Prev: [tip:perf/urgent] kprobes/x86: Fix removed int3 checking order
Next: [PATCH 1/2 V3] hwmon: (coretemp) Detect the thermal sensors by CPUID
From: Wolfram Sang on 11 May 2010 03:50
On Tue, May 11, 2010 at 08:36:36AM +0200, florian(a)mickler.org wrote:
> This changes the default of the option --git-all-signature-types to be
> disabled by default.
>
> The effect being, that only certain (currently Signed-Off-By:,
> Acked-By: and Reviewed-By:) tags are used to get adresses of potential
> maintainers.
>
> Motivated is this change by the desire to not 'spam' people unnecessary:
> A Tested-By or a Reported-By is not ment as a hint that those people
> want to/are able to review patches to the code in question.
>
> In a quest to find resilient statistics for this i came up with this:
>
> I produced a list of all the tag-signers not already covered
> with a signed-off/acked/reviewed tag somewhere in the last year of git history.
>
> Those were 650 addresses of "assumed non-developers".
>
> And to check if those "assumed non-developers" are professional
> testers/reporters worth cc'ing, i then counted their total appearences
> in the git log:
>
> 469 were mentioned only once.
> 123 were mentioned twice.
> 38 three times
> 8 four times
> 5 six times
> 5 five times
> 1 eight times
> 1 fourteen times
>
> I believe this supports my thesis, that the ''non-maintainer-tags'' are
> not actively useful for patch-review. (except probably the guy
> mentioned fourteen times...)
>
> But of course one could also find arguments to poke holes in this
> statistics, for example does this statistic not include code-locality:
> A tested-by on a patch that touches some specific piece of
> code can be more worth than a signed-off in another part of the tree.
>
> But... let's play it safe and let's err on the "safe" side
> meaning to not spam those people when in doubt. We already have the
> signed-off's and Maintainers file. So this should be ok. And if need be,
> the maintainers can always forward the patch.
>
> [i probably could make a diploma thesis out of this changelog :)]

:D

>
> Signed-off-by: Florian Mickler <florian(a)mickler.org>

I second this reasoning:

Acked-by: Wolfram Sang <w.sang(a)pengutronix.de>

--
Pengutronix e.K. | Wolfram Sang |
Industrial Linux Solutions | http://www.pengutronix.de/ |
 | 
Pages: 1
Prev: [tip:perf/urgent] kprobes/x86: Fix removed int3 checking order
Next: [PATCH 1/2 V3] hwmon: (coretemp) Detect the thermal sensors by CPUID