[Ksummit-2012-discuss] [ATTEND] Your upstream maintainer just isn't that into you...

Dan Carpenter dan.carpenter at oracle.com
Tue Jun 26 08:33:54 UTC 2012


On Tue, Jun 26, 2012 at 02:20:13PM +1000, NeilBrown wrote:
> On Mon, 25 Jun 2012 20:44:47 -0700 "H. Peter Anvin" <hpa at zytor.com> wrote:
> 
> > On 06/25/2012 06:44 PM, Dan Williams wrote:
> > > Can we do a better job of bounding the maximum latency for acceptance
> > > or rejection of a patch?  Are there barriers for maintainers asking
> > > for or accepting help, even on a temporary basis?
> > 
> > There shouldn't be.  In fact, anyone can help a maintainer by reviewing 
> > patches and giving them a Reviewed-by: if good and feedback if bad.
> > 
> 
> Maybe the issues are more subtle.
> Is asking for help a sign of weakness?  Is that a problem?
> Does asking for help impose on others?  Does that make us reticent to do it?
> Do we properly show appreciation when help is provided (BTW, thanks Dan for
> recent review of some RAID5 patches while I was on leave - I really
> appreciate it).
> 
> Do we have good role-models of maintainership?  Linus, Andrew Morton, Greg KH
> all do fantastic jobs, but appear a bit like one-man-bands.  Are there good
> examples of load-sharing that we can look to and emulate?
> 

Greg's not a one man show.  Sarah Sharp and Alan Stern review USB
patches.  Several of the trees under staging have active
maintainers.  I help review staging patches.  Alan Cox is stills
reviews tty patches and Jiri Slaby does as well.

> I'd be more than happy to have a co-maintainer (or 2) for md/raid, but the
> various people who have shown an interested and even an aptitude never
> stayed for long.  Did I scare them away by being too competent and not
> sharing enough?
> 

It could be that you're too competent.  Why help someone who is
managing fine on their own?

I'm not a formal maintainer for staging but I do do a bunch of
review.  When a patch is submitted there is a day or two for anyone
to comment before it gets merged.

Staging has a lot of traffic.  Most of it is refactoring, so it's
easy to review even if you don't know that code deeply.  I do
lighten the load by responding to newbies with completely mangled
patches.  I also do spot bugs.  And, of course, I complain about
changelogs and style issues.  Somebody has to do it.

I review kernel-janitor patches as well.  I enjoy helping newbies.
You do see people learn the system and graduate out of staging as
well and that feels good.

regards,
dan carpenter


More information about the Ksummit-2012-discuss mailing list