RFC: Page size on PPC/PPC64 builders

[David Woodhouse]

On Mon, 2008-03-03 at 11:17 -0500, Tom "spot" Callaway wrote:
> On Mon, 2008-03-03 at 11:07 -0500, Alan Cox wrote:
> > On Mon, Mar 03, 2008 at 10:17:38AM -0500, Tom spot Callaway wrote:
> > > > architectures. (In fact, the whole distinction between primary and
> > > > secondary architectures is IMHO a bad idea and should be dropped).
> > > 
> > > David, you keep saying this, and its rather unproductive and insulting.
> > 
> > Tom - why is repeatedly pointing out the obvious "insulting". Surely not
> > listening is insulting.
> 
> Its insulting to those of us who worked on the Secondary Architectures
> proposal, to the Fedora Steering Committee who ratified it, and the
> Fedora Board who signed off on it.

No more so than complaints about sysfs are personally insulting to akpm
and Linus.

Your proposal was largely useful, but severely misguided in some ways --
particularly in the way that it leads us to _preserve_ the distinction
between architectures rather than bringing new architectures 'into the
fold' as first-class citizens. And as it stands, those new architectures
don't really gain much at all over what they could have achieved as
standalone efforts.

> If I continuously pointed out in public that your kernel efforts are
> junk, and should be abandoned, I suspect you would be less than pleased.

I've never known Alan take technical criticism personally. Therefore I
believe -- and sincerely hope -- that your suspicion is mistaken.

And I've certainly never suggested that the process of bringing new
architectures into Fedora should be abandoned.

> The split between primary and secondary architectures enables secondary
> architectures to ramp up and grow, without unnecessarily burdening the
> Fedora infrastructure or packaging community.

On this we disagree.

Yes, it is a certain amount of extra work for the Fedora packaging
community when we find real generic bugs in their packages and they are
expected to fix them. No, I don't consider that 'unnecessary burden',
and I don't believe that any competent and conscientious packager should
object to it either. Or should we stop doing any QA at all, just in case
we find bugs and thus add 'unnecessary burden' for our packagers? 

By speaking for the Fedora packagers in such terms, it is you who are
insulting them, in my opinion.

It's not as if it's hard for them to exclude a given architecture and
resubmit the build, in the minority of cases where the issue is actually
arch-specific or even if they just can't be bothered.

>  I've reached out to David and tried to work with him on resolving
> specific technical concerns, and he still reverts back to an attitude
> of badmouthing the entire effort.

I have no intention of badmouthing the entire effort, as you well know.
And we've had productive discussions about technical concerns. It is
very disingenuous of you to pretend otherwise.

-- 
dwmw2