[PATCH] get_maintainer: Don't fallback to git by default

[PATCH] get_maintainer: Don't fallback to git by default

[Javier Martinez Canillas]

The get_maintainer.pl script could be used to get a list of people that
has to be in the copy list when posting patches for a given file but it
defaults to git fallback so the list returned isn't really relevant and
submitters are posting patches to random developers that just happened
to touch that file in the past.

Disable git fallback so get_maintainer.pl only returns the information
that is in MAINTAINERS by default.

Suggested-by: Wolfram Sang <wsa@the-dreams.de>
Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>

---
Hello,

This patch was suggested by Wolfram on this thread [0] but it is
something I thought before because I've seen people complaining
on a regular basis about being cc'ed in random patches and there
are tools like patman [1] that rely on get_maintainers.pl to post
the patches.

[0]: https://lkml.org/lkml/2015/10/26/444
[1]: http://git.denx.de/?p=u-boot.git;a=blob;f=tools/patman/README

Best regards,
Javier

 scripts/get_maintainer.pl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl
index 98bae869f6d0..b52ae26270e8 100755
--- a/scripts/get_maintainer.pl
+++ b/scripts/get_maintainer.pl
@@ -29,7 +29,7 @@ my $email_git = 0;
 my $email_git_all_signature_types = 0;
 my $email_git_blame = 0;
 my $email_git_blame_signatures = 1;
-my $email_git_fallback = 1;
+my $email_git_fallback = 0;
 my $email_git_min_signatures = 1;
 my $email_git_max_maintainers = 5;
 my $email_git_min_percent = 5;
-- 
2.4.3

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Laura Abbott]

On 10/26/2015 03:11 PM, Javier Martinez Canillas wrote:
> The get_maintainer.pl script could be used to get a list of people that
> has to be in the copy list when posting patches for a given file but it
> defaults to git fallback so the list returned isn't really relevant and
> submitters are posting patches to random developers that just happened
> to touch that file in the past.
>
> Disable git fallback so get_maintainer.pl only returns the information
> that is in MAINTAINERS by default.
>
> Suggested-by: Wolfram Sang <wsa@the-dreams.de>
> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
>
> ---
> Hello,
>
> This patch was suggested by Wolfram on this thread [0] but it is
> something I thought before because I've seen people complaining
> on a regular basis about being cc'ed in random patches and there
> are tools like patman [1] that rely on get_maintainers.pl to post
> the patches.
>
> [0]: https://lkml.org/lkml/2015/10/26/444
> [1]: http://git.denx.de/?p=u-boot.git;a=blob;f=tools/patman/README
>
> Best regards,
> Javier
>

We might need to do a sweep of files to make sure more of them are
listed in the maintainers file, else by default they only show LKML.
As an example, a bunch of files in lib are only showing LKML.
Defaulting to only sending to LKML may be swinging too far in the wrong
direction for reducing inbox size.

>   scripts/get_maintainer.pl | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl
> index 98bae869f6d0..b52ae26270e8 100755
> --- a/scripts/get_maintainer.pl
> +++ b/scripts/get_maintainer.pl
> @@ -29,7 +29,7 @@ my $email_git = 0;
>   my $email_git_all_signature_types = 0;
>   my $email_git_blame = 0;
>   my $email_git_blame_signatures = 1;
> -my $email_git_fallback = 1;
> +my $email_git_fallback = 0;
>   my $email_git_min_signatures = 1;
>   my $email_git_max_maintainers = 5;
>   my $email_git_min_percent = 5;
>

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Javier Martinez Canillas]

Hello Laura,

Thanks a lot for your feedback.

On 10/27/2015 08:00 AM, Laura Abbott wrote:
> On 10/26/2015 03:11 PM, Javier Martinez Canillas wrote:
>> The get_maintainer.pl script could be used to get a list of people that
>> has to be in the copy list when posting patches for a given file but it
>> defaults to git fallback so the list returned isn't really relevant and
>> submitters are posting patches to random developers that just happened
>> to touch that file in the past.
>>
>> Disable git fallback so get_maintainer.pl only returns the information
>> that is in MAINTAINERS by default.
>>
>> Suggested-by: Wolfram Sang <wsa@the-dreams.de>
>> Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
>>
>> ---
>> Hello,
>>
>> This patch was suggested by Wolfram on this thread [0] but it is
>> something I thought before because I've seen people complaining
>> on a regular basis about being cc'ed in random patches and there
>> are tools like patman [1] that rely on get_maintainers.pl to post
>> the patches.
>>
>> [0]: https://lkml.org/lkml/2015/10/26/444
>> [1]: http://git.denx.de/?p=u-boot.git;a=blob;f=tools/patman/README
>>
>> Best regards,
>> Javier
>>
> 
> We might need to do a sweep of files to make sure more of them are
> listed in the maintainers file, else by default they only show LKML.
> As an example, a bunch of files in lib are only showing LKML.
> Defaulting to only sending to LKML may be swinging too far in the wrong
> direction for reducing inbox size.
>

You are correct, disabling git fallback could be a too radical solution
for now until the MAINTAINERS file is more complete.

I'll do some auditing to the files in the source tree and see if I can
come with a patch series that improves the state of MAINTAINERS before
disabling this option by default.

Best regards,
-- 
Javier Martinez Canillas
Open Source Group
Samsung Research America

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Theodore Ts'o]

What if we have a git_fallback "auto" mode which only falls back to
using git if there is no entry in the MAINTAINERS file?  That might
address the concern which Laura raised, without blocking this
change until the MAINTAINERS file is fully populated.

Cheers,

						- Ted

Re: [PATCH] get_maintainer: Don't fallback to git by default

[joe]

On 2015-10-26 19:47, Theodore Ts'o wrote:
> What if we have a git_fallback "auto" mode which only falls back to
> using git if there is no entry in the MAINTAINERS file?  That might
> address the concern which Laura raised, without blocking this
> change until the MAINTAINERS file is fully populated.

That's what happens already right now.

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Laura Abbott]

On 10/26/15 9:57 PM, joe@perches.com wrote:
> On 2015-10-26 19:47, Theodore Ts'o wrote:
>> What if we have a git_fallback "auto" mode which only falls back to
>> using git if there is no entry in the MAINTAINERS file?  That might
>> address the concern which Laura raised, without blocking this
>> change until the MAINTAINERS file is fully populated.
>
> That's what happens already right now.

Not with the change proposed though

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Lee Jones]

On Mon, 26 Oct 2015, joe@perches.com wrote:
> On 2015-10-26 19:47, Theodore Ts'o wrote:
> >What if we have a git_fallback "auto" mode which only falls back to
> >using git if there is no entry in the MAINTAINERS file?  That might
> >address the concern which Laura raised, without blocking this
> >change until the MAINTAINERS file is fully populated.
> 
> That's what happens already right now.

I think it can be done better.

FYI, this patch came about because I keep complaining about submitters
sending me patches I really shouldn't have to care about.

Here is the most recent offending command output:

  $ ./scripts/get_maintainer.pl -f drivers/i2c/busses/Kconfig
  Wolfram Sang <wsa@the-dreams.de> (maintainer:I2C SUBSYSTEM,commit_signer:26/26=100%,authored:3/26=12%,removed_lines:3/11=27%)
  Lee Jones <lee.jones@linaro.org> (commit_signer:3/26=12%)
  Geert Uytterhoeven <geert@linux-m68k.org> (commit_signer:3/26=12%,authored:3/26=12%,removed_lines:2/11=18%)
  David Box <david.e.box@linux.intel.com> (commit_signer:3/26=12%,added_lines:11/146=8%)
  Jean Delvare <jdelvare@suse.de> (commit_signer:2/26=8%)
  Javier Martinez Canillas <javier@osg.samsung.com> (authored:2/26=8%,removed_lines:2/11=18%)
  Neelesh Gupta <neelegup@linux.vnet.ibm.com> (added_lines:11/146=8%)
  James Hogan <james.hogan@imgtec.com> (added_lines:10/146=7%)
  Ray Jui <rjui@broadcom.com> (added_lines:10/146=7%)
  Joachim Eastwood <manabian@gmail.com> (added_lines:10/146=7%)
  Andrew Bresticker <abrestic@chromium.org> (removed_lines:2/11=18%)
  Zhangfei Gao <zhangfei.gao@linaro.org> (removed_lines:1/11=9%)
  linux-i2c@vger.kernel.org (open list:I2C SUBSYSTEM)
  linux-kernel@vger.kernel.org (open list)

Anyone south of Wolfram and north of the I2C ML is unlikely to care.

Please find a way to make this better.  For the sake of our sanity.

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 291 bytes --]


> Anyone south of Wolfram and north of the I2C ML is unlikely to care.

Actually, driver maintainers should care about their portion of it.
Which leads me to the conclusion that driver maintainer entries in
MAINTAINERS should have the relevant Kconfig symbol as a keyword match,
probably.


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Lee Jones]

On Tue, 27 Oct 2015, Wolfram Sang wrote:
> > Anyone south of Wolfram and north of the I2C ML is unlikely to care.
> 
> Actually, driver maintainers should care about their portion of it.
> Which leads me to the conclusion that driver maintainer entries in
> MAINTAINERS should have the relevant Kconfig symbol as a keyword match,
> probably.

Good point, well put.  Although this situation is fairly unique to
Kconfig changes.  In general I know I don't want to be spammed with
changes to all the files I've ever adapted though.  If people wish to
act as reviewers, they should have a mention in MAINTAINERS for
get_maintiner.pl to pull information from:

R: Designated reviewer: FullName <address@domain>
   These reviewers should be CCed on patches.

Or a new marker, I: Interested parties.

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 315 bytes --]


> act as reviewers, they should have a mention in MAINTAINERS for
> get_maintiner.pl to pull information from:
> 
> R: Designated reviewer: FullName <address@domain>
>    These reviewers should be CCed on patches.

I think this should do.

Interested parties should start by scanning the mailing list :)

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Joe Perches]

On Tue, 2015-10-27 at 11:29 +0100, Wolfram Sang wrote:
> > act as reviewers, they should have a mention in MAINTAINERS for
> > get_maintiner.pl to pull information from:
> > 
> > R: Designated reviewer: FullName <address@domain>
> >    These reviewers should be CCed on patches.
> 
> I think this should do.
> 
> Interested parties should start by scanning the mailing list :)

I think the concept of reviewers is fine.

A couple weeks ago, I suggested a mechanism to have reviewers
concentrate their entries into separate REVIEWERS file(s)

https://lkml.org/lkml/2015/10/16/649

These REVIEWERS* files would also work for companies that have
people with specific areas of expertise that want internal review
before public exposure.

Kconfig and Makefile files in directories like drivers/i2c/busses
are outliers because all new individual drivers are added to both
files.

The mechanism to limit git-fallback use is a directory depth match.

Wolfram has a MAINTAINERS entry of:
	F:	drivers/i2c/
but that doesn't pattern doesn't match the directory depth of
file drivers/i2c/busses/Kconfig so git-fallback is used for files
that don't have specific MAINTAINERS.

It _might_ make some sense to limit activation of git history in
that specific case, but for some path like arch/, that limitation
might _not_ make sense.

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Joe Perches]

On Tue, 2015-10-27 at 08:21 +0000, Lee Jones wrote:
> On Mon, 26 Oct 2015, joe@perches.com wrote:
> > On 2015-10-26 19:47, Theodore Ts'o wrote:
> > >What if we have a git_fallback "auto" mode which only falls back to
> > >using git if there is no entry in the MAINTAINERS file?  That might
> > >address the concern which Laura raised, without blocking this
> > >change until the MAINTAINERS file is fully populated.
> > 
> > That's what happens already right now.
> 
> I think it can be done better.
> 
> FYI, this patch came about because I keep complaining about submitters
> sending me patches I really shouldn't have to care about.

Then you should add your name to the .get_maintainer.ignore file.

If you do, you'll only get listed by get_maintainer for files where
you have a specific entry in MAINTAINERS.

> Here is the most recent offending command output:
> 
>   $ ./scripts/get_maintainer.pl -f drivers/i2c/busses/Kconfig
>   Wolfram Sang <wsa@the-dreams.de> (maintainer:I2C SUBSYSTEM,commit_signer:26/26=100%,authored:3/26=12%,removed_lines:3/11=27%)
>   Lee Jones <lee.jones@linaro.org> (commit_signer:3/26=12%)
>   Geert Uytterhoeven <geert@linux-m68k.org> (commit_signer:3/26=12%,authored:3/26=12%,removed_lines:2/11=18%)
>   David Box <david.e.box@linux.intel.com> (commit_signer:3/26=12%,added_lines:11/146=8%)
>   Jean Delvare <jdelvare@suse.de> (commit_signer:2/26=8%)
>   Javier Martinez Canillas <javier@osg.samsung.com> (authored:2/26=8%,removed_lines:2/11=18%)
>   Neelesh Gupta <neelegup@linux.vnet.ibm.com> (added_lines:11/146=8%)
>   James Hogan <james.hogan@imgtec.com> (added_lines:10/146=7%)
>   Ray Jui <rjui@broadcom.com> (added_lines:10/146=7%)
>   Joachim Eastwood <manabian@gmail.com> (added_lines:10/146=7%)
>   Andrew Bresticker <abrestic@chromium.org> (removed_lines:2/11=18%)
>   Zhangfei Gao <zhangfei.gao@linaro.org> (removed_lines:1/11=9%)
>   linux-i2c@vger.kernel.org (open list:I2C SUBSYSTEM)
>   linux-kernel@vger.kernel.org (open list)
> 
> Anyone south of Wolfram and north of the I2C ML is unlikely to care.
> 
> Please find a way to make this better.  For the sake of our sanity.

Your sanity must be pretty darn fragile.

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Lee Jones]

On Tue, 27 Oct 2015, Joe Perches wrote:

> On Tue, 2015-10-27 at 08:21 +0000, Lee Jones wrote:
> > On Mon, 26 Oct 2015, joe@perches.com wrote:
> > > On 2015-10-26 19:47, Theodore Ts'o wrote:
> > > >What if we have a git_fallback "auto" mode which only falls back to
> > > >using git if there is no entry in the MAINTAINERS file?  That might
> > > >address the concern which Laura raised, without blocking this
> > > >change until the MAINTAINERS file is fully populated.
> > > 
> > > That's what happens already right now.
> > 
> > I think it can be done better.
> > 
> > FYI, this patch came about because I keep complaining about submitters
> > sending me patches I really shouldn't have to care about.
> 
> Then you should add your name to the .get_maintainer.ignore file.
> 
> If you do, you'll only get listed by get_maintainer for files where
> you have a specific entry in MAINTAINERS.

Oooo... I didn't even know that existed.  It's an idea, but I'd rather
have these kinds of issues sorted out for everyone, rather than being
a special case and putting myself in the grumpy-old-b***** file.  I'm
certainly not the only one suffering with this problem.

> > Here is the most recent offending command output:
> > 
> >   $ ./scripts/get_maintainer.pl -f drivers/i2c/busses/Kconfig
> >   Wolfram Sang <wsa@the-dreams.de> (maintainer:I2C SUBSYSTEM,commit_signer:26/26=100%,authored:3/26=12%,removed_lines:3/11=27%)
> >   Lee Jones <lee.jones@linaro.org> (commit_signer:3/26=12%)
> >   Geert Uytterhoeven <geert@linux-m68k.org> (commit_signer:3/26=12%,authored:3/26=12%,removed_lines:2/11=18%)
> >   David Box <david.e.box@linux.intel.com> (commit_signer:3/26=12%,added_lines:11/146=8%)
> >   Jean Delvare <jdelvare@suse.de> (commit_signer:2/26=8%)
> >   Javier Martinez Canillas <javier@osg.samsung.com> (authored:2/26=8%,removed_lines:2/11=18%)
> >   Neelesh Gupta <neelegup@linux.vnet.ibm.com> (added_lines:11/146=8%)
> >   James Hogan <james.hogan@imgtec.com> (added_lines:10/146=7%)
> >   Ray Jui <rjui@broadcom.com> (added_lines:10/146=7%)
> >   Joachim Eastwood <manabian@gmail.com> (added_lines:10/146=7%)
> >   Andrew Bresticker <abrestic@chromium.org> (removed_lines:2/11=18%)
> >   Zhangfei Gao <zhangfei.gao@linaro.org> (removed_lines:1/11=9%)
> >   linux-i2c@vger.kernel.org (open list:I2C SUBSYSTEM)
> >   linux-kernel@vger.kernel.org (open list)
> > 
> > Anyone south of Wolfram and north of the I2C ML is unlikely to care.
> > 
> > Please find a way to make this better.  For the sake of our sanity.
> 
> Your sanity must be pretty darn fragile.

Ticking time-bomb, me. ;)

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Joe Perches]

On Tue, 2015-10-27 at 10:49 -0700, Joe Perches wrote:
> On Tue, 2015-10-27 at 11:29 +0100, Wolfram Sang wrote:
> > > act as reviewers, they should have a mention in MAINTAINERS for
> > > get_maintiner.pl to pull information from:
> > > 
> > > R: Designated reviewer: FullName <address@domain>
> > >    These reviewers should be CCed on patches.
> > 
> > I think this should do.
> > 
> > Interested parties should start by scanning the mailing list :)
> 
> I think the concept of reviewers is fine.
> 
> A couple weeks ago, I suggested a mechanism to have reviewers
> concentrate their entries into separate REVIEWERS file(s)
> 
> https://lkml.org/lkml/2015/10/16/649
> 
> These REVIEWERS* files would also work for companies that have
> people with specific areas of expertise that want internal review
> before public exposure.
> 
> Kconfig and Makefile files in directories like drivers/i2c/busses
> are outliers because all new individual drivers are added to both
> files.
> 
> The mechanism to limit git-fallback use is a directory depth match.
> 
> Wolfram has a MAINTAINERS entry of:
> 	F:	drivers/i2c/
> but that doesn't pattern doesn't match the directory depth of
> file drivers/i2c/busses/Kconfig so git-fallback is used for files
> that don't have specific MAINTAINERS.
> 
> It _might_ make some sense to limit activation of git history in
> that specific case, but for some path like arch/, that limitation
> might _not_ make sense.

Another option would be to add another pattern to the
I2C section that matches all subdirectories of drivers/i2c/:
---
 MAINTAINERS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index fb7d2e4..c670e1f 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5104,6 +5104,7 @@ S:	Maintained
 F:	Documentation/devicetree/bindings/i2c/
 F:	Documentation/i2c/
 F:	drivers/i2c/
+F:	drivers/i2c/*/
 F:	include/linux/i2c.h
 F:	include/linux/i2c-*.h
 F:	include/uapi/linux/i2c.h

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Javier Martinez Canillas]

Hello,

On 10/28/2015 03:09 AM, Lee Jones wrote:
> On Tue, 27 Oct 2015, Joe Perches wrote:
> 
>> On Tue, 2015-10-27 at 08:21 +0000, Lee Jones wrote:
>>> On Mon, 26 Oct 2015, joe@perches.com wrote:
>>>> On 2015-10-26 19:47, Theodore Ts'o wrote:
>>>>> What if we have a git_fallback "auto" mode which only falls back to
>>>>> using git if there is no entry in the MAINTAINERS file?  That might
>>>>> address the concern which Laura raised, without blocking this
>>>>> change until the MAINTAINERS file is fully populated.
>>>>
>>>> That's what happens already right now.
>>>
>>> I think it can be done better.
>>>
>>> FYI, this patch came about because I keep complaining about submitters
>>> sending me patches I really shouldn't have to care about.
>>
>> Then you should add your name to the .get_maintainer.ignore file.
>>
>> If you do, you'll only get listed by get_maintainer for files where
>> you have a specific entry in MAINTAINERS.
> 
> Oooo... I didn't even know that existed.  It's an idea, but I'd rather
> have these kinds of issues sorted out for everyone, rather than being
> a special case and putting myself in the grumpy-old-b***** file.  I'm
> certainly not the only one suffering with this problem.
>

I agree with Lee here, if everyone has to be added to a file then that
solution does not scale. I bet that no one working on the kernel would
say that is OK on receiving random patches.

>>> Here is the most recent offending command output:
>>>
>>>   $ ./scripts/get_maintainer.pl -f drivers/i2c/busses/Kconfig
>>>   Wolfram Sang <wsa@the-dreams.de> (maintainer:I2C SUBSYSTEM,commit_signer:26/26=100%,authored:3/26=12%,removed_lines:3/11=27%)
>>>   Lee Jones <lee.jones@linaro.org> (commit_signer:3/26=12%)
>>>   Geert Uytterhoeven <geert@linux-m68k.org> (commit_signer:3/26=12%,authored:3/26=12%,removed_lines:2/11=18%)
>>>   David Box <david.e.box@linux.intel.com> (commit_signer:3/26=12%,added_lines:11/146=8%)
>>>   Jean Delvare <jdelvare@suse.de> (commit_signer:2/26=8%)
>>>   Javier Martinez Canillas <javier@osg.samsung.com> (authored:2/26=8%,removed_lines:2/11=18%)
>>>   Neelesh Gupta <neelegup@linux.vnet.ibm.com> (added_lines:11/146=8%)
>>>   James Hogan <james.hogan@imgtec.com> (added_lines:10/146=7%)
>>>   Ray Jui <rjui@broadcom.com> (added_lines:10/146=7%)
>>>   Joachim Eastwood <manabian@gmail.com> (added_lines:10/146=7%)
>>>   Andrew Bresticker <abrestic@chromium.org> (removed_lines:2/11=18%)
>>>   Zhangfei Gao <zhangfei.gao@linaro.org> (removed_lines:1/11=9%)
>>>   linux-i2c@vger.kernel.org (open list:I2C SUBSYSTEM)
>>>   linux-kernel@vger.kernel.org (open list)
>>>
>>> Anyone south of Wolfram and north of the I2C ML is unlikely to care.
>>>
>>> Please find a way to make this better.  For the sake of our sanity.
>>

Another option is to bump the git-fallback min % used as a threshold,
for example when using a min value of 20 the list is more sensible:

$ ./scripts/get_maintainer.pl -f --git-min-percent=20 drivers/i2c/busses/Kconfig 
Wolfram Sang <wsa@the-dreams.de> (maintainer:I2C SUBSYSTEM,commit_signer:26/26=100%,removed_lines:3/11=27%)
linux-i2c@vger.kernel.org (open list:I2C SUBSYSTEM)
linux-kernel@vger.kernel.org (open list)

>> Your sanity must be pretty darn fragile.
> 
> Ticking time-bomb, me. ;)
> 

Is not only Lee tbh, I see people complaining almost on a daily so
the current default is not good enough. This of course annoys more
to maintainers that already have a very heavy email load but it is
also inconvenient for developers like me since by doing a change
on a file that is frequently modified (i.e: Makefile, Kconfig, etc)
you earn a life of spam :)

Best regards,
-- 
Javier Martinez Canillas
Open Source Group
Samsung Research America

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 707 bytes --]


> Another option would be to add another pattern to the
> I2C section that matches all subdirectories of drivers/i2c/:
> ---
>  MAINTAINERS | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index fb7d2e4..c670e1f 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -5104,6 +5104,7 @@ S:	Maintained
>  F:	Documentation/devicetree/bindings/i2c/
>  F:	Documentation/i2c/
>  F:	drivers/i2c/
> +F:	drivers/i2c/*/

Hmm, what would be the drawback of extending the meaning of
'drivers/i2c/' to include 'drivers/i2c/*/'? To be honest, I assumed it
works this way already and I'd think I am not alone with that
assumption.

Thanks for looking into this!


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Joe Perches]

On Wed, 2015-10-28 at 00:25 +0100, Wolfram Sang wrote:
> > Another option would be to add another pattern to the
> > I2C section that matches all subdirectories of drivers/i2c/:
> > ---
> >  MAINTAINERS | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index fb7d2e4..c670e1f 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -5104,6 +5104,7 @@ S:	Maintained
> >  F:	Documentation/devicetree/bindings/i2c/
> >  F:	Documentation/i2c/
> >  F:	drivers/i2c/
> > +F:	drivers/i2c/*/
> 
> Hmm, what would be the drawback of extending the meaning of
> 'drivers/i2c/' to include 'drivers/i2c/*/'? To be honest, I assumed it
> works this way already and I'd think I am not alone with that
> assumption.

Some top level directories contain other subsystems.

A maintainer of something like arch/arm doesn't generally
want to be the maintainer of arch/arm/mach-<foo>

You _want_ a wider number of reviewers for patches to any
of those unmaintained subsystems.

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 332 bytes --]

> Some top level directories contain other subsystems.
> 
> A maintainer of something like arch/arm doesn't generally
> want to be the maintainer of arch/arm/mach-<foo>
> 
> You _want_ a wider number of reviewers for patches to any
> of those unmaintained subsystems.

I agree.

+F:	drivers/i2c/*/

makes sense then.


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Joe Perches]

On Wed, 2015-10-28 at 01:05 +0100, Wolfram Sang wrote:
> > Some top level directories contain other subsystems.
> > 
> > A maintainer of something like arch/arm doesn't generally
> > want to be the maintainer of arch/arm/mach-<foo>
> > 
> > You _want_ a wider number of reviewers for patches to any
> > of those unmaintained subsystems.
> 
> I agree.
> 
> +F:	drivers/i2c/*/
> 
> makes sense then.

The listing does become a bit unwieldy when there are
multiple sublevel directories.

It'd probably be nicer syntax to have some ... type
marking, but the number of cases just isn't very high
and most likely not worth the effort to me.

Re: [PATCH] get_maintainer: Don't fallback to git by default

[Lee Jones]

On Wed, 28 Oct 2015, Wolfram Sang wrote:

> > Some top level directories contain other subsystems.
> > 
> > A maintainer of something like arch/arm doesn't generally
> > want to be the maintainer of arch/arm/mach-<foo>
> > 
> > You _want_ a wider number of reviewers for patches to any
> > of those unmaintained subsystems.
> 
> I agree.
> 
> +F:	drivers/i2c/*/
> 
> makes sense then.

As both of my most recent issues have been pertaining to
drivers/i2c/*/ this would solve my current issue.  If anything else
crops up I'll be sure to let you know. ;)

For the line above:
  Acked-by: Lee Jones <lee.jones@linaro.org>

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog