Date: Fri, 9 Mar 2007 17:05:31 -0800 From: Ade Lovett <ade@FreeBSD.org> To: Doug Barton <dougb@FreeBSD.org> Cc: Jean-Yves Lefort <jylefort@FreeBSD.org>, freebsd ports <freebsd-ports@freebsd.org>, Kent Stewart <kstewart@owt.com>, Ade Lovett <ade@FreeBSD.org> Subject: Re: Ports 104877 causing big problems Message-ID: <FB399CF7-11E2-4CC9-8C91-7D6850B7B2D8@FreeBSD.org> In-Reply-To: <45F1EA6A.6070904@FreeBSD.org> References: <45F1DDE2.5030404@FreeBSD.org> <BE66AB56-E0B4-420A-910D-9C10DB9AF24D@FreeBSD.org> <45F1EA6A.6070904@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mar 09, 2007, at 15:14 , Doug Barton wrote: > Ade Lovett wrote: >> So, item (1): does the problem actually still exist with a port using >> the in-tree devel/libtool15 (via USE_AUTOTOOLS= libtool:15[:env]. If >> yes, empirical evidence will be required as an addendum to the >> PR. If >> no, then we're done. > > So it sounds like a reasonable way to proceed would be for Kent to > save a copy of his current libgpg-error +REQUIRED_BY file, then run > one of the commands that mezz suggested, and compare the before and > after pictures. If the problem is fixed, they should be substantially > different. Correct. This is the empirical evidence that needs to be determined and logged within the PR itself. Should it turn out that recent changes have not fixed the problem then, and only then, do we look at the appropriate solution. This would most likely be along the lines of an additional stanza to the USE_AUTOTOOLS construct rather than overloading GNU_CONFIGURE since: 1. There are most likely a number of ports that define GNU_CONFIGURE but which do NOT make use of libtool 2. When it comes to ports-wide operations (such as building indexes) we need to ensure that addition Mk/* infra-structural code is only brought in when needed. There is a non-zero cost to processing each Mk/bsd.*.mk file, so it is important to only bring these files in when absolutely necessary. 3. Ports that *are* affected by this issue (assuming the issue still exists) can be fixed in a more relaxed manner (eg: a conversion of GNU_CONFIGURE=YES to USE_AUTOTOOLS=configurehack [implying GNU_CONFIGURE=YES]) than a time-T switch. It will also allow for such affected ports to have PORTREVISIONs bumped by the respective maintainers so as to more clearly identify improved operation to the consumers of those ports. -aDe
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?FB399CF7-11E2-4CC9-8C91-7D6850B7B2D8>