On Feb 26, 2005, at 9:24 PM, Perry E. Metzger wrote:

> I discussed it with Christos and Matt Thomas briefly, and neither
> objected.

I discussed this issue with Christos on the phone today, and I think 
you have grossly mischaracterized the nature of your discussion of the 
topic with him.

> On the ktrace-lwp thing, I held off doing such stuff for several weeks
> on the basis that he was "just about to commit".

So you decided to make it harder in any case, even though there has 
been recent "sync with head" activity on that branch so Nick could test 
and/or fix any stragglers before doing the sync.

That's just totally disrespectful of other people's work, and I find 
your cavalier attitude towards it just appalling.

> On the 3.0 thing, the reason I'm doing whitespace and not __P is
> because whitespace is much safer

"Safer" (noting the "unexpected consequences" that did in fact pop up 
during your anti-whitespace jihad), and totally unnecessary.  I mean, 
you just acknowledged that you were conscious of an impending branch, 
yet you chose to modify hundreds of source files anyway.  Regardless of 
how "safe" the change is, that's just insane.

>  -- you can actually screw stuff up
> with de-__Ping. However, removing trailing whitespace is very low
> risk. I'll resume ANSIfication etc. after the branch happens.

...and I hope you get each and every such change approved by code 
reviewers who have a less myopic view of the state of the source tree.

I am formally raising this issue with core, and I would appreciate it 
if you would cease all of this whitespace-cleansing until they have 
rendered a decision as to the appropriateness of the actions.

-- thorpej