From owner-cvs-ports@FreeBSD.ORG Sat Aug 20 20:00:45 2011 Return-Path: Delivered-To: cvs-ports@FreeBSD.org Received: by hub.freebsd.org (Postfix, from userid 1033) id 0D06B106566B; Sat, 20 Aug 2011 20:00:45 +0000 (UTC) Date: Sat, 20 Aug 2011 20:00:45 +0000 From: Alexey Dokuchaev To: Mark Linimon Message-ID: <20110820200045.GA49034@FreeBSD.org> References: <201108201153.p7KBrPOV054618@repoman.freebsd.org> <4E4FD8A8.8070008@FreeBSD.org> <20110820162613.GA13064@FreeBSD.org> <20110820181937.GA1454@lonesome.com> Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline In-Reply-To: <20110820181937.GA1454@lonesome.com> User-Agent: Mutt/1.4.2.1i Cc: cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, Matthias Andree , ports-committers@FreeBSD.org Subject: Re: cvs commit: ports/misc/shared-mime-info Makefile X-BeenThere: cvs-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: CVS commit messages for the ports tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 20 Aug 2011 20:00:45 -0000 On Sat, Aug 20, 2011 at 01:19:37PM -0500, Mark Linimon wrote: > On Sat, Aug 20, 2011 at 04:26:13PM +0000, Alexey Dokuchaev wrote: > > Pav replied that we need to schedule few more -exp runs, but apparently > > portmgr@ was ENOTIME back then. > > Since your request, we've done -exps to update the ruby default, the > perl default, the apache default, and a number of other major things, > all of which were desirable for 9.0. The njobs change in my opinion > is more a nice-to-have. Understood (I presume). That's why my original inquiry was sent in relatively "quiet" period (after gmake update, AFAIR). I wholeheartedly support portmgr@'s desire to start taking care of it unless there are other sweeping changes pending on the queue. I just wanted to remind that -jX builds are long desired and once we have resources, we should bite the bullet and do it. On our side are number of GNU/Linux distros who had done this a while ago: either they have patches, of we can backport their versions MAKE_JOBS_UNSAFE=yes (i.e. it's a no-brainer) at least in the overlapping cases (read: popular ports). > All of this means that we can turn -exps far more quickly than we ever > have before. That's awesome! BTW, what about sparc64 boxen (several were offered via donations@) -- these days ports often get marked as BROKEN on sparcs, but our cluster does not seem to offer any for regular committers (contrary to ia64) to diagnose the problem/fix it. [1] [2] ./danfe [1] I am not even asking for jailed root access (yes, pun intended) to be able to install all necessary dependencies or (preferably) use marcuscom tinderbox. [2] Ditto for PowerPC.