Re: LSB Status
- To: debian-policy@lists.debian.org
- Subject: Re: LSB Status
- From: Grant Bowman <grantbow@svpal.org>
- Date: Sat, 5 Jan 2002 07:10:26 -0800
- Message-id: <[🔎] 20020105071025.A24680@svpal.svpal.org>
- In-reply-to: <20011203212509.A23069@azure.humbug.org.au>
- References: <9u32mq$iet$1@ncc1701.cistron.net> <XFMail.20011128085522.shalehperry@home.com> <20011202135455.B25729@wiggy.net> <20011202233522.B4702@azure.humbug.org.au> <20011203014909.D3534@wiggy.net> <20011203110006.B15472@azure.humbug.org.au> <20011203120430.H3534@wiggy.net> <20011203212509.A23069@azure.humbug.org.au>
* Anthony Towns <aj@azure.humbug.org.au> [011203 15:33]:
> On Mon, Dec 03, 2001 at 12:04:30PM +0100, Wichert Akkerman wrote:
> > Previously Anthony Towns wrote:
> > > Things break. That's what happen when things fail. You'll notice we don't
> > > guarantee anything better for our own init scripts.
> > LSB does so we will need to start caring. You can't selectively
> > implement the LSB, that would make the whole thing worthless.
>
> Eh? The LSB does no such thing:
>
> "An init.d shell script may declare using the "Required-Start:" header
> that it must not be run until certain boot facilities are provided. This
> infomration is used by the installation tool or the boot-time boot-script
> execution facility to ensure that init scripts are run in the correct
> order."
>
> The LSB is specifically designed *not* to require major changes to the
> way distributions or systems are set up to function.
This seems like one exception which is why I wanted to at least clarify
(sooner rather than later) what Debian's strategy is to implement it.
I am sure other items will come up along these lines, but we need to
start the process to start working through these items after more people
have had a chance to read the specification more closely.
> As a matter of implementation quality we may wish to do something to that
> effect, but it's not a requirement, and not doing it certainly doesn't
> make the whole thing worthless. And there's little point worrying about
> implementation quality when we don't have an implementation in the
> first place.
Actually, "Init files shall accept one argument, saying what to do" with
all of {start, stop, restart, reload, force-reload, status} being
listed. This indicates to me that this is a required change to be LSB
compliant and a highly visible one to sysadmins.
http://www.linuxbase.org/spec/gLSB/gLSB/iniscrptact.html
Even more specifically "Each LSB-compliant init.d script must source the
file /lib/lsb/init-functions. This file must cause the following shell
script commands to be defined..." There's also a requirement that may
have other ramifications. "...the LSB init.d files themselves should
only depend in /bin/sh features as defined by POSIX.2."
http://www.linuxbase.org/spec/gLSB/gLSB/iniscrptfunc.html
Cheers,
--
-- Grant Bowman <grantbow@svpal.org>
Reply to: