From owner-svn-ports-all@freebsd.org Wed Jan 27 11:18:19 2016 Return-Path: Delivered-To: svn-ports-all@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 26A74A6ED7C for ; Wed, 27 Jan 2016 11:18:19 +0000 (UTC) (envelope-from freebsd.contact@marino.st) Received: from shepard.synsport.net (mail.synsport.com [208.69.230.148]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 0531611CD; Wed, 27 Jan 2016 11:18:17 +0000 (UTC) (envelope-from freebsd.contact@marino.st) Received: from [192.168.1.21] (248.Red-83-39-200.dynamicIP.rima-tde.net [83.39.200.248]) (using TLSv1 with cipher DHE-RSA-AES128-SHA (128/128 bits)) (No client certificate requested) by shepard.synsport.net (Postfix) with ESMTP id 3FDD543552; Wed, 27 Jan 2016 05:18:15 -0600 (CST) Subject: Re: svn commit: r407270 - head/ports-mgmt/portmaster To: Fabian Keil , John Marino References: <201601261123.u0QBNcvL091258@repo.freebsd.org> <56A86A88.1070908@marino.st> <20160127120734.7000ed9d@fabiankeil.de> Cc: svn-ports-all@freebsd.org From: John Marino X-Enigmail-Draft-Status: N1110 Reply-To: marino@freebsd.org Message-ID: <56A8A775.50104@marino.st> Date: Wed, 27 Jan 2016 12:18:13 +0100 User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.5.0 MIME-Version: 1.0 In-Reply-To: <20160127120734.7000ed9d@fabiankeil.de> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit X-BeenThere: svn-ports-all@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: SVN commit messages for the ports tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 27 Jan 2016 11:18:19 -0000 On 1/27/2016 12:07 PM, Fabian Keil wrote: > John Marino wrote: >> It pulls in one small package. It's noticibly faster than poudriere and >> blows portmaster out of the water (parallel building) > > Is it faster than poudriere when doing the same task (building packages > from source), or when doing something else? Yes, it has a higher build throughput when using the same machine and parameters. Poudriere is not a slouch but I think of it as a battleship (geared for build clusters) and synth is like a corvette aimed at normal users. It's aimed to be more user friendly and it works ootb without configuration (unless partition space for default directories is limited in space) > I only looked at synth briefly (and stopped when I realized that > it's written in Ada), but my impression was that it's written with > a different use case in mind and is designed to reuse binary packages. The ability to use official packages was added like 2 days ago. It was the last feature added, and it's not the default behavior. It has several purposes, but a main one is to replace portmaster (and portupgrade) and be more performant alternative to poudriere. > The "officially build versions" may be absolutely fine for you, > but some people prefer not to use unreproducible binaries they > didn't build themselves: > https://reproducible-builds.org/ So these people have a price to pay. And these people can use poudriere which has no dependencies and these people can keep using portmaster until further notice. Finally, these people can use a prebuilt synth and then use it to rebuilt itself. It's just a matter of time and downloading. The most exotic dependence is gcc6-aux and that takes 10 minutes to build on my core i5. > >> lightweight refers to performance, not the fact that it has dependencies. > That's hardly obvious. well, now people know. John