Ondrej Medek <[EMAIL PROTECTED]> writes:
> Building scripts should also conform to posh or specify /bin/bash. Some time
> ago I've rebuild one package (gaim-extendedprefs or gaim-encryption, I don't
> remember) and it failed because my /bin/sh is a link to dash.
Incorrect.
--
To UNSUBSCRIBE, e
On Sep 06, Ondrej Medek <[EMAIL PROTECTED]> wrote:
> Building scripts should also conform to posh or specify /bin/bash. Some time
Wrong. /bin/sh scripts should surely conform to dash, but there is no
point is wasting time with posh.
> ago I've rebuild one package (gaim-extendedprefs or gaim-encry
Andreas Metzler wrote:
> Thomas Hood <[EMAIL PROTECTED]> wrote:
> [command -v]
> > It is not even useful as a which(1) replacement. Whereas "which"
> > prints the pathname of the first executable file on the PATH,
> > "command -v" prints the pathname of the first executable file on the
> > PATH _o
Thomas Hood <[EMAIL PROTECTED]> wrote:
[command -v]
> It is not even useful as a which(1) replacement. Whereas "which"
> prints the pathname of the first executable file on the PATH,
> "command -v" prints the pathname of the first executable file on the
> PATH _or_ the pathname of the first non-ex
Thomas Hood wrote:
> It is not even useful as a which(1) replacement. Whereas "which"
> prints the pathname of the first executable file on the PATH,
Note that the Debian 'which' is different than the which on other
systems. The traditional which is a csh script that intentionally
sources the us
Marco d'Itri wrote:
> Andreas Metzler wrote:
> > Thomas Hood wrote:
> > > Is there anything else which dash supports but posh does not?
> > command -v
>
> Which is the well known which(1) replacement, and basically mandatory
> in a sane Debian shell. I keep wondering why people bother with posh.
"
On Aug 04, Andreas Metzler <[EMAIL PROTECTED]> wrote:
> >> I would appreciate it if someone would create the "potash" shell,
> >> consisting of posh modified to implement "test -a", "test -o" and
> >> "local". Debian would probably run on that well enough for it to
> >> be used as /bin/sh, and it
Marco d'Itri <[EMAIL PROTECTED]> wrote:
> On Aug 01, Thomas Hood <[EMAIL PROTECTED]> wrote:
>> I would appreciate it if someone would create the "potash" shell,
>> consisting of posh modified to implement "test -a", "test -o" and
>> "local". Debian would probably run on that well enough for it to
This one time, at band camp, sean finney said:
> hi,
>
> On Wed, Aug 03, 2005 at 10:49:31AM +0200, Thomas Hood wrote:
> > Joey Hess wrote:
> > > Unless, of course, it's a shell library that would like to be reasonably
> > > portable without being wholey crippled by lack features that have been
> >
hi,
On Wed, Aug 03, 2005 at 10:49:31AM +0200, Thomas Hood wrote:
> Joey Hess wrote:
> > Unless, of course, it's a shell library that would like to be reasonably
> > portable without being wholey crippled by lack features that have been
> > in every shell worth the name for ages.
>
> I know that t
Joey Hess wrote:
> Unless, of course, it's a shell library that would like to be reasonably
> portable without being wholey crippled by lack features that have been
> in every shell worth the name for ages.
I know that the debconf library used "local" (#242011) and I agree
that "local" is an exam
Steve Greenland wrote:
> I personally think it ought to be left as is. Either comply, or change
> the first line to "#!/bin/bash", which will always be there on Debian
> system. Any script complicated enough to need local can probably survive
> bash startup.
Unless, of course, it's a shell library
On Aug 01, Steve Greenland <[EMAIL PROTECTED]> wrote:
> I personally think it ought to be left as is. Either comply, or change
> the first line to "#!/bin/bash", which will always be there on Debian
This is pointless because it negates the speed and memory footprint
advantages of dash.
> system.
"Brian M. Carlson" <[EMAIL PROTECTED]> writes:
> This is a public service announcement about Debian Policy section 10.4,
> which states in part:
>
> 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]
>
On 01-Aug-05, 11:05 (CDT), Marco d'Itri <[EMAIL PROTECTED]> wrote:
> And again, please let's focus on the *purpose* of policy. POSIX
> compatibility is not supposed to be a goal in itself, compatibility with
> other UN*X /bin/sh or support for a faster shell than bash may be.
The purpose of 10.4
>Thomas Hood <[EMAIL PROTECTED]> wrote:
>> I would appreciate it if someone would create the "potash" shell,
>> consisting of posh modified to implement "test -a", "test -o" and
>> "local". Debian would probably run on that well enough for it to
>> be used as /bin/sh, and it could become the de fa
On Aug 01, Thomas Hood <[EMAIL PROTECTED]> wrote:
> I would appreciate it if someone would create the "potash" shell,
> consisting of posh modified to implement "test -a", "test -o" and
> "local". Debian would probably run on that well enough for it to
> be used as /bin/sh, and it could become th
I wrote:
>> I don't know what kind of importance a policy clause can have if it
>> has "nil" practical impact.
Steve Langasek wrote:
> I mean that the practical *benefit* of such strict enforcement is nil.
> The *impact* is that it would be a royal waste of developer time to
> make all scripts com
On Aug 01, Thomas Hood <[EMAIL PROTECTED]> wrote:
> Well, if, contrary to fact, the idea were widely supported then posh
> could be adapted so that it implemented the minimum set of features
> that Debian expected sh scripts to have. Then posh could be used to
> test whether scripts were complian
On Aug 01, David Weinehall <[EMAIL PROTECTED]> wrote:
> I think we should require that a base install should be POSIX
> compliant; for everything else we can be a bit more lax.
I see lack of a rationale here.
> But unlike some others, I don't see the point of rejecting patches
> to fix XSI:isms/b
On Mon, Aug 01, 2005 at 12:40:43AM -0700, Steve Langasek wrote:
> On Mon, Aug 01, 2005 at 08:44:17AM +0200, Thomas Hood wrote:
> > Steven Langasek wrote:
> > > One might as well be able to expand "posh" as the "Pathologically
> > > Overstrict SHell"
>
> > Well, if, contrary to fact, the idea were
On Mon, Aug 01, 2005 at 08:44:17AM +0200, Thomas Hood wrote:
> Steven Langasek wrote:
> > One might as well be able to expand "posh" as the "Pathologically
> > Overstrict SHell"
> Well, if, contrary to fact, the idea were widely supported then posh
> could be adapted so that it implemented the min
Steven Langasek wrote:
> One might as well be able to expand "posh" as the "Pathologically
> Overstrict SHell"
Well, if, contrary to fact, the idea were widely supported then posh
could be adapted so that it implemented the minimum set of features
that Debian expected sh scripts to have. Then po
Marco d'Itri wrote:
> On Jul 29, Steve Langasek <[EMAIL PROTECTED]> wrote:
>
> > Well, please note that posh is not the only shell that lacks support for
> > local. IIRC, it also breaks down under one or more of dash and busybox sh.
> dash supports local, or at least supports it in the way it's u
On Jul 29, Steve Langasek <[EMAIL PROTECTED]> wrote:
> Well, please note that posh is not the only shell that lacks support for
> local. IIRC, it also breaks down under one or more of dash and busybox sh.
dash supports local, or at least supports it in the way it's used in
all the packages I main
On Jul 29, "Brian M. Carlson" <[EMAIL PROTECTED]> wrote:
> I have found no less than four packages which break with /bin/posh
> as /bin/sh, including one that refuses to be removed because of its
> brokenness. I am expecting many more.
Yes, like most of my packages.
posh does not provide any bene
On Fri, Jul 29, 2005 at 01:16:36PM +0200, Thomas Hood wrote:
> Brian M. Carlson wrote:
> > I would like to point out that the following are not POSIX features:
> > local
> > test -o
> > test -a
> Regarding "local", see #294962. I would like to see support for
> "local" added to posh and "local"
Brian M. Carlson wrote:
> I would like to point out that the following are not POSIX features:
>
> local
> test -o
> test -a
Regarding "local", see #294962. I would like to see support for
"local" added to posh and "local" added as an exception to the
"POSIX-only" rule laid down in 10.4.
See #26
This is a public service announcement about Debian Policy section 10.4,
which states in part:
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 sh
29 matches
Mail list logo
