[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: bug report dispute resolution request

On Wed, Dec 13, 2000 at 11:09:41PM -0500, Branden Robinson wrote:
> My argument is that either:
> 1) Herbert can feel (with or without justification from the relevant
> standards documents) that ash should not support the construction at issue
> without whitespace added to disambiguate it;
>   in which case  the bug is a wishlist item against ash which he can
>   close or tag as "wontfix".

This is indeed my opinion.  However, my conclusion is different from yours.
Remember that the whole point of having ash is so that we can have slim and
fast /bin/sh.  In order to achieve that, we need to make sure that all our
#!/bin/sh scripts do not invoke features which outside those that are defined
by POSIX.  Indeed, we have the following paragraph from our policy:

     Shell scripts (`sh' and `bash') should almost certainly start with
     `set -e' so that errors are detected.  Every script should use `set
     -e' or check the exit status of _every_ command.

     The standard shell interpreter ``/bin/sh'' can be a symbolic link to
     any POSIX compatible shell, if `echo -n' does not generate a newline.
     [1] Thus, shell scripts specifying ``/bin/sh'' as interpreter should
     only use POSIX features.  If a script requires non-POSIX features from
     the shell interpreter, the appropriate shell must be specified in the
     first line of the script (e.g., ``#!/bin/bash'') and the package must
     depend on the package providing the shell (unless the shell package is
     marked `Essential', e.g., in the case of `bash').

So my conclusion is that since ash's behaviour is allowed by POSIX, your
#!/bin/sh script (xserver-xfree86.config) then is in contravention with
the policy.  Thus I reassigned bugs back to you.  A decision which I still
think is the right one.

> 2) Herbert can feel that ash SHOULD support the construction at issue
> without whitespace added, as bash does;
>   in which case the bug is either a normal or wishlist item against ash,
>   whichever he feels is appropriate, and he can fix it himself or submit it
>   upstream.

If that were my opinion, I would have fixed the bug as a matter of urgency,
as I usually do with other ash bugs of this kind.

> Either way, there is nothing I can do to XFree86 to make ash decide to
> parse shell scripts differently.  That is the issue, and why I feel the bug
> belongs on ash's plate, not mine.  As I said before, I already changed the
> shell script in my package to work with the existing version of ash,
> because I don't want the package to have to Pre-Depend on a particular
> version of ash in the event Herbert agrees that ash should handle the
> construction the way bash does.  The outstanding issue is what ash should
> do, not whether my shell script is parseable by ash.

Well, if that were the case (I couldn't find any fixed packages yesterday,
can't check right now since auric is unreachable), you could've simply closed
bugs if the packages were in the pool, or wait until they are and then close
the bugs.

> I should note here that Debian's ash is a substantial fork from the
> upstream version.  The codebases are quite divergent since Herbert is not
> in the habit of submitting his changes upstream for consideration.

Only because the NetBSD maintainers have yet to reply to any one of my
messages.  Fortunately, the other upstream maintainers of my packages are
much more cooperative.
Debian GNU/Linux 2.2 is out! ( http://www.debian.org/ )
Email:  Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Reply to: