AutoQA: distro congestion?

[Michael Schwendt]

On Thu, 21 Apr 2011 11:50:15 -0700, AW wrote:

> To reiterate this again (thanks, Babe!) we've never claimed the process
> would catch all breakage, or expected it to. The point is to catch as
> much as we can,

In my point of view, "as much as we can" is not what is being done so far
at Fedora. 

What is implemented is a race between people, who are involved in rushing
out updates ("a day without risky upgrades is a boring day"), and other parts
of the Fedora user community, who may need a little bit more time to take
notice of new Test Updates [and possibly even more time to actually run
some tests or start evaluating the new packages].

The update acceptance criteria are half-hearted for some updates and
a hindrance for other updates.

> and the amount of issues we've caught with
> updates-testing is a lot more than 0.

That doesn't say anything about whether this would have been much
different if package maintainers were free to decide on how long to keep
their updates in updates-testing and how long to wait for pos/neg feedback
from testers.