Re: [PATCH 1/4] strip: move driver to staging

[david]

On Mon, 26 Oct 2009, Greg KH wrote:

> On Mon, Oct 26, 2009 at 10:18:20AM -0700, david@xxxxxxx wrote:
> > if someone were to claim 'maintainership' and then do nothing other than
> > complain if someone else were to change an API but not fix this in the
> > process, how would this be different than the current situation?
>
> A person "claiming maintainership" would then be responsible for keeping
> the API up to date and ensuring that the driver worked.  To do that,
> hardware would probably need to be present.

actually, I understood that the person changing the API was responsible 
for making the changes. when did this change?

> Do you have this kind of hardware and are willing to accept ownership of
> this driver?

no, I do not have the hardware, but if there are no bugs reported against 
this driverit would seem that having a 'maintainer' who made absolutly no 
changes to the driver (just allowing API changes by others to be 
implemented) would be the same thing as having no maintainer, but in the 
first case you are willing to have the driver in the kernel, in the other 
you want to rip it out.

it used to be (not that long ago) that when people said that the reason 
they didn't push their driver upstream into the kernel because there 
wasn't that much demand for it, the response was that we wanted drivers 
for everything, no matter how small the user base. I remember seeing posts 
from core developers saying that we had drivers for hardware where there 
were only single digit quantities ever built.

now it appears that you have to have 'enough' users (an amount undefined) 
or a person to specificly take maintainership of the driver to keep it in.


prior to the kernel summit, the criteria for having something moved out of 
the kernel into staging was for fairly significant problems (with a sloppy 
edge of 'or an unreasonable maintinance burden')

I don't think anyone who read that would have thought that 'an 
unreasonable maintinance burden' could be "I don't want to change this 
driver when I change an API"

for old hardware the driver _should_ be static except for API changes. the 
hardware isn't changing, why should the driver.

David Lang
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/