On Tue, 17 May 2011 11:06:51 +0000
Christoph Anton Mitterer <[email protected]> wrote:

> On Tue, 17 May 2011 19:14:27 +1000, Karl Goetz <[email protected]>
> wrote:
> > The current wording says bash must exist, so i guess this wasn't
> > such an issue (bash has printf as a builtin).
> > There is a bug about changing to reference posix instead of
> > specifying bash, perhaps this could be added as a prerequisite for
> > that change.
> a) Making bash a requirement should be dropped anyway.
> b) Does it say bash needs to be installed, or does it say /bin/sh
> needs to be bash?

It says:
> The following commands, or symbolic links to commands, are required
> in /bin.
[...]
> sh       The Bourne command shell
[...]
> If /bin/sh is not a true Bourne shell, it must be a hard or symbolic
> link to the real shell command.
> 
> The [ and test commands must be placed together in either /bin
> or /usr/bin.
> 
>     Rationale: For example bash behaves differently when called as sh
> or bash. The use of a symbolic link also allows users to easily see
> that /bin/sh is not a true Bourne shell.
> 
>     The requirement for the [ and test commands to be included as
> binaries (even if implemented internally by the shell) is shared with
> the POSIX.2 standard.

I guess we could add printf next to [ in that entry?

> c) In any case: As printf might be called with full path,.. it should
> be required IMHO.
> 
> It's not that big deal anyway, is it? I mean one more small binary...

i'm not fighting the change :)


> > If i was proposing changes I would rather suggest change the
> > wording to specify filesystem*s* (rather then the current
> > filesystem) to allow the practise of (for example)
> > using /mnt/sdc4 /mnt/sdb3 to rsync data between during recoveries.
> 
> But this would already open up the abuse from applications as the CG
> stuff I've mentioned before.

key word here is abuse - i don't believe they'll stop abusing it just
because we explicitly say they should. 

They are *already* failing to comply with the fhs on this.
> This directory [/mnt] must not be used by installation programs: a
> suitable temporary directory not in use by the system must be used
> instead.

([/mnt] added by me.)

> I'd simply say it is intended to be a single temporary mountpoint to
> be handled by the admin (and especially not used by any programs).
> If the admin now decides to makes subdirs, than this is up to him.

> > We can't know which filesystems this is in advance.
> That's why I say,.. put it completely in the hands of the distros.

Hm, not sure about this - same with your printf above - what about had
coded paths?

> > there are bugs about X11R6 and the */games directories, are there
> > other locations to look for?
> No I guess that's it.... IIRC there were also some /*/X11 dirs.
> Maybe one could also drop this.

A quick apt-file check doesn't show any */X11/ dirs that look
removable, unless its renaming X11-> Xorg in paths.
thanks,
kk

-- 
Karl Goetz, (Kamping_Kaiser / VK5FOSS)
Debian contributor / gNewSense Maintainer
http://www.kgoetz.id.au
No, I won't join your social networking group

Attachment: signature.asc
Description: PGP signature

_______________________________________________
fhs-discuss mailing list
[email protected]
https://lists.linux-foundation.org/mailman/listinfo/fhs-discuss

Reply via email to