of installations that have >>>> no need for any MySQL or XML functionality. Why compile these into the >>>> core by default? >>> >>> Actually, I now realize that changing 'CONFIGURE_ARGS=3D' to >>> 'CONFIGURE_ARGS+=3D' wouldn't help me, because --disable-all would >>> negate --enable-hash, which I was planning to specify. >> >> No, it won't negate it. --disable-all means "start with an empty set and >> for each --enable-foo add foo to the set". >> >> You can add to CONFIGURE_ARGS using a Makefile.local which is included >> at bsd.port.pre.mk time. > > Didn't know about Makefile.local, thanks! I tested the > 'CONFIGURE_ARGS+=3D' solution and it does work. I was under the > impression that order mattered. > > That method is slightly better than Makefile.local, because I try to > manage all port customizations with portconf. Otherwise, it's easy to > forget all the changes you made to get things working exactly right. I > also exclude /usr/ports from backups, so all settings would revert to > their defaults if I ever had to re-extract the entire tree. > >>> As for removing --enable-libxml, it seems to be required when building >>> the port (maybe for fpm?), but --enable-mysqlnd should definitely be >>> removed by default. >> >> As far as I can tell, --enable-mysqlnd does not add to dependencies, as: >> make -C /usr/ports/lang/php5 all-depends-list >> shows me as does: >> % ls /var/db/pkg|egrep 'php|mysql' >> php5-5.3.9 > > True, but it's still unnecessary. It increases the size of all > binaries (php, php-cgi, php-fpm) and adds a ton of useless information > to phpinfo() output. > > I think that making it an option that is enabled by default is > perfectly reasonable. I've attached a patch that does this. Tested it > on FreeBSD 9.0-RELEASE amd64. > > - Max Alex, So would it be possible to get this patch committed or are there some reasons why mysqlnd must be enabled in all cases? - Max