Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 24 Jul 2004 14:11:00 -0700
From:      John Merryweather Cooper <johnmary@adelphia.net>
To:        Alex Kapranoff <kappa@rambler-co.ru>
Cc:        Mike Makonnen <mtm@identd.net>
Subject:   Re: HEADS UP: change in ports rc.d script behaviour
Message-ID:  <20040724211100.GA49136@68-169-191-150.losaca.adelphia.net>
In-Reply-To: <41029BBC.3090405@rambler-co.ru>
References:  <20040724150905.GA998@rogue.acs-et.com> <41029BBC.3090405@rambler-co.ru>

next in thread | previous in thread | raw e-mail | index | archive | help

--WIyZ46R2i8wDzkSu
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Saturday, 24 July 2004 at 21:26:20 +0400, Alex Kapranoff wrote:
>Mike Makonnen wrote:
>
>>20040724:
>>        The rc.d/loclpkg script now "does the right thing" with respect
>>        to ports rc.d scripts. The most disruptive change is that scripts
>>        ending in *.sh are sourced in the same shell as rc.d/localpkg (as
>>        opposed to a subshell) and may cause the script to end prematurely
>>        if a script calls exit. Some broken scripts may not recognize the
>>        'fast' prefix to a command. The first problem can be fixed by sim=
ply
>>        renaming the script so that it doesn't have a '.sh' ending. Anoth=
er
>>        temporary work-arround is to completely remove the 'PROVIDE' line
>>        from the script. If you notice a broken ports rc.d script please
>>        notify the maintainer. Legacy scripts should not be affected.
>>        __FreeBSD_version bumped to 502124.
>>
>>
>>This update to the rc.d/localpkg attempts to normalize the behaviour
>>or ports rc.d scripts with that of the base system scripts. This may
>>break your ports scripts. Please report any breakage to the port's
>>maintainer so the script can be fixed.
>>
>What's the point of this change? What benefits do we get?
>Sorry for ignorance. I can only see that from now on, if any of the=20
>startup scripts terminates abnormally, we'll get
>a half-started system which in some cases could be extremely hard to repai=
r.

Well, while I respect the desire for cleanliness, architectural purity, etc=
., the reality is that rc.d/localpkg is already broken in -CURRENT for thos=
e of us who startup CUPS and similar ports.  EVERY reboot is a manual reboo=
t for me under the current system (i.e., I have to ^C out of the final star=
tup--because it stalls--and then run

# /etc/rc.d/localpkg restart

from root to get everything running again.  So any change in a direction th=
at makes things work on -CURRENT would be greatly appreciated.

jmc

--WIyZ46R2i8wDzkSu
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (FreeBSD)

iD8DBQFBAtBivBSBfjNhsrIRAusjAJ4nCMD41GB4tQXrg1OgDBAFSqHrogCffj35
Z20Jc37r+ehv4y4JXMUFnnM=
=b5cy
-----END PGP SIGNATURE-----

--WIyZ46R2i8wDzkSu--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040724211100.GA49136>