On Mon, Sep 15, 2003 at 07:31:54AM +1000, Luke Mewburn wrote:
> On Sun, Sep 14, 2003 at 02:12:35PM -0700, Matthew Jacob wrote:
>   | 
>   | This is a stupid and trivial printout change and in no way affects the
>   | actual behaviour. The pullup process is so expensive in NetBSD that I
>   | was waiting until changes that are more substantial came along to do
>   | that.
> 
> The pullup process is a matter of a developer forwarding a copy of the
> commit message to the appropriate release engineering ticket tracking
> pullup address, and wait for releng to apply it.

I know this. One can infer as much from the (pullup requested in ticket
#whatever) message one sees on -1-6 branch commits. I was merely requesting
that some developer take the time to do so.

> 
> It's not *that* "expensive".  And it has certainly resulted in a
> release branch that is more stable than for previous releases
> when we let developers commit at their whim to the it...

I am not disputing that. But the incendiary response I got from Matthew
Jacob seems to suggest that making the kernel print correct information is
more trouble than its worth. I thought NetBSD was about getting things
right, not doing things in a way that uses the least developer time.

As anyone can guess by the instability in current, it doesn't look like 2.0 is
coming anytime in the next six months, so releng can't have *that* much work
on their hands. I'm willing to wait, I just wanted to see if I could get the
wheel turning on getting that code pulled up. It did take until I mentioned
it for anybody to care enough to make the change in -current, after all. I
figured I was facing a similar situation.

-Sean

-- 
/~\ The ASCII
\ / Ribbon Campaign                   Sean Davis
 X  Against HTML                       aka dive
/ \ Email!