Date: Thu, 25 Aug 2016 21:18:54 +1000 From: Kubilay Kocak <koobs@FreeBSD.org> To: Chris Rees <crees@physics.org>, Kevin Lo <kevlo@FreeBSD.org>, ports-committers@freebsd.org, svn-ports-all@freebsd.org, svn-ports-head@freebsd.org Subject: Re: svn commit: r420828 - head/net/py-nnpy Message-ID: <d205d495-22d3-8de6-4ef6-dc127974f359@FreeBSD.org> In-Reply-To: <A7253FA3-A709-44D0-B9A8-1723F060E308@physics.org> References: <201608250020.u7P0K1GM086639@repo.freebsd.org> <5416C91F-F57A-4078-8E6A-A15CC195C1DC@physics.org> <50b13de1-ee2a-40ad-69e0-39eb6beed1fc@FreeBSD.org> <A7253FA3-A709-44D0-B9A8-1723F060E308@physics.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 25/08/2016 8:34 PM, Chris Rees wrote: > > > On 25 August 2016 11:32:13 BST, Kubilay Kocak <koobs@FreeBSD.org> > wrote: >> On 25/08/2016 6:26 PM, Chris Rees wrote: >>> You should really either expand BUILD_DEPENDS before assigning >>> it, with := or assign BUILD_DEPENDS to RUN_DEPENDS instead. Doing >>> that avoids pollution from later additions such as extra USES >>> only used for build etc. >> >> I think this may be prior advice, and I remember a flurry of >> mailing lists posts in the past about := that may or may not have >> been resolved permanently. See Also: 'MY_DEPENDS' in the Porters >> Handbook. >> >> The other option is to duplicate the lines, which I'm not impartial >> to either. > > How can this have been resolved?? Make evaluates lazily, which is > why it works for us! Exactly why I said I wasn't sure if it ever 'resolved' :) Some said := is cool, others said no, something something something, MY_DEPENDS turned up in the PH. I use none of the above so as not to bother *anyone*. > Chris >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?d205d495-22d3-8de6-4ef6-dc127974f359>