Subject: Re: PR/31877 CVS commit: pkgsrc/x11/qt3-libs
To: None <darwin-pkg-people@NetBSD.org, gnats-admin@netbsd.org,>
From: Min Sik Kim <minskim@NetBSD.org>
List: pkgsrc-bugs
Date: 06/08/2006 13:40:02
The following reply was made to PR pkg/31877; it has been noted by GNATS.
From: Min Sik Kim <minskim@NetBSD.org>
To: joerg@britannica.bec.de
Cc: gnats-bugs@NetBSD.org
Subject: Re: PR/31877 CVS commit: pkgsrc/x11/qt3-libs
Date: Thu, 8 Jun 2006 06:34:59 -0700
On Jun 8, 2006, at 2:35 AM, joerg@britannica.bec.de wrote:
>
> On Thu, Jun 08, 2006 at 01:30:02AM +0000, Min Sik Kim wrote:
>> The following reply was made to PR pkg/31877; it has been noted by
>> GNATS.
>>
>> From: Min Sik Kim <minskim@netbsd.org>
>> To: gnats-bugs@NetBSD.org
>> Cc:
>> Subject: PR/31877 CVS commit: pkgsrc/x11/qt3-libs
>> Date: Thu, 8 Jun 2006 01:27:27 +0000 (UTC)
>>
>> Module Name: pkgsrc
>> Committed By: minskim
>> Date: Thu Jun 8 01:27:27 UTC 2006
>>
>> Modified Files:
>> pkgsrc/x11/qt3-libs: Makefile
>>
>> Log Message:
>> Set PKG_FAIL_REASON when a buggy gcc is used (Apple's gcc 4.0.0
>> yields
>> an internal compiler error). This addresses PR pkg/31877.
>
> Please move this into hacks.mk or don't do this at all.
I didn't put it into hacks.mk because bsd.hacks.mk has the following:
# This Makefile fragment automatically includes the necessary hacks to
# make the package build correctly by including the package's hacks.mk
# file if it exists.
and setting PKG_FAIL_REASON does not "make the package build correctly."
As for your second option, I think it is still better to tell users
why this package fails than to end up with an internal compiler error
after wasting so much time in compiling, especially in the case of a
huge package like qt3-libs.
Regards,
Min