Re: [Qemu-devel] [PATCH] Add support for r6040 NIC

["Edgar E. Iglesias" <edgar.iglesias@gmail.com>]

On Wed, Aug 31, 2011 at 01:48:34PM -0500, Anthony Liguori wrote:
> On 08/31/2011 12:59 PM, Edgar E. Iglesias wrote:
> >On Wed, Aug 31, 2011 at 11:06:11AM -0500, Anthony Liguori wrote:
> >>On 08/31/2011 09:35 AM, malc wrote:
> >>>On Wed, 31 Aug 2011, Anthony Liguori wrote:
> >>>
> >>>>Upper case field names are not okay.  If you think coding style isn't clear,
> >>>>that's a bug in coding style.
> >>>
> >>>Sez hu? Coding style is garbage that should be thrown out of the window.
> >>>As for looking, yeah, i'm looking at usb with it's lovely hungarian
> >>>fields, should we stampede to "fix" it?
> >>>
> >>>If the one who's going to maintain the code is fine with whatever naming
> >>>is used so be it.
> >>
> >>No.  That's how we got into the coding style mess we're in in the
> >>first place.
> >
> >TBH, the codingstyle in QEMU is the least of "problems" we are facing.
> >We've got lack of documentation, lack of tests, lack of contributors,
> >etc, etc. IMO, those bring codingstyle issues into the pretty much
> >neglectable space.
> 
> I don't think we lack contributors.  Documentation and tests are
> really about discipline.  If we can't even be bothered to maintain
> consistency in variable naming, do you really expected that we can
> be disciplined in writing documentation and tests?

Yes I do. It's not white and black, it's not about making the code
completely inconsistent or 100 consistent. It's about find a level
of consistency that is acceptable and doesn't cost too much to
maintain.


> Is the next argument going to be that every subsystem should be able
> to have its documentation in it's preferred natural language such
> that the documentation for the block layer is in Esperanto?

Heh, I think you missunderstood me. I'm talking about details that
dont matter. Like, would you like to enforce American english or UK
english? More that kind of question.

> Consistent coding style makes the tree a single code base, instead
> of a bunch of independent islands.  This encourages sharing code and
> ideas across subsystems.  Too often, we reproduce the same thing and
> over again in different subsystems (and even different machine
> architectures).

I dont think the latter has to do with codingstyle consistensy. But
I agree with the former.

Cheers