Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 22 Dec 2011 18:57:33 +0000
From:      Chris Rees <utisoft@gmail.com>
To:        Ruslan Mahmatkhanov <cvs-src@yandex.ru>
Cc:        leslie@eskk.nu, perryh@pluto.rain.com, Kevin Oberman <kob6558@gmail.com>, freebsd-ports@freebsd.org, Martin Matuska <mm@freebsd.org>
Subject:   Re: Stop in /usr/ports/multimedia/x264
Message-ID:  <CADLo83-ZbMcRmi-KqCzPXDzQMfjb-jpS%2BU49amNr92h1nt5wfg@mail.gmail.com>
In-Reply-To: <4EF374E1.6050206@yandex.ru>
References:  <4EF1E3EF.9000405@eskk.nu> <CAN6yY1tOrhyng1o5%2Bo5ko3cEoT2MLqCUF=L47Gesv%2BNnDdrhFw@mail.gmail.com> <4ef30e7b.gG6UGw%2BkkNWpdv07%perryh@pluto.rain.com> <4EF2F920.8060007@FreeBSD.org> <CAN6yY1viu8vxkmOQQEUKk%2BvzM7vGJ8--tsw8hbYfCyBbPyYNpQ@mail.gmail.com> <4EF35F70.5070606@yandex.ru> <CADLo83_qrR6--=eV7EyCkq%2Bg%2B_Ew1u_YCCTRWV22zGWJj=uo4Q@mail.gmail.com> <4EF374E1.6050206@yandex.ru>

next in thread | previous in thread | raw e-mail | index | archive | help
On 22 Dec 2011 18:20, "Ruslan Mahmatkhanov" <cvs-src@yandex.ru> wrote:
>
> Chris Rees wrote on 22.12.2011 22:12:
> > On 22 Dec 2011 16:49, "Ruslan Mahmatkhanov" <cvs-src@yandex.ru> wrote:
> >>
> >> Kevin Oberman wrote on 22.12.2011 20:38:
> >>> The only thing I can think of is that the failing system is the
> >>> oldest, running 8.2 while the others were running 9.0-RC2 and
> >>> 9.0-Beta3.
> >>
> >> Yes, all is worked just fine (before the bash fix commit) there on
> >> 10.0-CURRENT i386. I saw that configure warnings too, but it still was
> >> built successfully. May be bash only needed on some older FreeBSD
> >> versions like 8.2, that didn't had all that improvements for /bin/sh,
> >> that were merged in 9 and 10, so maybe it is the reason why it builded
> >> fine on this systems w/o any additional fixes.
> >>
> >
> > True, since 9 sh supports == -- however I might label this compatibility
> > shims, rather than 'improvements'...
> >
> > Chris
>
> I'm talking about overall /bin/sh work, not only about compatibility
> with some `non-true' shells. I know many things was done and tests
> coverage is terrific.
> So may we imply dependency on shells/bash and specific configure patches
> only if OSVERSION < 900000?
>

Trouble is with these problems is that they can hide, and don't always
result in outright failure (i've experienced this with musicpd and Clamav).

The scripts need a proper audit before being declared sh-safe.

Chris



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CADLo83-ZbMcRmi-KqCzPXDzQMfjb-jpS%2BU49amNr92h1nt5wfg>