Re: [patch] checkpatch: relax spacing and line length

[Adrian Bunk <bunk@kernel.org>]

On Sun, Apr 06, 2008 at 06:54:54AM +0200, Jan Engelhardt wrote:
> ===
> total: 0 errors, 0 warnings, 36 lines checked
>
> d27a9f5.diff has no obvious style problems and is ready for submission.
> ===
> commit d27a9f5760fa231ab888f96e27355a001c88b239
> Author: Jan Engelhardt <jengelh@computergmbh.de>
> Date:   Sun Apr 6 06:49:01 2008 +0200
>
>     checkpatch: relax spacing and line length
>
>     We all had the arguments about 80 columns, so here goes a relax.
>     Checking for 95 (or perhaps something better?), but of course we
>     print "80" in the output, because if you happened to get to 95, it's
>     "really time" to break it.

No, it's really time to teach people that checkpatch is *not* a tool for 
janitors.

It's a tool for patch submitters and maintainers that automates a part 
of patch review.

When a patch has a few checkpatch warnings and the submitter can justify 
them (since the code would otherwise look bad) *that is OK*.

But if your driver has over 2000 lines over 80 columns or many lines 
over 95 columns it is not "really time to break it" - it is time to 
check *why* your code has that many that long lines.

>     This also relaxes the tab doctrine, because spaces DO make sense --
>     especially when you view the code with a tab setting of not-8.
>...

In the kernel all tabs are 8 spaces wide.

When you view the code with a different setting that's your fault.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed