Daniel Trebbien | 4 Dec 23:47 2010
Picon

git-svn mirror of the GNU Nano Subversion repository

Those who are interested in hacking on GNU Nano may be interested in a
git-svn mirror that I created from the Subversion repository:
https://github.com/dtrebbien/nano

If you would like to remake the mirror, I have written up detailed
instructions at:
https://github.com/dtrebbien/nano/wiki/Creating-a-git-svn-mirror-of-the-GNU-Nano-repository
Kamil Dudka | 6 Dec 20:12 2010
Picon

Re: [PATCH 1/2] check stat's result and avoid calling stat on a NULL pointer

On Sunday 28 November 2010 23:22:10 Mike Frysinger wrote:
> On Sunday, November 28, 2010 15:08:12 Kamil Dudka wrote:
> > +void stat_if_needed(const char *filename, struct stat **pstat)
>
> static
>
> > +    tmp = (struct stat *)nmalloc(sizeof(struct stat));
>
> not really specific to your patch, but the cast is useless and the sizeof
> would prob be better as sizeof(*tmp)
>
> > +    if (0 == stat(filename, tmp))
>
> order is reveresed from standard nano coding style

Sorry for the delay.  An updated patch is attached.  Thanks for review!

Kamil
_______________________________________________
Nano-devel mailing list
Nano-devel <at> gnu.org
http://lists.gnu.org/mailman/listinfo/nano-devel
Kamil Dudka | 6 Dec 20:19 2010
Picon

Re: [PATCH 2/2] use futimens() if available, instead of utime()

On Sunday 28 November 2010 23:17:37 Mike Frysinger wrote:
> typically people dont patch generated files (configure/config.h).  ignoring

Right, I didn't realize they were not maintained in the upstream SVN.

> that, this change doesnt handle the case where futimens() is ENOSYS. 
> rather than rewriting your own custom shims, how about using gnulib instead

What exactly do you mean by using gnulib?  Should I just import the related 
gnulib module(s) and post it here as a patch?  AFAIK there is no bootstrap 
for gnulib in nano at all yet.

Kamil
Kamil Dudka | 6 Dec 20:24 2010
Picon

Re: [PATCH 1/2] check stat's result and avoid calling stat on a NULL pointer

On Monday 06 December 2010 20:12:55 Kamil Dudka wrote:
> On Sunday 28 November 2010 23:22:10 Mike Frysinger wrote:
> > On Sunday, November 28, 2010 15:08:12 Kamil Dudka wrote:
> > > +void stat_if_needed(const char *filename, struct stat **pstat)
> >
> > static
> >
> > > +    tmp = (struct stat *)nmalloc(sizeof(struct stat));
> >
> > not really specific to your patch, but the cast is useless and the sizeof
> > would prob be better as sizeof(*tmp)
> >
> > > +    if (0 == stat(filename, tmp))
> >
> > order is reveresed from standard nano coding style
>
> Sorry for the delay.  An updated patch is attached.  Thanks for review!

Oops, meant to post this one.

> Kamil
_______________________________________________
Nano-devel mailing list
Nano-devel <at> gnu.org
http://lists.gnu.org/mailman/listinfo/nano-devel
Mike Frysinger | 6 Dec 21:06 2010
Picon

Re: [PATCH 2/2] use futimens() if available, instead of utime()

On Monday, December 06, 2010 14:19:03 Kamil Dudka wrote:
> On Sunday 28 November 2010 23:17:37 Mike Frysinger wrote:
> > that, this change doesnt handle the case where futimens() is ENOSYS.
> > rather than rewriting your own custom shims, how about using gnulib
> > instead
> 
> What exactly do you mean by using gnulib?  Should I just import the related
> gnulib module(s) and post it here as a patch?  AFAIK there is no bootstrap
> for gnulib in nano at all yet.

gnulib is considered generated code like "configure", so it shouldnt be 
committed.  which means using a gnulib bootstrap setup.  i think the scripts 
used by coreutils and such are fairly portable between projects ?
-mike
_______________________________________________
Nano-devel mailing list
Nano-devel <at> gnu.org
http://lists.gnu.org/mailman/listinfo/nano-devel
Kamil Dudka | 6 Dec 21:35 2010
Picon

Re: [PATCH 2/2] use futimens() if available, instead of utime()

On Monday 06 December 2010 21:06:22 Mike Frysinger wrote:
> gnulib is considered generated code like "configure", so it shouldnt be
> committed.  which means using a gnulib bootstrap setup.  i think the
> scripts used by coreutils and such are fairly portable between projects ?

I quickly looked at projects I am involved in.  coreutils and tar use a script 
named 'bootstrap', both sharing some code base with each other.  findutils 
uses a script named 'import-gnulib.sh', which is completely different from 
those above.

I am only afraid that requiring gnulib's bootstrap to be able to build 
development version of GNU nano, makes the project less nano.  There should 
be probably a wider consensus on it before introducing such a dependency.

Chris, what's your position on that?

Kamil
Eitan Adler | 6 Dec 21:40 2010

Re: [PATCH 2/2] use futimens() if available, instead of utime()

> I am only afraid that requiring gnulib's bootstrap to be able to build
> development version of GNU nano, makes the project less nano.  There should
> be probably a wider consensus on it before introducing such a dependency.
>
> Chris, what's your position on that?

I'm the maintainer of nano for FreeBSD.

I would prefer to not add an extra dependency for nano. However if it
is the only way to increase the portability of nano it might be
worthwhile (for some reason I doubt this) .

--

-- 
Eitan Adler
Mike Frysinger | 6 Dec 21:56 2010
Picon

Re: [PATCH 2/2] use futimens() if available, instead of utime()

On Monday, December 06, 2010 15:40:34 Eitan Adler wrote:
> > I am only afraid that requiring gnulib's bootstrap to be able to build
> > development version of GNU nano, makes the project less nano.  There
> > should be probably a wider consensus on it before introducing such a
> > dependency.
> > 
> > Chris, what's your position on that?
> 
> I'm the maintainer of nano for FreeBSD.
> 
> I would prefer to not add an extra dependency for nano. However if it
> is the only way to increase the portability of nano it might be
> worthwhile (for some reason I doubt this) .

it is only a dependency when building out of svn/git.  it does not affect the 
release at all.
-mike
_______________________________________________
Nano-devel mailing list
Nano-devel <at> gnu.org
http://lists.gnu.org/mailman/listinfo/nano-devel
Eitan Adler | 6 Dec 22:06 2010

Re: [PATCH 2/2] use futimens() if available, instead of utime()

> it is only a dependency when building out of svn/git.  it does not affect the
> release at all.

Must have missed that above - I'm fine with it then

--

-- 
Eitan Adler

Gmane