On Thu, Apr 16 2009, Josselin Mouette wrote:
> Le mercredi 15 avril 2009 à 23:05 -0500, Manoj Srivastava a écrit :
>> So, no, policy does not just document current practice. Policy
>> tries to document what is right.
>
> I think it should be both. When we do things right, they should be
>
Le mercredi 15 avril 2009 à 23:05 -0500, Manoj Srivastava a écrit :
> So, no, policy does not just document current practice. Policy
> tries to document what is right.
I think it should be both. When we do things right, they should be
specified in the Policy, and there’s no point specifyi
Le mercredi 15 avril 2009 à 17:12 -0500, Raphael Geissert a écrit :
> Then there must be some sort of missunderstanding. My intention was not to
> troll, but to demonstrate the implications of what you said. I would like
> to apologise for my previous message as I had understood something
> complet
On Wed, Apr 15 2009, Josselin Mouette wrote:
> Policy documents practice.
I wish people would not say that. It is not true; and hasn't
been. And, moreover, we would not _want_ that to be true; there should
be no excuse to justify wanting to enshrine broken or bad practices
into policy.
Josselin Mouette wrote:
> Le mercredi 15 avril 2009 à 11:44 -0500, Raphael Geissert a écrit :
[...]
>
>> And taking your statement to the extreme, it means that if zsh was used
>> as /bin/sh then no other shell interpreter could ever be used as /bin/sh
>> ever again but a fork of zsh.
>
> I’m pr
Josselin Mouette writes:
> Le mercredi 15 avril 2009 à 02:16 -0700, Russ Allbery a écrit :
>> The advantage of the current Policy approach is that we have some hope
>> of introducing a new /bin/sh down the road, and we don't require that
>> packages comply with bugs in dash that should be fixed i
Le mercredi 15 avril 2009 à 11:44 -0500, Raphael Geissert a écrit :
> > Policy documents practice. When that new /bin/sh exists, you can change
>
> bash is the current /bin/sh, from your statements I could imply that we
> should require all /bin/sh's to support: b0rken bash arrrays, shell
> regexe
Josselin Mouette wrote:
> Le mercredi 15 avril 2009 à 02:16 -0700, Russ Allbery a écrit :
>> > Actually it would be better to specify that scripts must work with both
>> > sh implementations available in Debian, being bash and dash, rather
>> > than making nothing more than a fork of the POSIX spe
Le mercredi 15 avril 2009 à 02:16 -0700, Russ Allbery a écrit :
> > Actually it would be better to specify that scripts must work with both
> > sh implementations available in Debian, being bash and dash, rather than
> > making nothing more than a fork of the POSIX spec.
>
> The advantage of the c
Josselin Mouette writes:
> It is not the role of the policy to specify the exact requirements of
> the /bin/sh implementation.
It is, however, the role of Policy to specify the minimum required feature
set that all scripts can assume.
> Actually it would be better to specify that scripts must w
Le mardi 14 avril 2009 à 18:45 -0500, Raphael Geissert a écrit :
> what feature provided by dash is being deprecated?
> Like Russ said, if there's any feature not covered by policy that is
> reasonable to be required please say so.
It is not the role of the policy to specify the exact requirements
On Apr 15, Raphael Geissert wrote:
> what feature provided by dash is being deprecated?
You are the one who started the thread. Please come back when you will
actually know what you are proposing exactly.
> Like Russ said, if there's any feature not covered by policy that is
> reasonable to be r
Marco d'Itri wrote:
[...]
> The point is not to eliminate bashism but dashism, and so far there are
> no demonstrated benefits to deprecating features available in dash but
> not in posh.
>
what feature provided by dash is being deprecated?
Like Russ said, if there's any feature not covered by pol
On Apr 14, Stefano Zacchiroli wrote:
> Beside "local" per se, what is exactly the problem? If you badly need
> bash-specific features you can use /bin/bash as the interpreter.
Every deviation from upstream has a cost, which needs to be justified
by a cost-benefits analisys.
The point is not to el
On Sun, Apr 12, 2009 at 09:35:37PM +0200, Marco d'Itri wrote:
> > * Bashisms-free archive
> This is useless. You are basically proposing to remove every usage of
> "local" and a few other directives for no good reason but at a huge cost.
Beside "local" per se, what is exactly the problem? If you b
Marco d'Itri wrote:
> On Apr 12, Raphael Geissert wrote:
>
>> I don't have a list at hand (I do have a sort of list in another machine
>> which is unreachable atm); but the first one that comes to my mind is the
>> use of 'type', it's output is unreliable and in some shell interpreters
>> it is
On Apr 12, Raphael Geissert wrote:
> I don't have a list at hand (I do have a sort of list in another machine
> which is unreachable atm); but the first one that comes to my mind is the
> use of 'type', it's output is unreliable and in some shell interpreters it
> is not implemented. Another one
Marco d'Itri wrote:
> On Apr 12, "Roberto C. Sánchez" wrote:
>
>> Policy specifically states that use of "local" is permitted [0]:
> So exactly what do you want to disallow which is supported by dash but
> not by policy, and for which purpose?
>
I don't have a list at hand (I do have a sort of
On Apr 12, "Roberto C. Sánchez" wrote:
> Policy specifically states that use of "local" is permitted [0]:
So exactly what do you want to disallow which is supported by dash but
not by policy, and for which purpose?
--
ciao,
Marco
signature.asc
Description: Digital signature
On Sun, Apr 12, 2009 at 09:35:37PM +0200, Marco d'Itri wrote:
>
> > * Bashisms-free archive
> This is useless. You are basically proposing to remove every usage of
> "local" and a few other directives for no good reason but at a huge cost.
>
Policy specifically states that use of "local" is perm
m...@linux.it (Marco d'Itri) writes:
> On Apr 12, Raphael Geissert wrote:
>> * Bashisms-free archive
> This is useless. You are basically proposing to remove every usage of
> "local" and a few other directives for no good reason but at a huge
> cost.
No, he's not -- look at the current Policy.
On Apr 12, Raphael Geissert wrote:
> * Dash as default /bin/sh
This is good.
> * Bashisms-free archive
This is useless. You are basically proposing to remove every usage of
"local" and a few other directives for no good reason but at a huge cost.
> needed to support dash as /bin/sh. The pourpos
22 matches
Mail list logo
